nanostructure explorer
The javascript nanostructure-explorer linked here makes available a number of (currently seven) "unknown" three-dimensional nanostructure models for real-time characterization by a high resolution (electron phase contrast) transmission electron microscope (TEM) with strong-phase-object (single-scattering) optics. A wide range of physical and diffraction contrast mechanisms are thereby made available, without the processing-time needed for multi-slice calculations.
Each time the page is loaded, both lens-setting and specimen-orientation are randomized, although a single focus-optimization will apply to all of the available specimens. Moreover the photos that you take as instrument operator are likely to be unique in detail, since each reload randomizes orientations modifiable only in one-degree increments. This is precise enough to characterize structures, but not to replicate images. Cool specimens to play with might include:
a silicon octahedron with {111} facets,
a single hemoglobin molecule,
a Si sphere with hemispheres twinned on (111),
some single-walled carbon nanotubes to which Pt particles are attached with help from single-strand DNA,
a group of randomly oriented nanocrystals,
an insulating nanoparticle with the damage-trail of a fission-fragment's ion explosion-spike passing through it,
a face-centered-cubic (fcc) icosahedral twin, an fcc sphere with a screw-dislocation through its center,
a Lennard-Jones "model-liquid" cluster after simulated annealing,
an interstellar graphite onion with an unlayered graphene core,
a strained Si/Ge interface with an edge dislocation array,a specimen wedge showing deBroglie phase-inversion bands due to thickness,
a foil with bowl-shaped strain able to show bend-contour spiders,
graphite spheres with & without random-layer-lattice disorder,
a foil with coherent precipitates, and dislocation-loop bounded stacking faults,
a crystalline foil with unstrained voids and/or strained amorphous precipitates,
the Au(111)/ZnO(001) interface underlying coherence-zone enhanced rod growth,
and what else?
General observation challenges might involve:
First I would practice a bit without "counting adjustment-steps" in each case, just to see what is involved. It might then be interesting to report on average how many steps it takes for you to arrive at a value satisfactory to you in each case. In the "without help" case you might also report the average final-defocus, and the average final amount of astigmatism that you end up with.
The good news is that once you are happy with the focus on the "with help" simulator linked at the top of the page here, all unknowns (e.g. 0 through 6) may be ready for more careful examination.
Magnification: Measure the width of the image panels on your device screen or printout to determine the magnification of the direct-space (left-side) images (of field-width ~112.9Å) as precisely as possible, to use with image and diffraction-pattern calculators plus the fact that X million (or Y thousand) times magnification means that a nanometer (or micron) on the object corresponds to X (or Y) millimeters in the image.
Observation challenges for "unknown specimen N" might then include: Specimen size and shape: Acquire data on the maximum and minimum projected widths (in Å) of the unknown specimen in a single image and/or in three perpendicular directions, and from this estimate the specimen's aspect-ratio, shape-category (i.e. equant, prolate, oblate or bladed), and average projected-area (in Å2) as well as (if possible) the unknown specimen's surface-area (in cm2) and volume (in cm3),
Crystallinity: Take data on one or more characteristic-periodicities (in Å) internal to the unknown specimen, and from this make a case for the number of single-crystal regions in the specimen and hence on whether the specimen is non-crystalline, poly-crystalline, bi-crystalline, or single-crystalline,
Candidate eliminations: If possible take "zone-axis-pattern" data from a single specimen-orientation on two lattice-spacings (in Å) associated with a single-crystal region of the unknown specimen, and on the angle (in degrees) between those periodicities. Diffraction is very powerful at saying NO. In other words periodicities not present in a particular crystal-structure, which e.g. "light up" our unknown crystal in a darkfield image, can rather strongly eliminate crystal-structure candidates (like those mentioned in the proposed specimen list above) from consideration. This webpage may help check your "two-spacing one-angle" measurements against some common candidates. Which structures can your observations rule out?Candidate indexings: Given "zone-axis pattern" data (e.g. two lattice-spacings and the angle between them) from one single-crystal region of an unknown specimen, find a candidate crystal-structure that might be used to Miller-index the periodicities in the pattern. From this also calculate a possible lattice-direction <uvw> in that candidate-structure for the beam orientation itself,
Oriented basis triplets: If possible take data on three (or more) non-coplanar lattice-spacings (in Å) in one (or all) single-crystal-region(s) of the unknown specimen, and on the angles (in degrees) between the corresponding lattice-spacing g-vectors. From this data, determine lattice-parameters (a, b, c, α, β & γ) for the whole (or part) of the reciprocal-lattice of one (or all) of those single-crystal-regions, as well as the oriented-basis-triplet which defines that region's orientation with respect to the specimen-stage.
and what else?
For instance, can you also find qualitative or quantitative evidence for alpha-helices, beta-pleated sheets, or oxygenated iron-atoms in a hemoglobin molecule, lattice strain amplitudes or directions, crystal shape-transforms in reciprocal space, specimen thickness-variations, bulk (e.g. point, line, planar, inclusion) or interface (e.g. surface reconstruction) defects in crystals, dislocation Burger's vectors, moiré contrast between overlapping lattices, etc.? Alternatively can you use zeros in the contrast transfer function, seen in the image power spectrum, to experimentally determine the spherical aberration coefficient, the damping envelope, or the focus-depth instability of the microscope itself? One might even explore diffraction statistics for randomly-oriented particles by hitting reload again and again, although at this point we've not yet taken pains to make sure that all orientations are equally probable.
Update: As of late July 2015, our explorer allows you to toggle between coherent-brightfield phase-contrast (HREM) and incoherent-darkfield amplitude-contrast (STEM) modes, as shown at left and right (respectively) in the figure above. This is equivalent to reversing the path of electrons through the lens-specimen system, and in the darkfield case rastering the focused beam across the specimen while recording "image pixels" with an annular-darkfield detector. Digital-darkfield analysis of the resulting lattice-images remains possible in both cases.
More specialized versions of this simulator, with interesting biomolecules (sans the hydrogen), with unlayered graphene specimens for our presolar grain work, and with thin film sulfides on amorphous SiO2, may be found: (i) here (for kinesin neck alpha-helices, rhodopsin, and DNA polymerase), (ii) here (for Pt-nanotube-ssDNA, hemoglobin iron labeled with gold, transmembrane proteins, and beta sheets), (iii) here (for some embedded graphene structures), (iv) here (for some sulfide/oxide composites), and (v) here (for a 23,760-atom quasi-crystal approximant, and some de Tomas "annealed" carbons).
Items to measure in biomolecule images not compromised by beam damage might include the diameter and length of alpha-helices, the lateral size of beta sheets, the diameter of DNA and the spacing between base pairs, the helical period or pitch of single-strand DNA, the thickness and lateral size of a lipid bilayer, the distance between isolated heavy atoms, and what else? Can you figure out which specimens correspond to what physical structures in this context?
Experimental image of a Pt nanocrystal in our HRTEM, along with its power spectrum (including a contrast transfer function footprint from the non-crystalline support continuous out to about 1/1.96 [cycles/Å]), for comparison.
The poster for our 2016 presentation at the national Microscopy and Microanalysis meeting in Columbus OH follows:
References:
P. Fraundorf. Nanostructure Explorer. University of Missouri Saint Louis Department of Physics and Astronomy. (2014-2016). Archived by WebCite® at
http://www.webcitation.org/6f3MFIW4x and http://www.webcitation.org/6huzAgg8L.
P. Fraundorf, Stephen Wedekind and Taylor Savage (2016) "Strong-phase-object nanoWorlds Online", working draft notes on use of this simulator, (Archived by WebCite® at http://www.webcitation.org/6fYxowwjR).
P. Fraundorf (2016) "Piecewise-potential nanoWorlds Online" (University of Missouri StL Dept Physics & Astronomy), working draft notes on deBroglie phase contrast. (Archived by WebCite® at http://www.webcitation.org/6f1QH9sac).
P. Fraundorf, Stephen Wedekind, Taylor Savage, David Osborn (2016) "Single-slice nanoWorlds Online" Microscopy and Microanalysis 22:S3, 1442-1443 Cambridge hal-01362470 pdf webpage.
Carla de Tomas, Irene Suarez-Martinez, Nigel A. Marks (2016) "Graphitization of amorphous carbons: A comparative study of interatomic potentials", Carbon 109, 681-693 this link data.
P. Fraundorf (2017) "Real-time digital-darkfield TEM determination of nanocrystal 3D-lattices", Microscopy and Microanalysis 23:S3 (in press) pdf.
Melanie Lipp, Taylor Savage, David Osborn and P. Fraundorf (2017) "Laboratory evidence of slow-cooling for carbon droplets from red-giant atmospheres", Microscopy and Microanalysis 23:S3 (in press) pdf.