POLARIMETER WITH MULTIPLE INDEPENDENT TUNABLE CHANNELS AND METHOD FOR MATERIAL ORIENTATION IMAGING

Abstract

A polarimeter and a method of analyzing and imaging microstructural material orientation of a sample are disclosed. The polarimeter, which is a partial Mueller-matrix polarimeter (pMMP), accesses multiple independent polarization channels by employing two independent polarization modulators configured to switch serially among multiple independent settings, wherein the combination of the settings of the first and second polarization modulators defines an independent polarization channel, and wherein an imaging detector produces a set of spatially registered images that are synchronized with the channels formed by the polarization modulators; and wherein a processor connected with a memory executes a classification algorithm stored in the memory that maps the set of images to one or more material orientation images by mapping the set of values for each detector pixel corresponding to the set of spatially registered images to a value of material orientation at each pixel coordinate using a machine-learning model, an electrodynamic model, or a combination thereof. The invention can thereby create material microstructural orientation images of diverse anisotropic materials, for instance polymer domains, fiber bundles or plys, and metallic crystalline grains.

Claims

1. A polarimeter for producing one or more material orientation images of a sample using multiple independent polarization channels comprising: a source of controlled electromagnetic radiation that produces a beam that propagates along a path terminating at an imaging detector with the sample positioned there between; a first polarization modulator positioned in the path preceding the sample and configured to switch serially among multiple independent settings; an electromagnetic-radiation collector positioned to direct electromagnetic radiation reflected from or transmitted by the sample to a second polarization modulator independent of the first polarization modulator; the second polarization modulator, configured to switch serially among multiple independent settings, wherein the combination of the settings of the first and second polarization modulators defines an independent polarization channel; the imaging detector, positioned to receive electromagnetic radiation from the sample transmitted through the second polarization modulator, wherein the imaging detector comprises pixels and produces a set of images that are spatially registered and synchronized with the channels formed by the first and second polarization modulators; and a processor connected with a memory, wherein the processor is configured to execute a classification algorithm stored in the memory that maps the set of spatially registered images to the one or more material orientation images by mapping a set of values for each detector pixel corresponding to the set of spatially registered images to a value of material orientation at each pixel coordinate using a model.

2. The polarimeter of claim 1 wherein the multiple independent polarization channels comprise at least three independent polarization channels.

3. The polarimeter of claim 1 wherein the setting of the first polarization modulator and the setting of the second polarization modulator are tunable.

4. The polarimeter of claim 1 wherein the model is a machine-learning algorithm trained on a database of Mueller matrices of samples with known material orientations.

5. The polarimeter of claim 1 wherein the model is an electrodynamic model.

6. The polarimeter of claim 1 wherein the sample is comprised of crystals and the orientation images are crystallographic-orientation images.

7. The polarimeter of claim 6 wherein the crystals are uniaxial crystals and the crystallographic-orientation images are c-axis images.

8. The polarimeter of claim 6 wherein the sample is metallic.

9. The polarimeter of claim 8 wherein the crystals are isotropic cubic crystals previously subjected to heat-tinting to produce anisotropic metal oxides.

10. The polarimeter of claim 8 wherein the sample is subjected to an external magnetic field.

11. The polarimeter of claim 1 wherein the sample is a diffuse reflector.

12. The polarimeter of claim 11 wherein the sample is curved or otherwise not flat.

13. The polarimeter of claim 1 wherein the sample is reflective and the polarimeter is arranged in a bistatic geometry with an arbitrary bistatic angle.

14. The polarimeter of claim 1 wherein the sample is reflective and the polarimeter is arranged in a monostatic geometry utilizing a beamsplitter.

15. The polarimeter of claim 14 wherein the polarimeter is packaged as a module that can be inserted into a conventional microscope.

16. The polarimeter of claim 1 wherein the polarimeter, excluding the sample assembly, is mounted on a tripod or other transportable platform.

17. A method of analyzing material orientation of a sample using a polarimeter of claim 1 having multiple independent polarization channels comprising the steps of: producing a beam from a source of controlled electromagnetic radiation that propagates along a path terminating at an imaging detector with the sample positioned there between; positioning a first polarization modulator configured to switch serially among multiple independent settings in the path preceding the sample and; positioning an electromagnetic-radiation collector to direct electromagnetic radiation reflected from or transmitted by the sample to a second polarization modulator independent of the first polarization modulator wherein the second polarization modulator is configured to switch serially among multiple independent settings, wherein the combination of the settings of the first and second polarization modulators define an independent polarization channel; positioning the imaging detector to receive the electromagnetic radiation from the second polarization modulator, wherein the detector comprises pixels and produces a set of images that are spatially registered and synchronized with the channels formed by the first and second polarization modulators; and executing a processor connected with a memory, wherein the processor is configured to execute a classification algorithm stored in the memory that maps the set of spatially registered images to one or more material orientation images by mapping a set of values for each detector pixel corresponding to the set of spatially registered images to a value of material orientation at each pixel coordinate using a model.

18. The method of claim 17 wherein the multiple independent polarization channels comprise at least 3 independent polarization channels.

19. The method of claim 17 wherein the setting of the first polarization modulator and the setting of the second polarization modulator are tunable.

20. The method of claim 17 wherein the model is a machine-learning algorithm trained on a database of Mueller matrices of samples with known material orientations.

21. The method of claim 17 wherein the model is an electrodynamic model.

22. The method of claim 17 wherein the sample is comprised of crystals and the orientation images are crystallographic-orientation images.

23. The method of claim 22 wherein the crystals are uniaxial crystals and the crystallographic-orientation images are c-axis images.

24. The method of claim 22 wherein the sample is metallic.

25. The method of claim 24 wherein the crystals are isotropic cubic crystals previously subjected to heat-tinting to produce anisotropic metal oxides.

26. The method of claim 24 wherein the sample is subjected to an external magnetic field.

27. The method of claim 17 wherein the sample is a diffuse reflector.

28. The method of claim 27 wherein the sample is curved or otherwise not flat.

29. The method of claim 17 wherein the sample is reflective and the polarimeter is arranged in a bistatic geometry with an arbitrary bistatic angle.

30. The method of claim 17 wherein the sample is reflective and the polarimeter is arranged in a monostatic geometry utilizing a beamsplitter.

31. The method of claim 17 wherein the polarimeter is packaged as a module that can be inserted into a conventional microscope.

32. The method of claim 17 wherein the polarimeter, excluding the sample assembly, is mounted on a tripod or other transportable platform.

33. A method of producing one or more material orientation images of a sample comprising: collecting a set of three or more polarized images using a polarimeter of claim 1 tuned to three or more corresponding channels; combining the polarized images to form orientation feature images that correspond to material orientation as described by a model.

34. The method of claim 33 wherein the model is a machine-learning algorithm trained on a database of Mueller matrices of samples with known material orientations.

35. The method of claim 33 wherein the model is an electrodynamic model.

36. The method of claim 33 wherein the sample is comprised of crystals and the orientation images are crystallographic-orientation images.

37. The method of claim 36 wherein the crystals are uniaxial crystals and the crystallographic-orientation images are c-axis images.

38. The method of claim 36 wherein the sample is metallic.

39. The method of claim 38 wherein the crystals are isotropic cubic crystals previously subjected to heat-tinting to produce anisotropic metal oxides.

40. The method of claim 38 wherein the sample is subjected to an external magnetic field.

41. The method of claim 33 wherein the sample is a diffuse reflector.

42. The method of claim 41 wherein the sample is curved or otherwise not flat.

Description

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0016] The accompanying drawings, which are incorporated into and form a part of the specification, illustrate one or more embodiments of the present invention and, together with the description, serve to explain the principles of the invention. The drawings are only for the purpose of illustrating one or more embodiments of the invention and are not to be construed as limiting the invention. In the drawings:

[0017] FIG. 1A-B is an illustration of an embodiment of the partial Mueller-matrix polarimeter (pMMP) of the current invention, featuring polarization modulators comprised of polarization crystals mounted in motorized rotary stages, in reflection (A) and transmission (B) configurations.

[0018] FIG. 2 is an illustration of a hexagonally close-packed (HCP) crystal unit cell in various orientations relative to the sample surface, which is depicted as a gray disk; the crystal c-axis is indicated by a solid line and illustrated in the orientations A: (,)=(0,0), B: (,)=) (0,45), C: (,)=(45,45), D: (,)=(90,45), and E: (,)=(180,90).

[0019] FIG. 3A-B is an illustration of grayscale crystal-orientation images of a titanium-alloy sample produced by an embodiment of the current invention; (A): c-axis plunge angle () image in which =0 is black and =90 is white, and (B) c-axis trend angle () image.

[0020] FIG. 4 is a cluster diagram that illustrates mapping from crystal orientation, for the case of a commercially-pure titanium sample, to selected polarized-reflectivity features specified by a pattern-recognition algorithm, wherein the orientations of the titanium crystals are indicated by an EBSD inverse pole figure and the reflectivity of a silver mirror is included for reference.

DETAILED DESCRIPTION OF THE INVENTION

[0021] The partial Mueller-matrix polarimeter (pMMP) of the current invention was introduced in U.S. Pat. Appln. Pub. No. 2019/0073561 and is further described with reference to FIG. 1. The pMMP used to perform material orientation imaging comprises a source of controlled electromagnetic radiation (EMR), preferably a laser (1), positioned in a path with an imaging detector (2) and a sample (3) positioned there between. The pMMP further comprises expansion and collimation optics (4) to produce an expanded and approximately collimated beam (5) that illuminates the sample. The illumination beam can be between 1-2 inches in diameter or larger to realize a large instantaneous FOV, and even larger images can be achieved by translating samples under the illumination beam, employing translation stages (6) on the sample assembly and stitching the resulting orientation images together using standard digital image-processing routines. For reflective samples the pMMP may assume a monostatic geometry, utilizing a beamsplitter similar to a conventional microscope, or a bistatic geometry with an arbitrary bistatic angle.

[0022] FIG. 1A illustrates an embodiment with a small bistatic angle, but the bistatic angle can also be larger, in which case the pMMP resembles a conventional ellipsometer. FIG. 1B illustrates an embodiment used for transmissive samples. If a reflective sample is highly polished, or metallographically polished, such that it reflects like a mirror, then the incident angle on the sample is half the bistatic angle, although the current invention can also be applied to rougher samples that reflect diffusely, in which case the incident angle can be variable in relation to the bistatic angle. Diffuse reflective samples include but are not limited to 3D-printed samples and any samples not subject to conventional polishing or fabrication by deposition to produce a mirror-like finish. Application to diffuse reflective samples also allows the polarimeter of the current invention to map crystal orientation on diffusely-reflecting curved samples and surfaces. For applications to 3D-printed samples the polarimeter and method can be configured to provide material orientation images of each layer or of selected layers produced during the printing process.

[0023] Regardless of its geometry, the pMMP comprises a first set of independent polarization modulators (7) configured to serially modulate the polarization state of the probe beam among a set of independent polarization states. The pMMP further comprises an EMR collector (8) positioned to direct EMR reflected or transmitted by the sample to a second polarization modulator (9) independent of the first set of polarization modulators. The EMR collector is large enough to enable imaging resolution adequate to resolve grains of sizes that commonly occur in industrial parts, for instance down to 10 microns or somewhat smaller. The combined settings of the first set of polarization modulators and the second polarization modulator are temporally-multiplexed and define multiple independent tunable polarization channels. The EMR detector (2), preferably an imaging detector such as a CCD or CMOS focal-plane array (FPA), is positioned to receive the EMR from the second polarization modulator, wherein the detector comprises pixels and produces a set of images that are synchronized with the set of channels formed by the first set of polarization modulators and the second polarization modulator. The pMMP further comprises a processor connected with a memory (10), wherein the processor is configured to execute a classification algorithm stored in the memory that provides an estimate of material orientation. In an imaging pMMP the material orientation is estimated at each image pixel and therefore at each coordinate on the sample.

[0024] In a first embodiment, which was introduced in U.S. Pat. Appln. Pub. No. 2019/0073561 and can achieve high-speed material orientation imaging, the first set of polarization modulators is located on a set of independent arms and the beam is directed serially among the first set of modulators by a first scanning element. The beam is then redirected, by reflection from an assembly of mirrors located on each independent arm, to a second scanning element that redirects the beam from each independent arm to a common path pointed at the sample. In this embodiment each member of the first set of independent polarization modulators, in combination with the second polarization modulator, defines an independent polarization channel. U.S. Pat. Appln. Pub. No. 2019/0073561 depicts one example of this embodiment.

[0025] In a second embodiment, which is illustrated in FIG. 1 and can achieve material orientation imaging using a polarimeter with lower size, weight, and power (SWaP) requirements, which can for instance be mounted on a tripod or other transportable platform, the members of the first set of polarization modulators are located in series on a common arm and are configured to switch serially among two or more independent settings, thereby modulating the polarization state of the beam among a set of independent polarization states. The second polarization modulator is likewise configured to switch serially among two or more independent settings. In this embodiment each setting of the first set of independent modulators, in combination with the setting of the second modulator, defines an independent polarization channel. The modulators can be one of several established devices, for UV, visible, or IR light for instance single polarization crystals or waveplates mounted in manual or preferably motorized rotary stages, which is the embodiment illustrated in FIG. 1, or two or more non-rotating polarization crystals mounted on a wheel or on a sliding linear stage.

[0026] Like a conventional optical microscope, the imaging pMMP of the current invention performs parallel imaging based on established optical designs utilizing commercial lenses and mirrors, although most of the lenses and mirrors of a pMMP must be either polarization-preserving or precalibrated in order to eliminate systematic measurement errors. According to the current invention a suitable conventional microscope can be converted into a pMMP suitable for material orientation imaging by adding a module or modules containing the first set of polarization modulators and the second polarization modulator, software for controlling the modulators, and software to implement classification or orientation mapping. The images collected in different polarization channels can be recorded, for instance by a digital FPA, and the set of polarized irradiances at each image pixel can be mapped to a material orientation. The light source must be highly polarized, but can be either coherent or diffuse; a polarization-preserving diffuser can be added to eliminate coherent artifacts.

[0027] The pMMP measurements are related to the polarization signature or Mueller matrix of the material, and the material dependence of the Mueller matrix can be modeled by several approaches, for instance using a machine-learning classifier trained on Mueller-matrix measurements of samples with known material orientations, or by using an electrodynamic model.

[0028] For material orientation imaging polarized images are collected in a set of 3 or more channels from which the recorded image irradiances are functionals of the polarization signature or Mueller matrix of the material at the prevailing optical frequency, incident angle, reflection or transmission angle, and sample finish or condition, among other parameters that can affect the signature. Channel images may be combined in compositions, for instance sums or differences of images or normalized images, in which the resulting image irradiances represent features of the sample that correspond to physical characteristics as determined by a model. Features include but are not limited to individual Mueller-matrix elements. Specific to crystallographic orientation imaging, orientation features correspond to local crystal orientation uniquely over a particular range of orientation, so that a measured orientation feature can be mapped to a unique crystal orientation. For example, mapping the c-axis of a uniaxial crystal can be defined mathematically by the two equations

F.sub.()=t()

and

F.sub.()=p(),

[0029] where F.sub. is the trend feature, is the Mueller matrix, t is the trend mapping function, and is the trend angle, while F.sub. is the plunge feature, p is the plunge mapping function, and is the plunge angle. The mapping functions t and p must be invertible over the relevant ranges of and respectively. For pMMP applications including material orientation imaging the features F depend on only a subset of the Mueller matrix ; a pMMP therefore measures only a subset of the Mueller matrix. In general, the pMMP of the current invention measures more diverse polarization states than a conventional PLM or PDI sensor but less than a complete Mueller-matrix polarimeter like those described in U.S. Pat. Nos. 4,306,809, 5,247,176, and 5,956,147, or in Hoover et al., Optics Express 24(17), 19881 (2016).

[0030] In image format the measurements and mappings are obtained and applied to each image pixel, usually independently, although multi-pixel mappings are also possible. For crystallographic orientation imaging the orientation features can be deduced by measuring a sufficient number of crystals with known orientations and projecting, using established machine-learning and pattern-recognition algorithms, the resulting measured features into a subspace wherein each orientation maps to a unique volume. This training approach is illustrated in FIG. 4. The orientation features can also be defined by feature dependence on crystal orientation according to a physical model, for instance an electrodynamic model. For c-axis mapping there can be separate orientation features for the crystal plunge and trend angles. There can also be separate orientation features for different orientation ranges, for instance an orientation feature appropriate for plunge angles near 0 (see FIG. 2) and a different orientation feature appropriate for plunge angles near 90, and one or more other orientation features for intermediate ranges. Whether using machine-learning or a physical model, the classifier must be trained on signature measurements of crystals with known orientations, with the orientations obtainable from EBSD or XRD measurements. The signature measurements for training are often made on a polarimeter that measures the complete Mueller matrix, such as those instruments described in U.S. Pat. Nos. 4,306,809, 5,247,176, and 5,956,147, or in Hoover et al., Optics Express 24(17), 19881 (2016). If the grains on which the classifier is to be trained are sufficiently large, then the training measurements can be made with a spot-probe polarimeter, in which the laser probe is focused to a small spot on the training samples, rather than an imaging polarimeter with an expanded beam. Once the classifier is trained for a particular material, alloy, or sample type, then for subsequent samples of the same type the classifier can be run and orientation images obtained with the pMMP of the current invention without the need for signature measurements and a training process, therefore without the need for EBSD or XRD measurements. The classifier estimates of crystal orientation at each image pixel can be depicted in grayscale images, for instance as demonstrated in FIG. 3, or in a single color image wherein the orientation angles, for instance the c-axis plunge and trend angles, are color-coded according to a preferred colormap. Color images are advantageous for representing the entire crystal orientation in a single image and for representing periodic orientation angles with smooth color transitions. The overall orientation angle can alternatively be represented by a glyph or oriented icon assigned to each grain in the image. Crystallographic orientation images, or their analogs for amorphous materials, can also be provided as digital image files that provide orientation angles for each pixel. Such digital image files are useful for material, crystallographic, and texture analysis using a diversity of models.

[0031] In at least one embodiment, and as readily understood by one of ordinary skill in the art, the apparatus according to the invention will include a general or specific purpose computer or distributed system programmed with computer software implementing the steps described above, which computer software may be in any appropriate computer language, including C++, FORTRAN, BASIC, Java, assembly language, microcode, distributed programming languages, etc. The apparatus may also include a plurality of such computers/distributed systems (e.g., connected over the Internet and/or one or more intranets) in a variety of hardware implementations. For example, data processing can be performed by an appropriately programmed microprocessor, computing cloud, Application Specific Integrated Circuit (ASIC), Field Programmable Gate Array (FPGA), or the like, in conjunction with appropriate memory, network, and bus elements.

[0032] Note that in the specification and claims, about or approximately means within twenty percent (20%) of the numerical amount cited. All computer software disclosed herein may be embodied on any computer-readable medium (including combinations of mediums), including without limitation CD-ROMs, DVD-ROMs, hard drives (local or network storage device), USB keys, other removable drives, ROM, and firmware.

[0033] Although the invention has been described in detail with particular reference to these embodiments, other embodiments can achieve the same results. Variations and modifications of the present invention will be obvious to those skilled in the art and it is intended to cover in the appended claims all such modifications and equivalents. The entire disclosures of all references, applications, patents, and publications cited above are hereby incorporated by reference.