A diffraction pattern of a gold nanocrystal formed from using a nano area beam of coherent X-rays. This reciprocal space diffraction image was taken by Ian Robinson's Group to be used in the reconstruction of a real space coherent X-ray diffraction image in 2007.
Coherent diffractive imaging (CDI) is a "lensless" technique for 2D or 3D reconstruction of the image of nanoscale structures such as nanotubes,[1] nanocrystals,[2] porous nanocrystalline layers,[3] defects,[4] potentially proteins,[5] and more.[5] In CDI, a highly coherent beam of X-rays, electrons or other wavelike particle or photon is incident on an object.
The beam scattered by the object produces a diffraction pattern downstream which is then collected by a detector. This recorded pattern is then used to reconstruct an image via an iterative feedback algorithm. Effectively, the objective lens in a typical microscope is replaced with software to convert from the reciprocal space diffraction pattern into a real space image. The advantage in using no lenses is that the final image is aberration–free and so resolution is only diffraction and dose limited (dependent on wavelength, aperture size and exposure). Applying a simple inverse Fourier transform to information with only intensities is insufficient for creating an image from the diffraction pattern due to the missing phase information. This is called the phase problem.
Imaging process
The overall imaging process can be broken down in four simple steps: 1. Coherent beam scatters from sample 2. Modulus of Fourier transform measured 3. Computational algorithms used to retrieve phases 4. Image recovered by Inverse Fourier transform
In CDI, the objective lens used in a traditional microscope is replaced with computational algorithms and software which are able to convert from the reciprocal space into the real space. The diffraction pattern picked up by the detector is in reciprocal space while the final image must be in real space to be of any use to the human eye.
To begin, a highly coherent light source of x-rays, electrons, or other wavelike particles must be incident on an object. This beam, although popularly x-rays, has potential to be made up of electrons due to their decreased overall wavelength; this lower wavelength allows for higher resolution and, thus, a clearer final image. Due to this incident light, a spot is illuminated on the object being detected and reflected off of its surface. The beam is then scattered by the object producing a diffraction pattern representative of the Fourier transform of the object. The complex diffraction pattern is then collected by the detector and the Fourier transform of all the features that exist on the object’s surface are evaluated. With the diffraction information being put into the frequency domain, the image is not detectable by the human eye and, thus, very different from what we’re used to observing using normal microscopy techniques.
A reconstructed image is then made through utilization of an iterative feedback phase-retrieval algorithm where a few hundred of these incident rays are detected and overlapped to provide sufficient redundancy in the reconstruction process. Lastly, a computer algorithm transforms the diffraction information into the real space and produces an image observable by the human eye; this image is what we would likely see by means of traditional microscopy techniques. The hope is that using CDI would produce a higher resolution image due to its aberration-free design and computational algorithms.
There are two relevant parameters for diffracted waves: amplitude and phase. In typical microscopy using lenses there is no phase problem, as phase information is retained when waves are refracted. When a diffraction pattern is collected, the data is described in terms of absolute counts of photons or electrons, a measurement which describes amplitudes but loses phase information. This results in an ill-posed inverse problem as any phase could be assigned to the amplitudes prior to an inverse Fourier transform to real space.[6]
Three ideas developed that enabled the reconstruction of real space images from diffraction patterns.[5] The first idea was the realization by Sayre in 1952 that Bragg diffraction under-samples diffracted intensity relative to Shannon's theorem.[7] If the diffraction pattern is sampled at twice the Nyquist frequency (inverse of sample size) or denser it can yield a unique real space image.[2] The second was an increase in computing power in the 1980s which enabled iterative hybrid input output (HIO) algorithm for phase retrieval to optimize and extract phase information using adequately sampled intensity data with feedback. This method was introduced[4] by Fienup in the 1980s.[8] In 1998, Miao, Sayre and Chapman used numerical simulations to demonstrate that when the independently measured intensity points is more than the unknown variables, the phase can be in principle retrieved from the diffraction pattern via iterative algorithms.[9] Finally, Miao and collaborators reported the first experimental demonstration of CDI in 1999 using a secondary image to provide low resolution information.[10] Reconstruction methods were later developed that could remove the need for a secondary image.
A simulated double wall nanotube (n1,m1)(n2,m2) can be used to test a CDI algorithm. First, a simulated nanotube is created (left) given the chiral numbers, (26,24)(35,25) in this case. Then a diffraction pattern is created by using the power spectrum function in Digital Micrograph software (middle). Finally, the algorithm is tested by reconstructing a final image (right). This work was performed by Ji Li and Jian-Min Zuo in 2007.
Reconstruction
In a typical reconstruction[2] the first step is to generate random phases and combine them with the amplitude information from the reciprocal space pattern. Then a Fourier transform is applied back and forth to move between real space and reciprocal space with the modulus squared of the diffracted wave field set equal to the measured diffraction intensities in each cycle. By applying various constraints in real and reciprocal space the pattern evolves into an image after enough iterations of the HIO process. To ensure reproducibility the process is typically repeated with new sets of random phases with each run having typically hundreds to thousands of cycles.[2][11][12][13] The constraints imposed in real and reciprocal space typically depend on the experimental setup and the sample to be imaged. The real space constraint is to restrict the imaged object to a confined region called the "support". For example, the object to be imaged can be initially assumed to reside in a region no larger than roughly the beam size. In some cases this constraint may be more restrictive, such as in a periodic support region for a uniformly spaced array of quantum dots.[2] Other researchers have investigated imaging extended objects, that is, objects that are larger than the beam size, by applying other constraints.[14][15][16]
In most cases the support constraint imposed is a priori in that it is modified by the researcher based on the evolving image. In theory this is not necessarily required and algorithms have been developed[17] which impose an evolving support based on the image alone using an auto-correlation function. This eliminates the need for a secondary image (support) thus making the reconstruction autonomic.
The diffraction pattern of a perfect crystal is symmetric so the inverse Fourier transform of that pattern is entirely real valued. The introduction of defects in the crystal leads to an asymmetric diffraction pattern with a complex valued inverse Fourier transform. It has been shown[18] that the crystal density can be represented as a complex function where its magnitude is electron density and its phase is the "projection of the local deformations of the crystal lattice onto the reciprocal lattice vector Q of the Bragg peak about which the diffraction is measured".[4] Therefore, it is possible to image the strain fields associated with crystal defects in 3D using CDI and it has been reported[4] in one case. Unfortunately, the imaging of complex-valued functions (which for brevity represents the strained field in crystals) is accompanied by complementary problems namely, the uniqueness of the solutions, stagnation of the algorithm etc. However, recent developments that overcame these problems (particularly for patterned structures) were addressed.[19][20] On the other hand, if the diffraction geometry is insensitive to strain, such as in GISAXS, the electron density will be real valued and positive.[2] This provides another constraint for the HIO process, thus increasing the efficiency of the algorithm and the amount of information that can be extracted from the diffraction pattern.
Algorithms
One of the most important aspects of coherent diffraction imaging is the algorithm that recovers the phase from Fourier magnitudes and reconstructs the image. Several algorithms exist for this purpose, though they each follow a similar format of iterating between the real and reciprocal space of the object (Pham 2020). Furthermore, a support region is frequently defined to separate the object from its surrounding zero-density region (Pham 2020). As mentioned earlier, Fienup developed the initial algorithms of Error Reduction (ER) and Hybrid Input-Output (HIO) which both utilized a support constraint for real space and Fourier magnitudes as a constraint in reciprocal space (Fienup 1978). The ER algorithm sets both the zero-density region and the negative densities inside the support to zero for each iteration (Fienup 1978). The HIO algorithm relaxes the conditions of ER by gradually reducing the negative densities of the support to zero with each iteration (Fienup 1978). While HIO allowed for the reconstruction of an image from a noise-free diffraction pattern, it struggled to recover the phase in actual experiments where the Fourier magnitudes were corrupted by noise. This led to further development of algorithms that could better handle noise in image reconstruction. In 2010, a new algorithm called oversampling smoothness (OSS) was created to use a smoothness constraint on the imaged object. OSS would utilize Gaussian filters to apply a smoothness constraint to the zero-density region which was found to increase robustness to noise and reduce oscillations in reconstruction (Rodriguez 2013).
Generalized Proximal Imaging (GPS)
Building upon the success of OSS, a new algorithm called generalized proximal smoothness (GPS) has been developed. GPS addresses noise in the real and reciprocal space by incorporating principles of Moreau-Yosida regularization, which is a method of turning a convex function into a smooth convex function (Moreau 1965) (Yosida 1964). The magnitude constraint is relaxed into a least-fidelity squares term as a means of lessening the noise in the reciprocal space (Pham 2020). Overall, GPS was found to perform better than OSS and HIO in consistency, convergence speed, and robustness to noise. Using R-factor (relative error) as a measurement for effectiveness, GPS was found to have a lower R-factor in both real and reciprocal spaces (Pham 2020). Moreover, it took fewer iterations for GPS to converge towards a lower R-factor when compared to OSS and HIO in both spaces (Pham 2020).
Two wave sources are coherent when their frequency and waveforms are identical; this property of waves allows for stationary interference in which the wave is temporally or spatially constant and the waves are either added or subtracted from one another. Coherence is important in the context of CDI as the coherence of the two sources allows for the continuous emission of waves to occur. A constant phase difference and the coherence of a wave are necessary in order to obtain any type of interference pattern.
Clearly a highly coherent beam of waves is required for CDI to work since the technique requires interference of diffracted waves. Coherent waves must be generated at the source (synchrotron, field emitter, etc.) and must maintain coherence until diffraction. It has been shown[11] that the coherence width of the incident beam needs to be approximately twice the lateral width of the object to be imaged. However determining the size of the coherent patch to decide whether the object does or does not meet the criterion is subject to debate.[21] As the coherence width is decreased, the size of the Bragg peaks in reciprocal space grows and they begin to overlap leading to decreased image resolution.
Energy sources
X-ray
Coherent x-ray diffraction imaging (CXDI or CXD) uses x-rays (typically .5-4keV)[5] to form a diffraction pattern which may be more attractive for 3D applications than electron diffraction since x-rays typically have better penetration. For imaging surfaces, the penetration of X-rays may be undesirable, in which case a glancing angle geometry may be used such as GISAXS.[2] A typical x-ray CCD is used to record the diffraction pattern. If the sample is rotated about an axis perpendicular to the beam a 3-Dimensional image may be reconstructed.[12]
Due to radiation damage,[5] resolution is limited (for continuous illumination set-ups) to about 10nm for frozen-hydrated biological samples but resolutions of as high as 1 to 2nm should be possible for inorganic materials less sensitive to damage (using modern synchrotron sources). It has been proposed[5] that radiation damage may be avoided by using ultra short pulses of x-rays where the time scale of the destruction mechanism is longer than the pulse duration. This may enable higher energy and therefore higher resolution CXDI of organic materials such as proteins. However, without the loss of information "the linear number of detector pixels fixes the energy spread needed in the beam"[11] which becomes increasingly difficult to control at higher energies.
In a 2006 report,[4] resolution was 40nm using the Advanced Photon Source (APS) but the authors suggest this could be improved with higher power and more coherent X-ray sources such as the X-ray free electron laser.
Simulated single wall carbon nanotube (left) is used to generate a diffraction pattern (middle) for reconstruction (right) algorithm testing. The top and bottom are different chirality tubes. This work was performed by Ji Li and Jian-Min Zuo in 2007.
Electrons
Coherent electron diffraction imaging works the same as CXDI in principle only electrons are the diffracted waves and an imaging plate is used to detect electrons rather than a CCD. In one published report[1] a double walled carbon nanotube (DWCNT) was imaged using nano area electron diffraction (NAED) with atomic resolution. In principle, electron diffraction imaging should yield a higher resolution image because the wavelength of electrons can be much smaller than photons without going to very high energies. Electrons also have much weaker penetration so they are more surface sensitive than X-rays. However, typically electron beams are more damaging than x-rays so this technique may be limited to inorganic materials.
In Zuo's approach,[1] a low resolution electron image is used to locate a nanotube. A field emission electron gun generates a beam with high coherence and high intensity. The beam size is limited to nano area with the condenser aperture in order to ensure scattering from only a section of the nanotube of interest. The diffraction pattern is recorded in the far field using electron imaging plates to a resolution of 0.0025 1/Å. Using a typical HIO reconstruction method an image is produced with Å resolution in which the DWCNT chirality (lattice structure) can be directly observed. Zuo found that it is possible to start with non-random phases based on a low resolution image from a TEM to improve the final image quality.
LEFT Volume representation of a particle formed by a collection of octahedral Si nanoparticles, RIGHT The central slice showing the high degree of porosity.
In 2007, Podorov et al.[22] proposed an exact analytical solution of CDXI problem for particular cases.
In 2016 using the coherent diffraction imaging (CXDI) beamline at ESRF (Grenoble, France), the researchers quantified the porosity of large faceted nanocrystalline layers at the origin of photoluminescence emission band in the infrared.[3] It has been shown that phonons can be confined in sub-micron structures, which could help enhance the output of photonic and photovoltaic (PV) applications.
In situ CDI
Incomplete measurements have been a problem observed across all algorithms in CDI. Since the detector is too sensitive to absorb a particle beam directly, a beamstop or hole must be placed at its center to prevent direct contact (Pham 2020). Furthermore, detectors are often constructed with multiple panels with gaps between them where data again cannot be collected (Pham 2020). Ultimately, these qualities of the detector result in missing data within the diffraction patterns. In situ CDI is a new method of this imaging technology that could increase resistance to incomplete measurements. In situ CDI images a static region and a dynamic region that changes over time as a result of external stimuli (Hung Lo 2018). A series of diffraction patterns are collected over time with interference from the static and dynamic regions (Hung Lo 2018). Because of this interference, the static region acts as a time invariant constraint that phases patterns together in fewer iterations (Hung Lo 2018). Enforcing this static region as a constraint makes in situ CDI more robust to incomplete data and noise interference in the diffraction patterns (Hung Lo 2018). Overall, in situ CDI provides clearer data collection in fewer iterations than other CDI techniques.
Related techniques
Various techniques for CDI have been developed over the years and utilized to study samples in physics, chemistry, materials, science, nanoscience, geology, and biology (6); this includes, but is not limited to, plane-wave DCI, Bragg CDI, ptychography, reflection CDI, Fresnel CDI, and sparsity CDI.
Ptychography is a technique which is closely related to coherent diffraction imaging. Instead of recording just one coherent diffraction pattern, several – and sometimes hundreds or thousands – of diffraction patterns are recorded from the same object. Each pattern is recorded from a different area of the object, although the areas must partially overlap with one another. Ptychography is only applicable to specimens that can survive irradiation in the illuminating beam for these multiple exposures. However, it has the advantage that a large field of view can be imaged. The extra translational diversity in the data also means the reconstruction procedure can be faster and ambiguities in the solution space are reduced.
Diffraction is the interference or bending of waves around the corners of an obstacle or through an aperture into the region of geometrical shadow of the obstacle/aperture. The diffracting object or aperture effectively becomes a secondary source of the propagating wave. Italian scientist Francesco Maria Grimaldi coined the word diffraction and was the first to record accurate observations of the phenomenon in 1660.
Interferometry is a technique which uses the interference of superimposed waves to extract information. Interferometry typically uses electromagnetic waves and is an important investigative technique in the fields of astronomy, fiber optics, engineering metrology, optical metrology, oceanography, seismology, spectroscopy, quantum mechanics, nuclear and particle physics, plasma physics, biomolecular interactions, surface profiling, microfluidics, mechanical stress/strain measurement, velocimetry, optometry, and making holograms.
Electron diffraction is a generic term for phenomena associated with changes in the direction of electron beams due to elastic interactions with atoms. It occurs due to elastic scattering, when there is no change in the energy of the electrons. The negatively charged electrons are scattered due to Coulomb forces when they interact with both the positively charged atomic core and the negatively charged electrons around the atoms. The resulting map of the directions of the electrons far from the sample is called a diffraction pattern, see for instance Figure 1. Beyond patterns showing the directions of electrons, electron diffraction also plays a major role in the contrast of images in electron microscopes.
Electron crystallography is a method to determine the arrangement of atoms in solids using a transmission electron microscope (TEM). It can involve the use of high-resolution transmission electron microscopy images, electron diffraction patterns including convergent-beam electron diffraction or combinations of these. It has been successful in determining some bulk structures, and also surface structures. Two related methods are low-energy electron diffraction which has solved the structure of many surfaces, and reflection high-energy electron diffraction which is used to monitor surfaces often during growth.
Selected area (electron) diffraction is a crystallographic experimental technique typically performed using a transmission electron microscope (TEM). It is a specific case of electron diffraction used primarily in material science and solid state physics as one of the most common experimental techniques. Especially with appropriate analytical software, SAD patterns (SADP) can be used to determine crystal orientation, measure lattice constants or examine its defects.
Electron holography is holography with electron matter waves. It was invented by Dennis Gabor in 1948 when he tried to improve image resolution in electron microscope. The first attempts to perform holography with electron waves were made by Haine and Mulvey in 1952; they recorded holograms of zinc oxide crystals with 60 keV electrons, demonstrating reconstructions with approximately 1 nm resolution. In 1955, G. Möllenstedt and H. Düker invented an electron biprism, thus enabling the recording of electron holograms in off-axis scheme. There are many different possible configurations for electron holography, with more than 20 documented in 1992 by Cowley. Usually, high spatial and temporal coherence of the electron beam are required to perform holographic measurements.
Computer-generated holography (CGH) is a technique that uses computer algorithms to generate holograms. It involves generating holographic interference patterns. A computer-generated hologram can be displayed on a dynamic holographic display, or it can be printed onto a mask or film using lithography. When a hologram is printed onto a mask or film, it is then illuminated by a coherent light source to display the holographic images.
The difference-map algorithm is a search algorithm for general constraint satisfaction problems. It is a meta-algorithm in the sense that it is built from more basic algorithms that perform projections onto constraint sets. From a mathematical perspective, the difference-map algorithm is a dynamical system based on a mapping of Euclidean space. Solutions are encoded as fixed points of the mapping.
Phase retrieval is the process of algorithmically finding solutions to the phase problem. Given a complex signal , of amplitude , and phase :
The European X-Ray Free-Electron Laser Facility is an X-ray research laser facility commissioned during 2017. The first laser pulses were produced in May 2017 and the facility started user operation in September 2017. The international project with twelve participating countries; nine shareholders at the time of commissioning, later joined by three other partners, is located in the German federal states of Hamburg and Schleswig-Holstein. A free-electron laser generates high-intensity electromagnetic radiation by accelerating electrons to relativistic speeds and directing them through special magnetic structures. The European XFEL is constructed such that the electrons produce X-ray light in synchronisation, resulting in high-intensity X-ray pulses with the properties of laser light and at intensities much brighter than those produced by conventional synchrotron light sources.
Ptychography is a computational method of microscopic imaging. It generates images by processing many coherent interference patterns that have been scattered from an object of interest. Its defining characteristic is translational invariance, which means that the interference patterns are generated by one constant function moving laterally by a known amount with respect to another constant function. The interference patterns occur some distance away from these two components, so that the scattered waves spread out and "fold" into one another as shown in the figure.
The contrast transfer function (CTF) mathematically describes how aberrations in a transmission electron microscope (TEM) modify the image of a sample. This contrast transfer function (CTF) sets the resolution of high-resolution transmission electron microscopy (HRTEM), also known as phase contrast TEM.
Hybrid input-output (HIO) algorithm for phase retrieval is a modification of the error reduction algorithm for retrieving the phases in coherent diffraction imaging. Determining the phases of a diffraction pattern is crucial since the diffraction pattern of an object is its Fourier transform and in order to properly invert transform the diffraction pattern the phases must be known. Only the amplitude however, can be measured from the intensity of the diffraction pattern and can thus be known experimentally. This fact together with some kind of support constraint can be used in order to iteratively calculate the phases. The HIO algorithm uses negative feedback in Fourier space in order to progressively force the solution to conform to the Fourier domain constraints (support). Unlike the error reduction algorithm which alternately applies Fourier and object constraints the HIO "skips" the object domain step and replaces it with negative feedback acting upon the previous solution.
Single particle analysis is a group of related computerized image processing techniques used to analyze images from transmission electron microscopy (TEM). These methods were developed to improve and extend the information obtainable from TEM images of particulate samples, typically proteins or other large biological entities such as viruses. Individual images of stained or unstained particles are very noisy, and so hard to interpret. Combining several digitized images of similar particles together gives an image with stronger and more easily interpretable features. An extension of this technique uses single particle methods to build up a three-dimensional reconstruction of the particle. Using cryo-electron microscopy it has become possible to generate reconstructions with sub-nanometer resolution and near-atomic resolution first in the case of highly symmetric viruses, and now in smaller, asymmetric proteins as well. Single particle analysis can also be performed by inductively coupled plasma mass spectrometry (ICP-MS).
Phase-contrast X-ray imaging or phase-sensitive X-ray imaging is a general term for different technical methods that use information concerning changes in the phase of an X-ray beam that passes through an object in order to create its images. Standard X-ray imaging techniques like radiography or computed tomography (CT) rely on a decrease of the X-ray beam's intensity (attenuation) when traversing the sample, which can be measured directly with the assistance of an X-ray detector. However, in phase contrast X-ray imaging, the beam's phase shift caused by the sample is not measured directly, but is transformed into variations in intensity, which then can be recorded by the detector.
Fourier ptychography is a computational imaging technique based on optical microscopy that consists in the synthesis of a wider numerical aperture from a set of full-field images acquired at various coherent illumination angles, resulting in increased resolution compared to a conventional microscope.
Precession electron diffraction (PED) is a specialized method to collect electron diffraction patterns in a transmission electron microscope (TEM). By rotating (precessing) a tilted incident electron beam around the central axis of the microscope, a PED pattern is formed by integration over a collection of diffraction conditions. This produces a quasi-kinematical diffraction pattern that is more suitable as input into direct methods algorithms to determine the crystal structure of the sample.
In crystallography, direct methods is a set of techniques used for structure determination using diffraction data and a priori information. It is a solution to the crystallographic phase problem, where phase information is lost during a diffraction measurement. Direct methods provides a method of estimating the phase information by establishing statistical relationships between the recorded amplitude information and phases of strong reflections.
Convergent beam electron diffraction (CBED) is an electron diffraction technique where a convergent or divergent beam of electrons is used to study materials.
Jianwei (John) Miao is a Professor in the Department of Physics and Astronomy and the California NanoSystems Institute at the University of California, Los Angeles. He performed the first experiment on extending crystallography to allow structural determination of non-crystalline specimens in 1999, which has been known as coherent diffractive imaging (CDI), lensless imaging, or computational microscopy. In 2012, Miao applied the CDI method to pioneer atomic electron tomography (AET), enabling the first determination of 3D atomic structures without assuming crystallinity or averaging.
↑ JR Fienup (1987). "Reconstruction of a complex-valued object from the modulous of its Fourier transform using a support constraint". J. Opt. Soc. Am. A. 4: 118–123. Bibcode:1987JOSAA...4..118Y. doi:10.1364/JOSAA.4.000118.
↑ J Miao; D Sayre; H. N. Chapman (1998). "Phase Retrieval from the Magnitude of the Fourier transform of Non-periodic Objects". J. Opt. Soc. Am. A. 15 (6): 1662–1669. Bibcode:1998JOSAA..15.1662M. doi:10.1364/JOSAA.15.001662.
↑ J Miao; P Charalambous; J Kirz; D Sayre (1999). "Extending the methodology of x-ray crystallography to allow imaging of micromere-sized non-crystalline specimens". Nature. 400 (6742): 342–344. Bibcode:1999Natur.400..342M. doi:10.1038/22498. S2CID4327928.
↑ Leili Baghaei; Ali Rad; Bing Dai; Diling Zhu; Andreas Scherz; Jun Ye; Piero Pianetta; R. Fabian W. Pease (2008). "X-ray diffraction microscopy: Reconstruction with partial magnitude and spatial a priori information". Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures. 26 (6): 2362–2366. Bibcode:2008JVSTB..26.2362B. doi:10.1116/1.3002487.
↑ Baghaei, Leili; Rad, Ali; Dai, Bing; Pianetta, Piero; Miao, Jianwei; Pease, R. Fabian W. (2009). "Iterative phase recovery using wavelet domain constraints". Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures. 27 (6): 3192. Bibcode:2009JVSTB..27.3192B. doi:10.1116/1.3258632. S2CID10278767.
↑ A. A. Minkevich; T. Baumbach; M. Gailhanou; O. Thomas (2008). "Applicability of an iterative inversion algorithm to the diffraction patterns from inhomogeneously strained crystals". Phys. Rev. B. 78 (17): 174110. Bibcode:2008PhRvB..78b4110M. doi:10.1103/PhysRevB.78.174110.
This page is based on this Wikipedia article Text is available under the CC BY-SA 4.0 license; additional terms may apply. Images, videos and audio are available under their respective licenses.
