Sphere 3D Model Doesn't Show Hydrogen For Specific Molecules

[JRomero0613]

For https://github.com/phetsims/QA/issues/506 the sphere 3D models for dimercaptoboron (BH2S2) and thioxoborinothioic acid (BHS2) do not show the hydrogen atoms. The hydrogen atoms can be seen in the ball-and-stick 3D model, just not on the sphere 3D model. To replicate go to the playground screen and navigate to the 5th panel with hydrogen, boron, sulfur, silicon, and phosphorus (this is the only panel with sulfur). Create either of the two aforementioned molecules using a boron with two sulfur on either side, and either one or two hydrogen attached to the sulfur (pictures of these molecules can be seen below as a reference). Next view the 3D model using the 3D button next to the molecule name. By default the sphere model should show up first. For both of the molecules mentioned the hydrogen is not visible in the sphere model, if the model is changed to the ball-and-stick model the hydrogen atom(s) are visible. I have tried to test other various molecules to see if this is a widespread problem, but so far only the two mentioned molecules exhibit unanticipated behavior. Below I have attached the molecular composition for the two molecules, as well as both 3D model views of the model. This issue was discovered on Windows 10 on Chrome but appears on Firefox and Edge on Windows 10, as well as iOS 13.3.1. This indicates it is not a browser or device specific issue. I will be looking for more molecules that exhibit unanticipated behavior in the 3D model views and will update if any more are found. Thioxoborinothioic acid molecular composition: Thioxoborinothioic acid ball-and-stick 3D model: Thioxoborinothioic acid sphere 3D model: Dimercaptoboron molecular composition: Dimercaptoboron ball-and-stick 3D model: Dimercaptoboron sphere 3D model:

[loganbraywork]

It seems that the larger molecules overlap with the hydrogen. When you attach hydrogen to Boron for example, they Hydrogen appears to be inside of the Boron atoms.

[Denz1994]

More context around this issue can be found at this comment chain after this comment.

I believe the issue is we currently don't have a neat way of handling radicals in our space filled modeling. This is because we never initially, filtered out these molecule types in the original pub chem data pull and the space filled model uses that pub chem data.

The inconsistency of showing the hydrogens mentioned above is due to the sizes of borane vs silicon. Borane is just small enough to still show some of the hydrogens, but silicon is so large it engulfs the hydrogens. I would expect to see no hydrogens with a radical containing Chlorine as well.

We can either go with the teacher's notes solution posed in https://github.com/phetsims/build-a-molecule/issues/150#issuecomment-582000683 and make sure to filter out radical data in #158. Or I'll have to find some special case handling for radicals, which won't be a minor change.

Thoughts @arouinfar or @ariel-phet? Raising priority because this will require ThreeJS implementation changes.

[ariel-phet]

@Denz1994 I am fine with the teacher tips solution and looking to eventually filter out radical data.

[Denz1994]

Got it, because the content of this issue is covered in the tagged issues above. I'll close this one.

This issue should be considered another data point for #158.