Wikipedia talk:WikiProject Elements/Quantum graphics

Quantum graphics
"Quantum graphics" I called our new topic on how to show electron configuration better. As we know, just circles-with-dots is not very accurate or usefull. It originated on my talkpage, I'll copy relevant texts here. Double sharp has stretched the topic into the super heavy elements. -DePiep (talk) 17:18, 26 August 2013 (UTC) This topic is a new development in WT:ELEMENTS and may run for a longer time and with continuing edits (discussions, proposals). Also it may slow down this Talkpage to reload, since it will contain graphics and larger tables. I propose to split this topic off to a page like Elements/Quantum graphics page (name suggestion). This WP page(s) also allows us to organize it more project-like (instead of discussion-like timeline threads we are supposed to use on this talkpage). On this talkpage (WT:ELEM), an introduction and some teasing graphics can stay. Plus of course links. -DePiep (talk) 09:05, 12 September 2013 (UTC)
 * Split off proposal
 * Not a bad idea. I don't really want another Metalloids... ;-) Double sharp (talk) 09:15, 12 September 2013 (UTC)
 * It is this page now: Wikipedia talk:WikiProject Elements/Quantum graphics. -DePiep (talk) 19:57, 11 October 2013 (UTC)

Electron configuration images
reproduction of ; contributions by Double sharp and me. -DePiep (talk) 17:25, 26 August 2013 (UTC)
 * (introductionary talk omitted here)
 * We need a scheme that shows "[He] 2s2 2p2" good. DePiep (talk) 23:17, 24 August 2013 (UTC)
 * I agree. But it's very difficult. I can envision for that some vaguely circular areas labelled 1s, 2s, and 2p, but that becomes an insane mess of circles. I also would not want to have a big shaded area for each shell, because we don't know exactly how the energy levels of the shells are for most of the elements (and bad things happen to this scheme in the 9th period because 9p1/2 and 8p3/2 are about the same level, so insane overlapping happens and I'm not sure how you will do that!


 * For a better graph I must study some quantum mechanics first. (eh, let me coin it "quantum graphics"). Is should be as useful as the Feynman diagram. We could use colors for electrons that are: incomplete (missing) in a shell, the one added compared its left neighbor. The [Xe] detail could be clickable (linked) separately (using some image technique). We could leave out details (like your 9p1/2 ... point) from the element graph, and make that clickable. Have you seen a graph system somewhere that could be used for all elements? -DePiep (talk) 08:27, 26 August 2013 (UTC)
 * How do you hide the core? I mean, even if 7p is full, 7d is not, so I'm not sure how you're going to deal with that.
 * Here's my idea: Get rid of the circles. I mean, there's no way to have circular orbits without causing visual havoc. Instead have a linear view like at the solar system template. Colour-code the shells 1–9. We have 1s first, then 2s further out, then 2p, 3s, 3p, 3d and 4s overlapping, 4p, 4d and 5s overlapping, 4f/5d/6s, 6p, 5f/6d/7s, 7p1/2, 7p3/2 (do note: 7s falls further down for elements 113++, while 6d stays up. 114 has it reversed, but not by much. In general (periods 4 to 8) the ns electrons should fall inwards in the np-block, by inert pair effect; 8s should start fading in the 7d-block.), 8s, a mess of 5g/6f/7d/8p1/2, 9s, 9p1/2/8p3/2, a complete mess containing 6g, 7f, 9d and numerous other blurry ones (because no ref gives anything about what else should be here). Then it should fade out after that. Then we would have a sort of schematic diagram of approximate energy levels (see Ununseptium for an exact version).
 * Actually, this could be a great as a GIF scrolling through elements 1–172 and then 173–184 can just insert electrons (I'm thinking silver spheres to mark them out) into the big mess of colour at the right and fade out gradually. Electron configurations should be given in each frame, along with name and symbol (obviously). Just shove the nucleus almost out of the frame so we have lots of room for the electrons. We should rush through 173–183 without predicted electron configurations. The more I think about this, the more I think it would be a great animation for extended periodic table. It works there better as this is either very wide or very tall (would prefer wide for an article, not tall).
 * If you like it, tell me and I will give you the necessary details. It will probably take a while, but it will be so worth it. Double sharp (talk) 14:29, 26 August 2013 (UTC)
 * Oh yeah, and I forgot to mention: higher subshells that are not filled or being filled yet are greyed out. When relativistic effects start taking over at high Z, visually show the higher subshells (except s) slowly splitting. Double sharp (talk) 14:35, 26 August 2013 (UTC)
 * Yes, getting rid of the circles. It cannot be picture-like, we must step to a scheme. I also expect we cannot use the cloud-shapes. And YESSSSS to that gif! Popping in the electrons! So we need a tough principled base that covers H to 184 the same way. An idea On how to handle details (i.e., links to detail topics): the base drawings can have a visual box that says like "[Xe]". Put in a template, that box has a transparent cover image, that links to the detail page (a Xenon section?). Other details can have a different cover+link. (We use this trick in NavPeriodicTable). Will take a look at the solar system template in a minute. -DePiep (talk) 15:45, 26 August 2013 (UTC)
 * You mean Distance from Sun using EasyTimeline I guess. -DePiep (talk) 15:49, 26 August 2013 (UTC)
 * Do I understand correctly: each element has its own image right? The Uus picture you mention (into the Island of S) has multiple elements. A gif animation only shows them after another; this works if we keep the same frames & positions for the graphic items, or maybe squeeze over a period. Size for now doesn't matter (indeed we can always decide to put is outside of the infobox), to get width. OTOH, oriented vertically solves something because of easier scrolling? I keep in mind. -DePiep (talk) 16:18, 26 August 2013 (UTC)
 * Can we separate the nine shell color blocks, making a stairs with nine colored steps (they overlap in the end in the l-to-r sense). When undecided (SHE), we can mix them vertically (fading/mixing colors at borders maybe, or grey). -DePiep (talk) 16:47, 26 August 2013 (UTC)

For my idea on noble gas structures in the frames for each element (and the gif): we can keep moving the shown area to the right, and then once a noble gas config disappears from view we have an arrow to it at the extreme left with the label [Xe] or something.

Actually, what we really ought to do for subshell splitting is: e.g. for 7p, there are three 7p orbitals. At H, they should be equally separated in one clear region; when we get to the point that they are actually going to be filled, we should have the lowest one tugged down below the others. At some point the label 7p should get torn into 7p1/2 (the lower one) and 7p3/2 (the upper two, that stay together). Because really 7p is 7p1, since the subscript is effectively the azimuthal quantum number.

O DePiep, please give us nine distinct colours that will show up well here. Double sharp (talk) 10:40, 27 August 2013 (UTC)


 * Actually, yeah, showing the individual orbitals gives us the magnetic quantum number as well. Spin quantum number can be notated with a small up- or down-arrow on the electron in question. Now our diagram can show all four quantum numbers. Hooray!
 * (I really like how this scheme applies just as well to all the elements.)
 * For the eV for energy levels, you can stop specifying exact values after around Lr (since AFAIK we don't have them). Double sharp (talk) 11:05, 27 August 2013 (UTC)
 * I think though that when we have predictions (e.g. E113, E117) we should not be afraid to use them. :-) Double sharp (talk) 12:11, 27 August 2013 (UTC)
 * I'd like to have one graphic solution for all predictions. Grey, asterisk or brackets? tbd. -DePiep (talk) 12:39, 27 August 2013 (UTC)
 * A word about details: we should build the concept first, as detailed as we need. That is, positions for 118 (eh, 184) electrons are reserved. Also their (individual) quantum numbers should be there, I guess. also the SHE issues should have a graphic solution. A big and cluttered graph that will be -- but systematically correct. Then, after we agree on that, we can talk about folding the [noble gas] shells, omitting unused shells per element, reduce text & clutter, gif movie effects, maybe move details to outside of the graph, &tc, but only after the plan is complete. -DePiep (talk) 12:39, 27 August 2013 (UTC)

Graph-01
What we could do, first throw. Phosphorus demo. Does one get the electron configuration, by quantum mechanics? Feel free to edit.

Base electron setup is at element cell-ec/base118 (data filled; most recent & most complete. One can copy-paste the table from there). More examples on element cell-ec/sandbox; master template is element cell-ec. -DePiep (talk) 16:47, 27 August 2013 (UTC)

graph-01. Introduces
 * - show nucleus, write atomic number and symbol
 * - n-colors for shells
 * - use All Four qn defining values (n,l,m,s)
 * - line up unoccupied electrons (by positions; grey)
 * - simple fold of noble shells into a single item
 * - introduce eV scale with 0 (but not shown to scale yet)
 * - allow editor notes (we)

Todo next
 * - use eV scale: set each electron on its eV distance (so not the same distance steps).
 * - 0 position eV is where?
 * - Overlapping positions when same eV for two electrons might occur
 * - Check and improve spin values, they are just up/down altering now. Spin_quantum_number says something about this (final paragraph). -DePiep (talk) 08:05, 29 August 2013 (UTC)

-DePiep (talk) 22:30, 27 August 2013 (UTC)

argh, not that wide! we can collapse up and down arrows into one orbital and have it with both up and down arrows if both are full, or just up if one alone is. We can have big brackets showing the shell number for each colour.

Also, please: there should be some cue that 3d and 4s are about the SAME energy level. overlapping maybe. you could have each shell on one horizontal level, making overlaps in eV easier to see. Noble gas notation still works. Double sharp (talk) 10:10, 28 August 2013 (UTC)
 * - re width: don't bother current width. These are just paper cutouts to play with. Templates are much easier to adjust than svg graphs (for me). We are developing what information to show and how to show it graphically first. Once we have a complete & correct graph, chunky, we 1. put all its data in a table and 2. use a script that writes the data+graphs in svg format and churns out 184 images -- much smaller & smarter). The bars could be more like this.
 * So we do not start with collapsing the spin twins or so for space reasons.
 * - re same E level (e.g. 3d and 4s). Absolutely. The intention is to put the electrons to the eV scale. That is: distance consistent with their eV. If two are overlapping, we could shift one up (vertical stacking). We could even move a whole n-shell on row upward (completely above the n-1 row). Would use nine rows then - not impossible (reduce bars to circles maybe). Is there a datalist that says the eV for each electron position? -DePiep (talk) 21:12, 28 August 2013 (UTC)
 * - Any use to mark the &#x2113; levels? With what kind of graphic item? The s-p-d-f-g pattern might show (or not). -DePiep (talk) 07:31, 29 August 2013 (UTC)
 * was thinking more of dotted lines representing each nl subshell where each dot represents an orbital with a spin-up and a spin-down electron. you know how I love conciseness. :-P Double sharp (talk) 16:16, 30 August 2013 (UTC)
 * First we want to define each electron by full 4QN: n-l-m-s (am I too shortwriting?). Folding shell details, like the very common n-&#x2113;number-of-electrons or by writing [NG], we can do later on easily -- afterwards. It is easy to fold data, it is nigh impossible to add data details later.
 * My issues now are: 1. how to show the ...-m-s values graphically in detail and 2. what are the eV values per electron. I expect issues with incomplete spin twins (looks like we cannot state whether a single one is an "up" or "down" cat) -DePiep (talk) 17:07, 31 August 2013 (UTC)

What is the eV of an electron, defined individually by it's 4QN: "n ℓ mℓ ms"? -DePiep (talk) 21:23, 4 September 2013 (UTC)
 * Re Double sharp
 * Let me tell you beforehand what I will judge you sketch by. I'll be fair.
 * 1: How does it show the electron by: n ℓ mℓ ms?
 * 2: How does it show the electron filling sequence? (by eV, by f-block irregularities, by Mandelung)?
 * 3: How does it help PT-illiterates to understand the PT?
 * 99: How are the foldings (aggregations) done?

-DePiep (talk) 22:37, 5 September 2013 (UTC)
 * New abbreviation: IP = ionization energy (literally, "ionization potential", which is synonymous).
 * 1. n is shown in the shell number and colour. ℓ is shown as spdfg letters. mℓ is not shown: in boron, do we even know whether the outermost electron is in 2px, 2py or 2pz? This is why I didn't include it – it could be even resonant, in which case it makes no sense to claim which p orbital it is in. It might make sense for d and above, because 3dz2 (mℓ = 0) doesn't look like the others (similar stuff happens for f and g). But because it cannot really be applied universally I elected not to show it. ms is present in the up or down arrow on each electron (those grey balls).
 * 2. Electron filling sequence – I have quite a few fundamental misgivings with that concept! You see, naturally you don't get electron filling with proton filling ("oh, I add a proton and an electron, and voilà, I have the next-Z-up atom"). The only process you can get naturally that compares is really ionization. In that case, it shows which electrons happen to be higher-energy, and hence which are ionized first. Note that this is for the neutral atom! Hence it will not follow the 2nd and following ionization energies, because – using Sc as my generic example (I'll use B for a generic small atom and Sc for a generic medium or large-sized atom) – the 1st ionization energy is the energy to get from Sc to Sc+, and the 2nd is from Sc+ to Sc2+. But our diagrams only show the energies for Sc. (Perhaps useful spin-off: diagrams for ions like Sc+, Sc2+ ... it would be useful. Sc20+ has a far higher nuclear charge than H, which should mean that the 1s electron travels faster and is closer to the nucleus and so should have a higher IP.)
 * 3. It should show them what is the highest-energy aka valence electrons for an element? (There should be some sort of fade-out for electrons that do not participate chemically, as this doesn't always correspond to the previous noble gas configuration! For boron, it would be 1s only. For gallium, it would be 3d and anything lower (which does NOT include 4s!). For element 113, it's 7s, 5f, the first six electrons of 6d, and anything below that.) Filling order – yeah, the gif would show that too. And it also tells them the actual energies of the electrons in the atom (something which I never actually mentioned to you; they're all negative!).
 * 99. (any particular reason for this, instead of 4?) Generally, we fold together orbitals that aren't filled and have about the same energy. When those orbitals actually start getting filled, we stop folding them and show the exact energies of the electrons in them (which doesn't make sense when they aren't filled yet, which is why we don't show this then). Also, using 5d instead of 5d3/2 and 5d5/2 is equally a form of "folding". It only makes a difference when we actually get there and relativistic effects are important (which they are, in that region). Double sharp (talk) 10:14, 12 September 2013 (UTC)
 * Great. I will read again & again, and I will use. We have time. -DePiep (talk) 20:24, 15 September 2013 (UTC)

Graph-02
I think I might have to make a picture of what I originally imagined. It will be up here soon. :-) Double sharp (talk) 13:01, 3 September 2013 (UTC)
 * I'd like to see that one. Make it a cheap sketch, as I do here ;-) . -DePiep (talk) 20:57, 4 September 2013 (UTC)

I sincerely apologize for the hurriedness and typo-fullness of this (made up in Microsoft Word!!). I'm not an graphics-maker person for a reason...it's not really complete (it should have a ruler-scale like yours), but I think you can get the idea. :-) Double sharp (talk) 13:04, 10 September 2013 (UTC)

A challenge I did not mention in the pic: if you're going to have anywhere near a decent aspect ratio and show all the shells and subshells (even 1s!) then the scale has to start logarithmic before going linear! (You can see those values.) Double sharp (talk) 13:07, 10 September 2013 (UTC)
 * Great in the essence! Will be back later on. -DePiep (talk) 15:18, 10 September 2013 (UTC)
 * Made this one into subsection "Graph-02", and repositionit right after . -DePiep (talk) 08:22, 12 September 2013 (UTC)

Dimensions to be covered
The graph should be able to show these items independently. I assume for now we add text where needed:
 * Removed here: older experimental list. -DePiep (talk) 09:49, 27 August 2013 (UTC)

This table may change over time, following the discussion Notes and usefull links
 * electron configuration
 * Quantum_number used here
 * Aufbau principle, f-block, Madelung rule
 * energy levels shown (n+l rule)
 * Periodic table (electron configurations) shows filling sequence nicely
 * Please, no Aufbau. By the time you get past element 120 it practically ceases to work. And if we are going full-quantum then the aberrant electron configurations of Cr/Ni/Cu/etc. are expected! Double sharp (talk) 13:59, 27 August 2013 (UTC)
 * OK. -DePiep (talk) 08:02, 29 August 2013 (UTC)


 * About four quantum numbers (4QN) notation. Magnetic quantum number not yet clear. My question:
 * "3d5" means "five 3d electrons". (Mn). Used in our infoboxes, and outside e.g. by pdg.
 * "3d?" to mean: "electron(s) in 3d with m&#x2113;=+2". (That could be 1 or 2 electrons by spin IIC). How/where to position this number in text?
 * -DePiep (talk) 08:02, 29 August 2013 (UTC)
 * I've set it to n ℓ mℓ ms. mℓ should be in range -4 ... +4. -DePiep (talk) 09:07, 29 August 2013 (UTC)

Markings
Possible markings to add to the graph are:

n-Colors
Just some n-colors for us to start with. They are too dark when using as background for text. They will be changed near the end of this development into more fine tuned colors. Whenever needed: Example: -DePiep (talk) 11:40, 27 August 2013 (UTC)
 * {{legend|#00d|2=n=1 #00d}}
 * {{legend|#f00|2=n=2 #f00}}
 * {{legend|#0d0|2=n=3 #0d0}}
 * {{legend|#808|2=n=4 #808}}
 * {{legend|#f88|2=n=5 #f88}}
 * {{legend|#0dd|2=n=6 #0dd}}
 * {{legend|#080|2=n=7 #080}}
 * {{legend|#dd0|2=n=8 #dd0}}
 * {{legend|#d0d|2=n=9 <tt>#d0d</tt>}}
 * {{legend|#000|2=white <tt>#000</tt>}}
 * {{legend|#ccc|2=grey (c) <tt>#ccc</tt>}}
 * {{legend|#fff|2=black <tt>#fff</tt>}}
 * {{legend|#d0d|2=Some text}} &rarr; {{legend|#d0d|2=Some text}}
 * For inline text, use &rarr;  prevents a linebreak, see.


 * Wait, I was under the impression that they weren't going to be backgrounds! I thought of them as text colours, and colours of lines that would represent the individual orbitals. Double sharp (talk) 11:56, 27 August 2013 (UTC)
 * Most importantly, the graphic should have as little text as possible. The graph should explain it by itself (I am also hoping we can do without a legend!). So I expected no text needed on the colors. Also, using colored font, as you write, I have bad fresh memories (one cannot distinguish three or four meaningfully, really, especially when used atop colored background - how wellchosen they may be.) How/wherefor did you plan use colored text? -DePiep (talk) 12:19, 27 August 2013 (UTC)
 * I suggest that for now, to get the conceptual layout, write the text in black below any graphical item. Almost every text item in here has a graphical equivalent (or must have)! -DePiep (talk) 12:24, 27 August 2013 (UTC)


 * Explained: I have chosen these nine colors to prevent visual confusion (colors too similar), and the order is such that neighboring colors are very opposite (greens are not neighbors). shows the intended effect (e.g., neighboring 3d and 4s colors are contrasting more extreme). I have chosen them dark, to make contrast with a light background (white canvas). OTOH, they are not suitable to use as bg color for text (too dark). Text atop of these colors is not intended or requested here anyway. Using three-number codes instead of six is for easy typing in develop time (#080 corresponds to #008800. Writing #ab06f3; is different right?). -DePiep (talk) 09:37, 12 September 2013 (UTC)