Technology Field

The present invention relates to a process for testing the 3D morphology of particles.

Background

Analysis of particle characterization plays an important role for interface and colloid science. There’re several methods to characterize and describe morphology of particles, such as scanning electron microscopy (SEM), transmission electron microscopy (TEM) and Atom force microscopy (AFM). However, these characterization methods are kinds of number-based counting methods which tested from the relatively low numbers of particles which will bias the understanding for particle morphology. In addition, conventional image-based analysis techniques cannot obtain adequate 3-dimensional morphological information for both inside and outside of particles. Therefore, there’s a strong need to develop an efficiency method to obtain and statically characterize 3 dimensional morphological of a large amount of particles.

Summary of the Invention

One object of the present invention is to provide a process for testing the 3D morphology of particles, which can efficiently and statically characterize 3 dimensional morphological of a large amount of particles.

It has been surprisingly found that the above objects can be achieved by following embodiments:

1. A process for testing the 3D morphology of particles which comprises following steps:

(i) optionally coating the particle surface with a metal thin layer and/or embedding the particle in epoxy resin for tomographic scanning contrast enhancement;

(ii) tomographic scanning the particles to create multiple tomographic slices;

(iii) subjecting the multiple tomographic slices to image processing;

(iv) performing single particle segmentation; and

(v) carrying out statistic and analysis of geometric parameters of every single particle.

2. The process according to item 2, wherein the tomographic scanning in step (ii) is carried out by XRM, MRI or OCT.

3. The process according to item 1 or 2, wherein the number of slices created in step (ii) is in the range from 20 to 10,000, preferably from 50 to 8,000, more preferably from 100 to 5,000.

4. The process according to any of items 1 to 3, wherein the voxel size of the slice is in the range from 0.1 ³ to 100 ³ pm ³ , preferably from 0.2 ³ to 50 ³ pm ³ , more preferably from 0.4 ³ to 10 ³ pm ³ . 5. The process according to any of items 1 to 4, wherein the segmentation in step (iv) comprises

(iv-1) binarizing the tomographic slices by a threshold method;

(iv-2) identifying the free space part and the non-free space part; and (iv-n) isolating particle one by one.

6. The process according to any of items 1 to 5, wherein the segmentation in step (iv) comprises

(iv-1) binarizing the tomographic slices by a threshold method;

(iv-2) identifying the free space part and the non-free space part;

(iv-3) applying a local 3D thickness calculation on the non-free space part;

(iv-4) for each voxel in the non-free space part applying a threshold (b) on the local calculation result to remove the particle voxel which has smaller 3D thickness value than the threshold (b); and

(iv-n) isolating particle one by one.

7. The process according to any of items 1 to 6, wherein step (iv-1) is carried out by a machine learning model.

8. The process according to item 6 or 7, wherein the threshold (b) in step (iv-4) is 0-10%, or 2- 10% of maximum of cross-section area of the scanning volume.

9. The process according to any of items 1 to 8, wherein the geometric parameters comprise surface area, sphericity, bounding box, particle volume, orientation and any combination thereof.

10. The process according to item 9, wherein the sphericity comprises sphericity A and/or sphericity B:

Sphericity A = Surface area of volume equivalent sphere of particle I Surface area of particle, Sphericity B = Radius of volume equivalent sphere of particle I Radius of circumscribed sphere of particle.

11. The process according to any of items 1 to 10, wherein the particle is formed from polymer, metal, inorganic compound, organic compound or any combination thereof.

12. The process according to any of items 1 to 11 , wherein the particle is derived from polymer foams, preferably from elastomeric polymer foams, more preferably from the foamed TPU. 13. The process according to any of items 1 to 12, wherein the average particle size D50 of the particles is no more than 10 mm, preferably no more than 5 mm, more preferably no more than 1 mm, for example in the range from 20 to 800 pm.

14. The process according to any of items 1 to 13, wherein the bulk density of the particles is in the range from 0.15 to 0.45 g/cm ³ , preferably from 0.18 to 0.35 g/cm ³ , more preferably from 0.18 to 0.3 g/cm ³ .

15. The process according to any of item 1 to 14, which is used in 3D printing, coating and catalyst.

According to the present invention, the process of the present invention can efficiently and quantitively characterize 3 dimensional morphological of a large amount of particles.

Description of the Drawing

Figure 1 shows the phericity A histogram for 2 samples.

Figure 2 shows the scattering plot for sphericity A vs particle volume.

Embodiment of the Invention

The undefined article “a”, “an”, “the” means one or more of the species designated by the term following said article.

In the context of the present disclosure, any specific values mentioned for a feature (comprising the specific values mentioned in a range as the end point) can be recombined to form a new range.

One aspect of the present invention is directed to a process for testing the 3D morphology of particles which comprises following steps:

(i) optionally coating the particle surface with a metal thin layer and/or embedding the particle in epoxy resin for tomographic scanning contrast enhancement;

(ii) tomographic scanning the particles to create multiple tomographic slices;

(iii) subjecting the multiple tomographic slices to image processing;

(iv) performing single particle segmentation; and

(v) carrying out statistic and analysis of geometric parameters of every single particle.

The thickness of the metal thin layer in step (i) can be in the range from 8 nm to 50 nm (for example 10 nm, 15 nm, 20 nm, 25 nm, 30 nm, 40 nm or 50 nm), or from 12 to 30 nm. The metal can be platinum. The tomographic scanning in step (ii) can be carried out by the experimental techniques such as x-ray tomography (including micro x-ray tomography and nano x-ray tomography), Focused Ion Beam Scanning Electron Microscopy, Nuclear Magnetic Resonance and Neutron tomography.

In an embodiment, the tomographic scanning in step (ii) is carried out by XRM, MRI or OCT.

The number of slices created in step (ii) can be in the range from 20 to 10,000 (for example 20, 50, 80, 100, 200, 500, 800, 1000, 2000, 4000, 5000, 6000, 8000 or 10,000), preferably from 50 to 8,000, more preferably from 100 to 5,000, or from 200 to 4000, or from 500 to 3000.

The volume of slice is typically represented by regular 3D volume elements referred to in the art as “voxels”. Generally, each voxel is cubic, having sides of equal length in the x, y, and z directions. In step (ii), the voxel size of the slice is in the range from 0.1 ³ to 100 ³ pm ³ (for example 0.1 ³ pm ³ , 0.2 ³ pm ³ , 0.5 ³ pm ³ , 0.6 ³ pm ³ , 0.7 ³ pm ³ , 0.8 ³ pm ³ , 1 ³ pm ³ , 2 ³ pm ³ , 5 ³ pm ³ , 8 ³ pm ³ , 10 ³ pm ³ , 12 ³ pm ³ , 15 ³ pm ³ , 18 ³ pm ³ , 20 ³ pm ³ , 30 ³ pm ³ , 40 ³ pm ³ , 50 ³ pm ³ , 60 ³ pm ³ , 70 ³ pm ³ , 80 ³ pm ³ , 90 ³ pm ³ or 100 ³ pm ³ , preferably from 0.2 ³ to 50 ³ pm ³ , more preferably from 0.4 ³ to 10 ³ pm ³ or from 0.4 ³ to 8 ³ pm ³ , or from 0.4 ³ to 5 ³ pm ³ , or from 0.4 ³ to 2 ³ pm ³ .

Each voxel has an associated numeric value, or amplitude, that represents the relative material properties of the imaged sample. The range of these numeric values or amplitude, commonly known as the grayscale range, the granularity of the values (e.g., 8 bit or 16 bit values), and the like. For example, the voxels of a typical x-ray tomographic image volume represented by 16 bit data values can have amplitudes ranging from 0 to 63535.

In step (iii), voxels may have their original amplitude value modified, for example by image processing such as artifact reduction or noise filtering, to minimize artifacts and noise generated during acquisition. For example, all the slices are normalized for brightness and contrast. Then an Kuwahara filter (for example 3*3) was applied on each slice for noise smoothing.

Segmentation is generally useful for performing feature identification, and can be performed by way of an automated numerical process, or by hand-picking values. Either approach involves assessing the characteristics of an image, derivative, or constructed volume, for example the characteristics of voxel amplitude, voxel amplitude connectivity or disconnectedness, or shape of connected or disconnected amplitude bodies. One example of a segmentation process is referred to in the art as thresholding (a). In this context, thresholding (a) is commonly utilized to separate free space part from non-free space part within an image volume. A threshold value is chosen within the voxel amplitude range, such that voxels having amplitudes below this threshold value are quantized to a specific numeric value denoting free space part, while voxels having amplitudes above that threshold are quantized to another numeric value denoting non-free space part. In this instance, thresholding (a) will convert a grayscale image volume to a derivative volume in which each voxel has one of two possible numeric values, commonly 0 and 1 (binarizing). Thresholding (a) can be applied any number of times, or using any number of different threshold values, to denote various features within a grayscale image.

In an embodiment, a machine learning model (random forest model or ll-net model) was trained to perform the single particle segmentation. For example, machine learning model (random forest model or ll-net model) was trained to segment the slice into particle phase and air phase (binary the volume space).

In an embodiment, the segmentation in step (iv) comprises (iv-1) binarizing the tomographic slices by a threshold method; (iv-2) identifying the free space part and the non-free space part; and (iv-n) isolating particle one by one.

In an embodiment, step (iv-1) is carried out by the machine learning model, or step (iv-1) and step (iv-2) are carried out by the machine learning model.

In an embodiment, the segmentation in step (iv) comprises

(iv-1) binarizing the tomographic slices by a threshold method;

(iv-2) identifying the free space part and the non-free space part;

(iv-3) applying a local 3D thickness calculation on the non-free space part;

(iv-4) for each voxel in the non-free space part applying a threshold (b) on the local calculation result to remove the particle voxel which has smaller 3D thickness value than the threshold (b); and

(iv-n) isolating particle one by one.

In step (iv-3), a local 3D thickness calculation is applied on the result of the non-free space part. Each voxel in the non-free space part would get a number from the calculation to represent the cross section area at that voxel position. In step (iv-4), a threshold (b) is applied on the local thickness calculation result to remove the particle voxel which has smaller 3D thickness value than the threshold (b). The threshold (b) in step (iv-4) is 0-10% (for example 1%, 2%, 5%, 8% or 10%), or 2 to 10% of maximum of crosssection area of the scanning volume (i.e., the particle). In a specific example, a threshold (b) for 10 * 0.7 ² pm ² was applied on the local thickness calculation result to generate particle volume segmentation.

Sometimes two or more or several small particles are connected to each other through a slender belt, especially when the particles are polymer materials. If said particles are regarded as one particle, it will interfere with the results. The interference can be efficiently removed by applying a local 3D thickness calculation on the non-free space part and applying a threshold (b) on the local calculation result for each voxel in the non-free space part to remove the particle voxel which has smaller 3D thickness value than the threshold (b).

Step (iv-n) comprises isolating particle one by one. For each none-free part volume, if all connected voxel belongs to the free part volume, it would be considered as a single isolated particle.

In step (v), statistic and analysis of geometric parameters of every single particle is carried out. The geometric parameters can comprise surface area, sphericity, bounding box, particle volume, orientation and any combination thereof.

In this regard, each particle volume (i.e., the actual volume of the particle), the surface area of particle, bound (i.e., the diameter of the circumscribed sphere of particle), bounding box and orientation are calculated. Such statistic and analysis can be carried out by opensource Porespy package.

In an embodiment, sphericity can comprise sphericity A and/or sphericity B, which can be calculated based on following equations:

Eq.1: Sphericity A = Surface area of volume equivalent sphere of particle I Surface area of particle

Eq.2: Sphericity B = Radius of volume equivalent sphere of particle I Radius of circumscribed sphere of particle.

According to the present invention, radius of volume equivalent sphere of particle is the radius of a sphere having volume equal to actual volume of the particle.

The surface area of volume equivalent sphere of particle can be calculated as follows: 4 x TT X (radius of volume equivalent sphere of particle) ² . With respect to the circumscribed sphere of a particle, the diameter of the circumscribed sphere of the particle is the longest dimension of the particle, i.e., the straight line distance between the two farthest points of the particle.

In an embodiment, at least one, preferably both of average sphericity A and average sphericity B of TPE particles is no more than 0.55, for example 0.5, 0.45, 0.4, 0.35 or 0.3, preferably no more than 0.5.

In an embodiment, the average sphericity A of particles is no more than 0.55, preferably no more than 0.5, no more than 0.45 or no more than 0.4 and/or average sphericity B of particles is no more than 0.5, preferably no more than 0.4, no more than 0.35 or no more than 0.3.

According to the present invention, the term “average sphericity” as used in this disclosure means number average sphericity.

Usually, at least one, preferably both of the average sphericity A and average sphericity B of TPE particles are above 0.05, above 0.08 or above 0.1. In an embodiment, at least one, preferably both of the average sphericity A and average sphericity B of particles are in the range from 0.05 to 0.6, preferably from 0.08 to 0.55, or from 0.1 to 0.5, or from 0.1 to 0.45.

In an embodiment, the average particle size D50 of the particles can be in the range from 10 pm to 1 mm, for example 10 pm, 20 pm, 50 pm, 80 pm, 100 pm, 150 pm, 200 pm, 250 pm, 300 pm, 350 pm, 400 pm, 500 pm, 600 pm, 700 pm, 800 pm, 900 pm or 1 mm, preferably from 10 to 800 pm, from 10 to 600 pm, from 10 to 400 pm, from 10 to 250 pm, or from 20 to 800 pm, from 20 to 600 pm, from 20 to 400 pm, from 20 to 250 pm, or from 30 to 800 pm, from 30 to 600 pm, from 30 to 400 pm, from 30 to 250 pm, or from 40 to 800 pm, from 40 to 600 pm, from 40 to 400 pm, from 40 to 250 pm, or from 50 to 800 pm, from 50 to 600 pm, from 50 to 400 pm, or from 50 to 250 pm.

The average particle size D50 can be tested according to ISO 13320-1.

According to the present invention, the process of the present invention can test the 3D morphology of at least 200 particles (for example 300, 400, 500, 800, 1000, 2000, 5000, 10,000 particles), or at least 400 particles, or from 200 to 10,000 particles, or from 400 to 5000 particles at the same time.

In an embodiment, the particle is formed from polymer, metal, inorganic compound, organic compound or any combination thereof. In an embodiment, the particle is derived from polymer foams, preferably from elastomeric polymer foams, more preferably from the foamed TPU.

The elastomeric polymer can include for example thermoplastic elastomer (TPE). The TPE in general contain polymeric blocks (usually referred to as "hard" blocks or A blocks) and amorphous polymeric blocks (usually referred to as "soft" blocks or B blocks). Each soft block is linked to at least two hard blocks. Melting or softening of the hard blocks permits viscous flow of the polymeric chains, resulting in thermoplastic behavior.

The TPE can be selected from thermoplastic polyurethane elastomer (TPU), thermoplastic copolyester elastomer (TPC), thermoplastic styrene elastomer (TPS), polyether block amide (PBA), thermoplastic vulcanite (TPV), thermoplastic polyolefin (TPO), and combinations thereof, preferably selected from thermoplastic polyurethane elastomer (TPU), polyether block amide (PBA), and combination thereof.

Thermoplastic polyurethane elastomer (TPU):

TPUs and processes for their production are well known. By way of example, TPUs can be produced via reaction of (a) isocyanates with (b) compounds reactive toward isocyanates and having a molar mass of from 500 to 10000 and, if appropriate, (c) chain extenders having a molar mass of from 50 to 499, if appropriate in the presence of (d) catalysts and/or of (e) conventional auxiliaries and/or conventional additives.

According to the invention, the TPU has a weight average molecular weight of at least 0.1 x10 ⁶ g/mol, preferably at least 0.4x10 ⁶ g/mol and in particular greater than 0.6x10 ⁶ g/mol. The upper limit to the weight average molecular weight of the TPU is generally determined by the processability and also the desired property spectrum. The number average molecular weight of the TPU is preferably not more than 0.8x10 ⁶ g/mol. The average molecular weights indicated above for the TPU and also for the formative components (a) and (b) can be determined by means of gel permeation chromatography.

Thermoplastic Polyester Elastomers:

TPE may alternatively be a thermoplastic polyester elastomer, also known as a TPC. Thermoplastic polyester elastomer comprises polyesters segments and polyether segments. The polyesters segments can be produced by the reaction of dicarboxylic acid and derivative thereof (such as terephthalate) and diols (such as butanediol). The polyether segments can comprise polyalkylene (ether) glycol or polyol.

Thermoplastic polystyrene elastomers:

TPE may alternatively be a thermoplastic polystyrene elastomer, also known as TPS. Thermoplastic polystyrene elastomers are typically based on A-B-A type block structure where A is a hard phase and B is an elastomer. Usually, the thermoplastic polystyrene elastomer comprises ethylene, propylene, butadiene, isoprene units or combination thereof. In an embodiment, the thermoplastic polystyrene elastomer is a styrene ethylene butylene styrene block copolymer. In an embodiment, the thermoplastic polystyrene elastomer is a poly(styrene-butadiene-styrene), a poly(styrene-ethylene-co-butylene-styrene), a poly(styrene-isoprene-styrene), any copolymer thereof, and any blend thereof. Non-limiting examples of thermoplastic polystyrene elastomers are Kraton D and Kraton G.

Thermoplastic Vulcanizate Elastomers: TPE may alternatively be a thermoplastic vulcanate elastomer, also known as a TPV. A nonlimiting example of a thermoplastic vulcanate elastomer is Santoprene from ExxonMobil.

Polyether block amides:

The thermoplastic elastomer may alternatively be a polyether block amide, also known as a PBA. Suitable PBA may be obtained by copolymerization of a cyclic lactam with a dicarboxylic acid and a diamine to prepare a carboxylic acid-functional polyamide block, followed by reaction with a polyether diol (polyoxyalkylene glycol).

Cyclic lactam can comprise e-caprolactam, pyrrolidone or w-laurolactam or combination thereof.

Dicarboxylic acid can comprise aromatic dicarboxylic acids including, for example, phthalic acid, iso- and terephthalic acid or esters thereof; aliphatic dicarboxylic acids including, for example, cyclohexane- 1 ,4-dicarboxylic acid, adipic acid, sebacic acid, azelaic acid and decanedicarboxylic acid as saturated dicarboxylic acids, and maleic acid, fumaric acid, aconitic acid, itaconic acid, tetrahydrophthalic acid and tetrahydroterephthalic acid as unsaturated aliphatic dicarboxylic acids.

Diamine can have the general formula H ₂ N — R” — NH ₂ where R” may be aromatic and aliphatic. R” may have 2 to 30, preferably 2 to 20 or 2 to 16 carbon atoms. Diamine can comprise ethylenediamine, tetramethylenediamine, pentamethylenediamine, hexamethylenediamine, or decamethylenediamine, 1 ,4-cyclohexanediamine, m-xylylenediamine and combination thereof.

Polyether diol can comprise polyethylene glycol (PEG), polypropylene glycol (PPG) or polypropylene ether glycol (PPEG), polytetramethylene glycol (PTMG or PTHF), polytetramethylene ether glycol, and combinations thereof. Polymerization may be carried out, for example, at temperatures of from 180 °C to 300 °C. A non-limiting example of a polyether block amide is Ves- tamid E from Evonik.

Thermoplastic Polyolefin Elastomers:

The thermoplastic elastomer may alternatively be a thermoplastic polyolefin elastomer, also known as a TPO. A non-limiting example of a thermoplastic polyolefin elastomer is Engage from Dow.

Process of preparing the polymer particles

In an embodiment, the polymer particles can be prepared by pulverizing the polymer foam.

According to a preferred embodiment, the temperature of the polymer foam during the pulverization is lower than the T _g of the polymer by at least 20 °C (for example 20 °C, 40 °C, 60 °C, 80 °C, or 100 °C), preferably by at least 40 °C or by at least 60 °C. In an embodiment, the temperature of the polymer foam during the pulverization is lower than -20 °C, or lower than - 50 °C, or lower than - 80 °C. In an embodiment, the polymer foam is treated with liquid nitrogen before or during the pulverization. A person skilled in the art can determine the time for treating the polymer foam by liquid nitrogen as long as the polymer foam is sufficiently cooled.

According to the present invention, the process can further comprise sieving the pulverized material. The polymer particles with the target particle size can be obtained via sieving.

According to the present invention, the polymer foam can be crushed before pulverization to obtain small particles having average particle size less than 10 cm, preferably less than 5 cm.

According to the present invention, the polymer particles are derived from the elastomeric polymer foams, especially the foamed TPE material. TPEs are those as described above.

To effect foaming, it is possible to use blowing agents. Suitable blowing agents are, for example, low-boiling liquids. Liquids which are inert toward the organic polyisocyanate and preferably have a boiling point below 200°C are suitable.

In an embodiment, the polymer is foamed with the supercritical fluid. The foaming with supercritical fluid can be carried out by following steps:

(1) treating the polymer with the supercritical fluid at a foaming temperature, and

(2) foaming the treated polymer obtained in step (1) to produce a polymer foam.

The supercritical fluid can be selected from carbon dioxide, water, Ci-Ce-alkane, ethylene, propylene, methanol, ethanol, acetone, nitrogen, or a combination thereof, preferably carbon dioxide.

In an embodiment, the average pore size of the polymer foam is in the range from 20 to 800 pm (for example 20 pm, 30pm, 50 pm, 80 pm, 100 pm, 150 pm, 200 pm, 300 pm, 500 pm, 600 pm, or 800 pm), preferably from 30 to 600 pm or from 40 to 500 pm.

The process of the present invention can be used in the field of 3D printing, coating and catalyst. The process of the present invention can efficiently and quantitively characterize 3 dimensional morphological of a large amount of particles. The particles with the desired 3 dimensional morphological can be used in for example 3D printing, the production of coating and the production of catalyst.

Examples

Materials

- ET-4: Prepared by grinding of foamed TPU 13011 pellet from BASF (BASF commercial product: Infinergy X1125-130 II, bulk Density of the pellet is 0.12-0.14 g/cm ³ , average pellet weight is 26 mg and average pore size is 130 pm), the average particle size of the powders is 160 pm, the bulk density of the powders is 0.18g/cm ³ . The EP-4 Powders were prepared by cryogenic grinding under -196 °C, grinding speed 100m/s (linear velocity) and sieved to required size. - FS-1088A: TPU powder from Farsoon, Bulk density of FS-1088A: 0.56g/cm3, average particle size D50 of FS-1088A: 83 pm.

Example 1 :

Powder samples were first coated with 20nm platinum layer by sputtering method to enhance the X-ray contrast and dispersion.

The 3D morphology information of powder was measured by Zeiss Xradia 610 Micro CT. Xray source: 50.13 (kV), 4W. Specifically, Micro CT scanning created about 1000 slices of the sample with 0.7 ³ pm ³ voxel size. First, all the slices were normalized for brightness and contrast. Then a 3*3 Kuwahara filter was applied on each slice for noise smoothing. A machine learning model (random forest model) was trained to segment the slice into polymer phase and air phase (binary the volume space). 3D local thickness calculation was applied on the polymer phase segmentation result. A threshold for 10 * 0.7 ² pm ² was applied on the local thickness calculation result to generate single isolated particle. If such step was ignored, the contacted neighbour particles would be identified as one particle. For example, in sample “FS-1088A”, the total polymer volume was 10106952.17 |im ³ . If 3D local thickness threshold was included, the average volume would be 18647.51 |im ³ , and the particle number would be 542. If such step was ignored, the particle number would be only 12 and average volume would be 842246.01 |im ³ , which was obvious wrong result.

For each particle 3D volume (i.e., the actual volume of the powder), the surface area of powder, bound (i.e., the diameter of the circumscribed sphere of powder) were calculated by opensource Porespy package. The sphericity A of powders was calculated based on following equations: Sphericity A = Surface area of volume equivalent sphere of powder I Surface area of powder

The results were shown in Figures 1 to 2. Figure 1 showed the phericity A histogram for 2 samples, wherein the X axis was the sphericity and Y axis was the paticle counts number. Figure 2 showed the scattering plot for sphericity A vs particle volume, wherein the X axis was the particle volume and the Y axis was the sphericity A.