About two years ago, Robert Moerland's Blender illustration was used on the cover of the scientific 'Optica' magazine. This month, he was featured again!
At Delft University, we placed light emitting molecules at different distances from a thin layer of Indium Tin Oxide (ITO). ITO is popular for being both transparent and mildly conducting and therefore used a lot in displays and touch screens.
When the light emitting molecules are not too close to the ITO, 10 nm or more, they shine at full strength. But, when the molecules get very close to the ITO, less than 10 nm, the conduction electrons in the ITO start to sense the presence of the molecules via their electric field, and a strong current is generated inside the ITO. In that case, instead of emitting light, the molecule will lose its energy as heat. The amount of energy that is converted into heat instead of light is strongly dependent on the distance to the ITO: at 1 nm distance, almost all the energy of the molecule is directly transformed into heat and no light can be seen exiting the ITO. As this happens over the length scale of 10 nm, you can reverse the reasoning and measure the intensity to know the distance!*
The cover picture
The cover picture illustrates what happens between distances of 0 to 10 nm. Ball-and-stick objects represent the molecules, which glow brighter the further they are from the surface. A ribbon with marks implies a ruler for nanometer scale distances, which is implemented by measuring the intensity of the molecules.
The model is very simple (see screenshot), and was rendered with Cycles. It essentially is a cube, a plane, a curve and a model for the molecule (I am not good at modelling). The model for the molecule was converted from the PDB (protein data bank) file format to VRML with pdb2vrml by David N. Blauch. The balls that represent the atoms are a mix of a glossy shader with roughness and an emission shader. I used the Fresnel node to make the balls emit light only at the edges.
The cube that depicts the ITO has a texture used for bump and color, which is an actual measurement of the height of a real ITO slide. At first I only used this as a bump material, but then it wasn’t very visible. Therefore I mixed the height map into the color of the glass material of the ITO as well, darkening it slightly where it is lower.
I wasn’t, and still am not, aware of a good method to create a glow on (essentially) non-emissive objects. To get the glow that I wanted, I duplicated the molecules to a new layer and assigned an emission shader to them with a reasonable intensity. I wanted the original (non-emitting) molecules to be visible in the center of the glow. So I passed render layer of the emitting molecules in parallel through a glare node with a size of 6, and a glare node with a size of 9. Then I subtracted the size 6 from the size 9, leaving a hole in the center where the non-emitting molecules would show. I then added this to the final output.
Finally, I also wanted some light to escape from the ITO slide itself, at the edges. For full control, I again duplicated this slide to a new layer, and discarded the bump and color map. I made the glass surface have some roughness and added a point lamp inside. I only wanted light scattering from the edges, so I changed the top and bottom surfaces of this slide to completely smooth glass, and kept the roughness on the sides.
Finally I added some lens flares and did some burning and dodging in post.
*For the purists, we actually use the excited-state lifetime to probe the distance. The intensity drop is a related phenomenon.
The results were published on 22 January in the online open-access (free to read) journal Optica. The work is on the cover of the February issue. A plain English explanation of the work can be found here.