Wikipedia talk:WikiProject Chemistry/Archive 56

Archive 50

←

Archive 54

Archive 55

Archive 56

Formatting: mini-diagram mid-sentence in footnote

In trying to make Aramid more comprehensible, and source the official definition, I ended up needing to add a footnote quoting the FTC document. This includes the phrase

at least 85% of the amide (    ) linkages,

with an actual diagram between the parentheses. In the end I resorted to describing the diagram, and editing the parentheses back in when someone thought they were a typo and deleted them.

The source and my quote from it are in the footnotes to the sentence In order to meet the FTC definition of an aramid, at least 85% of these linkages must be attached to two aromatic rings. at Aramid#Terminology and chemical structure.

The diagram seems key to the definition, and to the understanding of non-chemists wanting to make sense of what the "amide linkages" are, so if anyone knows a sensible way to represent it in the footnote I'd be grateful. Musiconeologist (talk) 20:00, 16 January 2025 (UTC)

This can be done, but I would warn against cluttering the article with placing a diagram in a footnote. Especially when the article already has a nice rendering of an amide polymer in that section. That floated diagram+caption and the wikilink should be more than sufficient for readers.

That being said, here's how to do what you want: {{NoteTag|Example text [[File:Example.svg]] Example text}}; just wrap your diagram and text in the {{efn}} or {{NoteTag}} template and transclude {{Notelist}} or {{Notefoot}} at the bottom of the article (See Template:Notelist for some documentation on footnotes). Of course, you'll have to format the image file so that it looks proper inline with a text block. ― Synpath 23:01, 16 January 2025 (UTC)

Thanks. I was wondering really whether there might be some markup similar to LaTeX for generating it typographically rather than embedding an image. As a non-chemist wanting to be absolutely sure I was understanding things correctly before clarifying the explanation, I found the existing image didn't go far enough, as there was more than one way to choose two adjacent groups as the "linkage". The diagram in the FTC reference was what made it unambiguous for me. Though that was, of course, before I'd made the description more explanatory. (Also, I don't like the fact that half the footnote describes their image, making it very long.)

Probably what it really needs is simply for one of the linkages in the image to be highlighted. Musiconeologist (talk) 23:35, 16 January 2025 (UTC)

Formatting line formula in ChemBox

@Musiconeologist and Synpath: Congrats on the formatting innovations above. Here is a vaguely related challenge. Many inorganic compounds have super complicated structures. So complicated that pictures of the lattice are almost useless for most readers. Check out this fragment of Sr(SCN)₂:

In view of the near uselessness of these complicated images, we often provide a linear formula in the ChemBox. The linear formula does not inform about the 3-D structure, but at least it communicates the components of the structure. We especially rely on this approach for compounds composed of polyatomic ions, like lead thiocyanate because the linear formula reminds the readers that SCN's come as a package.

You will notice that the line equation in lead thiocyanate is an image (png, svg, etc). Why can't I just type Pb²⁺ (SCN^-)₂ into that slot of the ChemBox? Or maybe I can?--Smokefoot (talk) 00:27, 17 January 2025 (UTC)

That is a good point that this should be easy to do with chembox, but isn't as far as I can tell at the moment. You can shoehorn in {{chem2}} as an image caption so long as an image is included. Trying to omit the image by reducing its width doesn't really work as you might suspect. It might be worth adding this option as a new parameter to the chembox, but if this ImageCaption bypass (diff) is adequate then maybe this is fine? ― Synpath 01:04, 17 January 2025 (UTC)

Turns out you can just throw {{chem2}} directly into the ImageFile parameter and it just works. Took a second to figure out why, but looking through chembox the images are formatted through Module:InfoboxImage and when you pass that code anything that starts with a "<" character it doesn't do any processing and just returns what you gave it. Often, {{chem2}} returns html starting with <span class="chemf nowrap">. It's also why you can't just type Pb²⁺ (SCN^-)₂ and get something approaching half-decent.

This behaviour is hacky, inelegant and basically an oversight - it shouldn't be relied on. Its confusing that it works at all and a parameter with clearer, understandable meaning should be used instead. There's an undocumented parameter called ImageCaptionAll that is more appropriate for sticking text in, but its improperly implemented and can't be used atm. ― Synpath 05:48, 17 January 2025 (UTC)

Hah, the parameter was orphaned by some TfD cleanup: Wikipedia:Templates for discussion/Log/2023 July 25#Template:Chembox image caption all. Might be worth asking an admin for a copy of the deleted page to see if its worth anything. ― Synpath 05:59, 17 January 2025 (UTC)

Nice job on strontium thiocyanate, Synpath. That trick could be applied to many articles!--Smokefoot (talk) 15:00, 17 January 2025 (UTC)

Piancatelli rearrangement

Someone has published a paper on the Piancatelli rearrangement through MDPI with plagerises our article. MDPI papers are sometimes OK and sometimes not, but this is quite blatant, they've lifted various images. Project Osprey (talk) 13:36, 14 January 2025 (UTC)

References

Amazing. A message to the journal editor would be in order. I avoid citing MDPI articles, although sometimes their articles looks useful. One suggestion, add it to Plagiarism from Wikipedia.--Smokefoot (talk) 14:19, 14 January 2025 (UTC)

Remember that the images are Creative Commons, so reuse is perfectly OK assuming appropriate credit has been given. I suggesting posting a question on Wikimedia Commons, Commons:Village pump/Copyright. However, I suspect the image use is OK.

I don't review or publish with MDPI after an editor's secretary tried to pressure me to change my review of a paper as they needed articles. I would not waste time on this, i dont think you will gain enough. Ldm1954 (talk) 14:24, 14 January 2025 (UTC) Removed as date noted below makes this not relevant.Ldm1954 (talk) 14:34, 14 January 2025 (UTC)

Are you sure the plagiarism goes in that direction? That paper in Molecules was published in 2013 when the Wikipedia article was a two-sentence stub. Marbletan (talk) 14:27, 14 January 2025 (UTC)

Sorry, I misread the day as the year - but in that case this is major plagiarism on our part (and from a poor source as well).

I think I've traced it. The content change was a single IP edit, but the pictures all came from User:Megnic25. I'm guessing they're the same person. They're a student from Wiki Ed, UC Santa Barbara 2016 (Wikipedia:Wiki_Ed/UC_Santa_Barbara/CHEM_239_(Spring_2016)). The Catellani reaction came out of the same class, which I've never head of. We normally catch these things sooner. Project Osprey (talk) 15:12, 14 January 2025 (UTC)

@Project Osprey: Could you give some specific examples of copyvios in that article? I see strong similarities between their scheme 5 and our File:Yin mechanism.jpg; anything else? Renerpho (talk) 03:19, 21 January 2025 (UTC)

Structural chemistry: WikiEd strikes again?

For what I would think is a central chemical topic, Structural chemistry is not a great page -- largely written as part of a class. While the parts on characterization look OK (I have not checked sources), I think a better general description and linking is needed. Maybe best to respond at the relevant talk page. Ldm1954 (talk) 14:02, 27 January 2025 (UTC)

@CHEM 300 UBC CJA and Ian (Wiki Ed): a lot of WikiEd stuff is awful, including the material emanating from UBC. These students are in over their heads, and their instructors rarely supervise. So the blind are not even led by the blind, they are not led. Some of us try to monitor their assignments as sort of a to-to list.

By some perverse logic, some Wikipedians seem to encourage the unsupervised homework activity, probably assuming that the product is better than it is.

Bottom line is that you could view Structural chemistry as a blank slate, IMHO.--Smokefoot (talk) 14:42, 27 January 2025 (UTC)

The page as it was before the major additions from the WikiEd class did not have much to it, though I don't know how much of an improvement this is. Reconrabbit 14:51, 27 January 2025 (UTC)

Radon and Radium spectral lines

Are our spectral lines for Radon and Radium wrong? We're using this and this since 2013. And yet, our Radon spectrum looks exactly like the Radium spectrum, but with a few extra lines. I thought this could be because the two are chemically similar somehow (just two apart in the periodic table), but then Xenon and Barium should have similar spectra, too... I thought spectral lines were as unique as a finger print. Any ideas?

All these images were uploaded by the same user, who unfortunately has been inactive since 2018. According to the file descriptions, all of our spectra images we use in the info-boxes for the elements don't come from some text book or something like that, but were created by them using a free plug-in for Matlab.[1]

Some general questions: How do we know the spectrum of, say Einsteinium (an element so rare we can hardly have observed it directly). And why do we trust someone whose entire credentials are that they are "interested in antennas",[2] and have created a small program that's been on version 1.0.0.0 since its inception 15 years ago, and that is entirely undocumented, for images that appear in the info-box of every single element on dozens of Wikipedia projects? Shouldn't we refer to a scholarly source for this kind of stuff? Renerpho (talk) 01:14, 14 January 2025 (UTC)

Thanks to Nrco0e for making me aware of this issue. Renerpho (talk) 01:17, 14 January 2025 (UTC)

Credit also goes to Randall Munroe's xkcd: Radon (13 January 2025) for sparking this find. Renerpho (talk) 16:54, 14 January 2025 (UTC)

Probably one should use NIST as a source for spectral lines instead. Double sharp (talk) 01:46, 14 January 2025 (UTC)

@Double sharp: You came close to noticing the discrepancy in September 2022 (Special:Diff/1109788615/1109853291); quote: "I haven't checked it against the pictures, but NIST has data." Renerpho (talk) 02:20, 14 January 2025 (UTC)

@Renerpho: Heh, I didn't remember that. Probably I should've checked back then indeed. I guess I was probably pressed for time then; well, better late than never...I think we should at least remove the wrong pictures first. Double sharp (talk) 02:26, 14 January 2025 (UTC)

@Double sharp: These are added automatically, using a template: Template:Infobox element/symbol-to-spectral-lines-image. I don't know how to remove them. Renerpho (talk) 02:52, 14 January 2025 (UTC)

Never mind, I've removed them. Renerpho (talk) 02:56, 14 January 2025 (UTC)

I might have used too strong language when I said they were "wrong"; e.g. File:Sodium spectrum visible.png has a pair of yellow spectral lines, they're just not as obvious as they should be. My preference would be to keep the current versions until we can do a direct 1:1 swap with new versions; the NIST page that @Double sharp found has a tool to generate these, so we should be able to do this quickly, provided they're public-domain. Preimage (talk) 02:57, 14 January 2025 (UTC)

@Preimage: It takes a single edit to collectively add them back (by reverting me), preferably once they've been re-done. It's better to have no images than ones that are possibly wrong! Or, if you want to check them manually, add them back one by one, that works as well. I have to sign off for now, but I'll check back later. Renerpho (talk) 03:05, 14 January 2025 (UTC)

provided they're public-domain -- They're probably not copyrighted, see [3]. Whether that license amounts to PD, I'm not entirely sure. Renerpho (talk) 03:13, 14 January 2025 (UTC)

The NIST ASD is Standard Reference Data, but made to be freely available. There's also this passage from [4]: NIST web pages are provided as a public service by the National Institute of Standards and Technology (NIST). With the exception of material marked as copyrighted, information presented on NIST sites are considered public information and may be distributed or copied. Use of appropriate byline/photo/image credits is requested. The ASD isn't marked up as copyrighted so the spectra should be fair game with attribution. ― Synpath 03:41, 14 January 2025 (UTC)

Rydberg–Ritz combination principle suggests using NIST Atomic Spectra Database - Lines Holdings. Their spectral lines for neutral Rn and Ra look completely different. As another check, the extremely strong pair of Na spectral lines at 589 and 590 nm is not at all obvious on File:Sodium_spectrum_visible.png. So yeah, we have up to 97 spectral-line errors in elements articles. Probably worth notifying WT:WikiProject_Elements if you haven't done so already. Preimage (talk) 01:50, 14 January 2025 (UTC)

Oh dear... Renerpho (talk) 01:58, 14 January 2025 (UTC)

Besides, I assumed that the author of the spectra were trying to produce absorption lines. If so, the lines should be black instead of white colour. I agree that all the spectra should be remade. -- Flycici (talk) 02:04, 14 January 2025 (UTC)

For reference, this issue was first raised in 2022, on Talk:Radon/Archive 2#Emission spectra?. Renerpho (talk) 02:05, 14 January 2025 (UTC)

@Preimage: WikiProject Elements has already been notified (thanks LaundryPizza03). Renerpho (talk) 02:10, 14 January 2025 (UTC)

Links above were not working for me. I assume they do not for others as well. Here are links to the NIST ASD for convenience:

Atom	Lines	Image
Na I	link	link
Rn I	link	link
Ra I	link	link

I spent a moment trying to find licensing info on the images, but turned up nothing. I'd hope they're public domain. Cheers ― Synpath 03:04, 14 January 2025 (UTC)

I can confirm the links [5]and [6], provided by Preimage, don't work for me either. Renerpho (talk) 03:06, 14 January 2025 (UTC)

Yep, I screwed those up - looks like copying and pasting from the browser bar didn't work for the version of the page I was on. Thanks @Synpath and @Renerpho. Preimage (talk) 06:15, 14 January 2025 (UTC)

I strongly support using traceable data. The images in question come from a generator program that notes "Included with this function is spectrum data for nearly all the elements from Hydrogen to Uranium. See the readme.txt file for further information.". But downloading the matlab package isn't working for me. Someone please check if it identifies the source of its data. DMacks (talk) 03:44, 14 January 2025 (UTC)

@DMacks: I cannot download it either, but I can make and access an online copy. The spectra are stored as txt-files elementname.txt in a sub-directory called "Data", but without a reference that identifies the source. First entry is wavelength (in Angstrom), second entry is intensity, and the data is sorted by wavelength. README.txt says that because the data has been collected from a variety of sources, data for different elements may have been normalised to different values. This sounds to me like NIST may have been one of their sources, but not the only one, and that they had to manually edit the data. It also says that this is the Matlab version of a Java program dicharge.java by John Talbot 1997, hosted by several sites on the internet. That would be this, but dicharge.java comes without data (no spectral lines).

The first few lines of radium.txt are:

4194.090 8 | 4244.720 8 | 4340.640 100 | 4436.270 20 | 4533.110 30 | 4641.290 8 | 4682.280 100 | 4699.280 8 | 4825.910 100 | 4856.070 10 | 4859.410 10 | 4927.530 10 | 5097.560 10 | 5205.930 10 | 5283.280 10 | 5320.290 10 | 5399.800 10 | 5400.230 20 | 5406.810 20 | 5482.130 8 | 5501.980 10 | 5553.570 10 | 5555.850 20 | 5616.660 10 | 5660.810 50 | 5813.630 20 | 6200.300 30 | 6336.900 10 | 6446.200 20 | 6487.320 20 | 6593.340 10 | 6719.320 10 | 6980.220 20

This is the spectrum given by NIST for Ra II. I am unsure if that's appropriate; I think it should be Ra I?

radon.txt has

4226.060 10 | 4307.760 80 | 4335.780 7 | 4349.600 100 | 4435.050 40 | 4459.250 50 | 4508.480 50 | 4577.720 50 | 4609.380 50 | 4721.760 30 | 5722.580 6 | 6061.920 10 | 6200.750 6 | 6380.450 6 | 6557.490 10 | 6606.430 10 | 6627.230 15 | 6669.600 6 | 6704.280 8 | 6751.810 20 | 6806.790 6 | 6836.950 8 | 6837.570 8 | 6891.160 10 | 6998.900 10 | 4194.090 8 | 4244.720 8 | 4340.640 100 | 4436.270 20 | 4533.110 30 | 4641.290 8 | 4682.280 100 | 4699.280 8 | 4825.910 100 | 4856.070 10 | 4859.410 10 | 4927.530 10 | 5097.560 10 | 5205.930 10 | 5283.280 10 | 5320.290 10 | 5399.800 10 | 5400.230 20 | 5406.810 20 | 5482.130 8 | 5501.980 10 | 5553.570 10 | 5555.850 20 | 5616.660 10 | 5660.810 50 | 5813.630 20 | 6200.300 30 | 6336.900 10 | 6446.200 20 | 6487.320 20 | 6593.340 10 | 6719.320 10 | 6980.220 20

The first half of this agrees perfectly with Rn I, indicating that the two are based on the same source in this case. NIST's reference is a 1933 paper, so that makes sense. The second half is a perfect copy of Ra II, explaining why the two spectra look so similar. This is clearly not intentional, as the data is supposed to be sorted by wavelength. I checked if there are more files that are not sorted correctly, and found about two dozen other elements that appear to have the same issue. This doesn't mean that they're all wrong, or that the remaining elements are all correct, just that it doesn't look like an isolated problem. Renerpho (talk) 15:15, 14 January 2025 (UTC)

TL;DR: The reason why our spectra for radon and radium looked so similar was that our spectrum of radium was that of Ra II, while our spectrum of radon was a combination of Rn I and Ra II. About two dozen other elements may also be combinations of multiple elements. The spectra for the remaining elements may not be those of the neutral atom.

I am shocked that these spectra have been online for nearly 15 years, and have been prominently featured on dozens of Wikipedia projects for well over a decade, without anyone noticing they're complete rubbish. Renerpho (talk) 15:36, 14 January 2025 (UTC)

I find that the scale at the bottom of the NIST-generated images is hard to read at 220px size. Maybe make a custom version that is hue-accurate (pure spectral colors cannot be represented exactly in sRGB) and with a labeled λ-axis? –LaundryPizza03 (dc̄) 04:41, 14 January 2025 (UTC)

Maybe we can recycle the spectra that were uploaded in 2013 for that purpose? Create a version without lines. Renerpho (talk) 16:10, 14 January 2025 (UTC)

Here: File:Spectrum_no_lines.png Renerpho (talk) 19:28, 14 January 2025 (UTC)

If these will be recreated wholesale we should likely opt for a scalable SVG. Something like File:Linear visible spectrum.svg could be a good option to build off of. Unfortunately, the link in the description for sourcing is broken, but the linked R code has additional sourcing. It looks like the file has been generated thoughtfully. ― Synpath 20:54, 14 January 2025 (UTC)

Sounds good. I suggest to just cut it to a range of 400 to 700 nm. Renerpho (talk) 21:12, 14 January 2025 (UTC)

If we're going for SVG files then it should be possible to create those from the raw data (the spectral lines) directly. I don't have experience with creating SVG files, but if that's possible then we should take that approach. Renerpho (talk) 21:13, 14 January 2025 (UTC)

It's definitely possible to do a direct conversion from data to SVG, since SVG is composed of primitives like lines of specified RGB colors with specified XY-coordinate end-points. The only visual questions are what line-thickness, how to map the intensities, and a default size so that axis-label text is legible. DMacks (talk) 22:43, 14 January 2025 (UTC)

I wanted to provide another useful source from NIST: Handbook of Basic Atomic Spectroscopic Data

The tables given in this NIST source are of the same format as previously used to generate the images using the MATLAB program.

Here is some detail on how I located this source from the program mentioned above. The Talbot 1997 discharge.java code may not provide raw data, but some is given on the Talbot code's parent website here. This references "Spectra of neutral and singly ionized elements from the Astronomical Data Center (ADC) catalog A6016, by Reader J., Corliss Ch.H. :1981, 'Line Spectra of the Elements', CRC Handbook of Chemistry and Physics; NSRDS-NBS 68." While that reference may not be available, it is listed as a source at NIST's Basic Atomic Spectroscopic Data Handbook, with the full version of that handbook given above. Dmh430 (talk) 01:52, 15 January 2025 (UTC)

This NIST source also discusses the nomenclature e.g. Ra I and Ra II; I refers to the neutral element and II to its singly-ionized form. Dmh430 (talk) 01:54, 15 January 2025 (UTC)

Dear all, I took the liberty to create some samples of Ra I emission and absorption spectra. The data are copied from the NIST webpage. (Again, are they in the public domain? I would think so because they are natural facts.) Also, the hue here is different from the ones provided by @Renerpho and @LaundryPizza03. Hope I have not moved to fast and violated any intellectual right. -- Flycici (talk) 19:16, 15 January 2025 (UTC)

Also the strength of the lines should be proportional to their transition probabilities. But perhaps let's not over-complicate things... Cheers. -- Flycici (talk) 19:18, 15 January 2025 (UTC)

That's Rn I. But yeah, comparing to the spectrum generated by the NIST tool it is clear that the relative-strength data is important for realistic emission spectra, and that we should include the full range from 380–780 nm. For Rn I, there are three particularly strong lines in the excluded far-red zone and one near 434.96 nm. –LaundryPizza03 (dc̄) 04:55, 16 January 2025 (UTC)

Also, I still think that a marked scale would be useful for readers, including color-blind ones and those unfamiliar with the precise colors along the visible spectrum. –LaundryPizza03 (dc̄) 06:54, 16 January 2025 (UTC)

Agreed. Even when you put the obvious accessibility benefits aside, having a numerical scale just seems useful. Renerpho (talk) 04:15, 18 January 2025 (UTC)

I’m not sure accessibility is an issue here. Color blind persons may have trouble distinguishing different frequencies of color but they would have no trouble understanding (1) the number of lines nor (2) the relative widths of gaps nor (3) the brightness or width or whatever we use to indicate intensity. I think this is all one needs to understand an emission spectrum. YBG (talk) 16:35, 18 January 2025 (UTC)

Relative-strength data could be shown with thickness of these lines (for example File:Emission spectrum-H.svg), or with brightness of these lines. Nucleus hydro elemon (talk) 13:32, 16 January 2025 (UTC)

I'd prefer brightness over thickness. It's more realistic, and thick lines could potentially hide other lines that are nearby. Renerpho (talk) 04:13, 18 January 2025 (UTC)

You will likely have to use both width and brightness as the contrast range between low intensity lines and high intensity would be too hard to see if you use a narrow line for the brightest. Graeme Bartlett (talk) 05:05, 18 January 2025 (UTC)

I have another question for these spectral lines. If they mean to be emission spectra, aren't the backgrounds supposed to be black? --Nucleus hydro elemon (talk) 15:39, 18 January 2025 (UTC)

Yes, there should be no background colours in a real emission spectrum, but I believe these faded colours are included as an implicit scale for wavelength rather than a number line. ― Synpath 16:12, 18 January 2025 (UTC)

@Synpath: Brilliant idea (pun intended). But it might be nice to have a small strip that shows a fully illuminated spectrum. YBG (talk) 16:23, 18 January 2025 (UTC)

A strip of colour spectrum above a numbered wavelength axis seems like the best option to me as well, barring any infobox real-estate concerns. ― Synpath 16:50, 18 January 2025 (UTC)

I try to create a sample with emission spectrum, fully illuminated spectrum, and a numbered wavelength axis (File:Atomic spectrum of hydrogen.svg). However, I don't think the colours of these lines are correct. Is the 410nm line dark blue instead of violet? Nucleus hydro elemon (talk) 07:56, 19 January 2025 (UTC)

Agree the colors need a tweak. According to Violet (color)#Optics, we should be representing 400nm radiation using Shades of violet#Electric violet. We then have a choice between using Image:Linear visible spectrum.svg directly (to represent a fully illuminated spectrum of uniform intensity), or adjusting it to have perceptually uniform brightness (which would be my preference; I originally suggested Image:Spectrum4websiteEval.png for this purpose, but this doesn't have spectral markings, and and may not even be spectrally linear). Curiously, the NIST images seem to be using an overly reddish shade of violet for very short wavelengths, almost as if their color range wraps back around from 380nm to 780nm. Lastly, are you log-transforming intensity as the NIST images do? (Makes faint lines easier to see, though requires selecting an arbitrary unit comparison value.) Preimage (talk) 09:12, 19 January 2025 (UTC)

As I have no idea how to do with intensities, I ignored it when creating this spectrum. I also need an answer for the range of visible spectrum: should I use 400~700nm, 380~750nm, or 380~780nm? Nucleus hydro elemon (talk) 09:27, 19 January 2025 (UTC)

NIST uses 380–780 nm as their definition of the visible spectrum. –LaundryPizza03 (dc̄) 12:01, 19 January 2025 (UTC)

I changed 400nm into violet, and 380~400nm are dark violets.

For the intensity of lines, I define NIST's relative intensity of 5 as α = 0, while the brightest line as α = 1. Alpha of other lines are given by $\alpha =log_{\frac {x}{5}}({\frac {y}{5}})$, where $x$ is the intensity of the brightest line, while $y$ is the intensity of the other line. Nucleus hydro elemon (talk) 15:02, 19 January 2025 (UTC)

@Nucleus hydro elemon: "Wavelengths of the 4 lines shown are from NIST" (from the file description) -- if that's from before you extended the range to 380-780 nm then that's fine; otherwise I don't understand it.

I like the plot design! Could you create an example of a more complicated spectrum? Radon would be fitting. We can then decide what's better, the PNG images generated by NIST, or the SVG files. Renerpho (talk) 14:51, 20 January 2025 (UTC)

Here is the spectrum of radon. Some lines are extremely dark to be seen. Nucleus hydro elemon (talk) 06:51, 21 January 2025 (UTC)

None of the lines are visible in the thumbnail. You will have to fatten them up. Graeme Bartlett (talk) 11:04, 21 January 2025 (UTC)

It will be around this size if appeared on infobox

I changed the lines to become thicker, now at least the red 706nm line and several blue lines are visible. Nucleus hydro elemon (talk) 12:54, 21 January 2025 (UTC)

A little nitpick with this line width change is that now there are two overlapping violet lines at 384 nm. Nrco0e (talk • contribs) 18:00, 21 January 2025 (UTC)

It is hard to solve as this issue is related to the thickness of the lines. Too thin, then none of the lines is visible. Too thick, then the overlapping issue arises. I think the issue of overlapping will be inevitable when there are many lines in the spectrum. Nucleus hydro elemon (talk) 23:01, 21 January 2025 (UTC)

Instead of a rectangular slit, we could use a point spread function that slopes off, such as a triangle or a normal distribution. Then the information of multiple lines will not be lost when they overlap. Graeme Bartlett (talk) 11:56, 24 January 2025 (UTC)

I try to create the spectrum of sodium with triangles, but the svg turns out to be glitched. All the lines are at maximum intensity in Wikipedia, but if you download the original file, the original intensities returned back. Nucleus hydro elemon (talk) 08:27, 26 January 2025 (UTC)

The fill-opacity settings are not terminated with a semi-colon. I assume that some ways of visualizing SVG are more forgiving with syntax errors like this. Also, this svg has each path element separated with a line of ccc" /> which is being ignored by the rendering process. ⇌ Synpath 15:03, 26 January 2025 (UTC)

I really like where this is going, and one of the things I noticed while going through the NIST ASD simulated em spectra was that they have radial gradients applied to them that roughly scale with intensity. This is similar to at least some images of em spectra (see the helium one I linked below) and provides an extra dimension aside from colour to distinguish band intensity. It can also allow neighbouring bands to be distinguished more easily as each can have different contours. The key problem I have with this is deciding what the base colour is, then deciding how to scale colour "brightness" and band-width according to intensity as I don't know the relevant physics (though I can guess as to what applies). ⇌ Synpath 22:51, 21 January 2025 (UTC)

Are we redesigning the wheel?

Before we get further into the the details on how to represent emission spectra in a figure we should agree that it is necessary to do that. The PNG images generated on the NIST website are free to use with attribution and simulate the look of an emission spectrum accurately (or at least accurately enough for Wikipedia). For an editor like me it will be quicker to download all the neutral element spectrums, reupload them to commons with a proper description and then link them through the template. A good amount of clicking, but probably quicker and more accurate than programming an emission spectrum generator and trying to decide how to scale line heights, perceived luminosity and sampling colour from a spectrum. ⇌ Synpath 16:30, 19 January 2025 (UTC)

That's not unreasonable. Let's do it. –LaundryPizza03 (dc̄) 17:45, 19 January 2025 (UTC)

I've decided to use this as a chance to learn some Commons file uploading tech (Commons:Guide to batch uploading) and am semi-automating this through a python script. If nothing else I'll learn how to automatically add templated file descriptions to images, which has always been the worst part about uploading images of proteins for me. I hope to be done by the end of the day, but as always: an xkcd. ⇌ Synpath 18:46, 19 January 2025 (UTC)

Let me know if you need help on commons with cats, tagging, or other organization (pretty much anything *once it's on* commons, not *getting it onto* commons). DMacks (talk) 19:59, 19 January 2025 (UTC)

Before I start testing my script to see if it works I figured I should gather opinions on the file description and other metadata. I'd rather get it right now rather than have to correct however many files this ends up being. This seem good?

{{Information
  |description    = Simulated emission spectrum of neutral <element_name> (<element> I). Data retrieved from National Institute of Standards and Technology, 
Atomic Spectra Database (NIST ASD).
  |date           = {{ISOdate|2025-01-19}}
  |source         = Kramida, A., Ralchenko, Yu., Reader, J., and NIST ASD Team (2024). NIST Atomic Spectra Database (ver. 5.12), Online. Available: [https://physics.nist.gov/asd] (2025, January 19). National Institute of Standards and Technology, Gaithersburg, MD. DOI: [https://doi.org/10.18434/T4W30F]
  |author         = 
  |permission     = [https://www.nist.gov/open/copyright-fair-use-and-licensing-statements-srd-data-software-and-technical-series-publications#data]
  |other versions = 
}}

⇌ Synpath 21:46, 19 January 2025 (UTC)

Feedback, in order of the fields... You can leave the |date= as just YYYY-MM-DD (no wrapper conversion template) because {{Information}} on Commons handles formatting and automatically. A better formatting of a DOI link is [[doi:10.18434/T4W30F]], both on enwiki and Commons, rather than an explicit and reader-visible URL. A |author= value is required. You could offload some of the |source= details to it (distinguishing where you got it from how it was actually created?). Good link to the source's permissions statement, but it's not clear what specific part(s) of it applies. For example, if you are asserting that it the image itself is work of an agency of the US government (obviously simple facts-of-nature data are not license-protectable), include {{PD-USGov}}. DMacks (talk) 23:12, 19 January 2025 (UTC)

== {{int:filedesc}} ==
{{Information
  |description = Simulated emission spectrum of neutral <element_name}> (<element> I). Image retrieved from National Institute of Standards and Technology, Atomic Spectra Database (NIST ASD).
  |date = 2025-01-19
  |source = Image generated through: [https://physics.nist.gov/PhysRefData/ASD/lines_form.html]
  |author = Kramida, A., Ralchenko, Yu., Reader, J., and NIST ASD Team (2024). NIST Atomic Spectra Database (ver. 5.12), Online. Available: [https://physics.nist.gov/asd] (2025, January 19). National Institute of Standards and Technology, Gaithersburg, MD. DOI: [[doi:10.18434/T4W30F]]
  |permission = 
  |other versions =
}}
=={{int:license-header}}==
{{PD-USGov-NIST}}

Thanks for the advice! I've incorporated it in the description above. I'll be uploading a test set as soon as I get OpenRefine reading in structured data, but let me know if there's something to change. ⇌ Synpath 01:06, 20 January 2025 (UTC)

Looks nice! You can omit the explicit "DOI: " prefix to the doi:.... link because the link adds that automatically. The "Image retrieved from" statement should go in probably |source= since source means "where it came from" whereas description means "what it illustrates". Butthe fact that it's NIST data might remain, since the fact that it's traceable data vs simulation vs whatever could be seen as part of its identity. DMacks (talk) 01:27, 20 January 2025 (UTC)

Made the changes. The script works and I ran a test upload with hydrogen to beryllium through OpenRefine and it looked like everything uploaded fine. Yet the files aren't on Commons. Oh well, I must be missing a step or need other permissions and wasn't served an error. I'll try again another day. ⇌ Synpath 03:42, 20 January 2025 (UTC)

Can your tool also set categories (probably one for each specific element and one for this set of images). DMacks (talk) 17:05, 20 January 2025 (UTC)

Got the test upload to work, something silly on my end of course. Also, I forgot to include that I was adding the category c:Category:Simulated atomic emission spectra in my upload. The category list can be easily extended to include one for the element name. Adding one specifically for these NIST ASD simulated spectra is an option I didn't consider. Is that overcategorizing with what I'm adding already? File names are going to be formatted 4 (Be I) NIST ASD emission spectrum.png ⇌ Synpath 22:53, 20 January 2025 (UTC)

Thanks for your work, Synpath! I have to say though, I don't find the NIST plots (like 1 (H I) NIST ASD emission spectrum.png) particularly appealing. They are quite dark, the spectral lines are stylized to appear darker at the ends (I'm not sure why they do that but I find it distracting), and the plots lack numbers on the horizontal scale. I much prefer the automatically created SVG files by Nucleus hydro elemon, like Atomic spectrum of hydrogen.svg. Renerpho (talk) 02:12, 21 January 2025 (UTC)

@Synpath: I somewhat prefer the NIST ASD plot, though I would like it better if its scale bar were slightly brighter and more informative (specifically including the nm values like in the SVG version you uploaded). I find the continuous color bar in the SVG version too bright and distracting from the thin emission lines. Regarding on Renerpho's comment about the back fade being distracting in NIST plot, I wouldn't mind if that gets removed or not. I'm indifferent about it. Nrco0e (talk • contribs) 02:27, 21 January 2025 (UTC)

@Nrco0e: One advantage of the SVG plot is that it's user-generated (and somewhat automated), rather than directly downloaded from NIST as an image. The design of the SVG plots is completely up to us. I think that changing the scale in the downloaded PNG files is more tricky. @Nucleus hydro elemon: I agree that the continuous horizontal spectrum in your plots may be too distracting. Can you change that, and maybe make the numbers a bit larger to make it more readable at small image scale? Renerpho (talk) 02:33, 21 January 2025 (UTC) For transparency, I've asked Nrco0e off-wiki to chime in here; they were the one who initially noticed the problem. Renerpho (talk) 02:38, 21 January 2025 (UTC)

I make the scale larger and the continuous spectrum smaller in File:Atomic spectrum of hydrogen.svg. I will create radon's atomic spectrum soon, but I'm not sure if it is still needed as Synpath had uploaded spectral line images for all 97 elements. Nucleus hydro elemon (talk) 05:31, 21 January 2025 (UTC)

It looks much better now! Nrco0e (talk • contribs) 06:33, 21 January 2025 (UTC)

(Coming from an offsite discussion with Renerpho and nrco0e) I agree with the above sentiment as well. My thought process was that if these are going to be featured in infoboxes, the info contained therein should be relatively straightforward and trivial to see without squinting (within reason); my idea would be somewhat large(r) font numbers for each line labeling the wavelength, rather than a very small scale at the bottom which needs enlargement to make any useful info out of. I'm also not a fan of the NIST complete lack of labeling aside from an implicit unlabeled scale. exoplanetaryscience (talk) 02:40, 21 January 2025 (UTC)

If only more elements had nice demonstrations of their spectra (Genuine Helium spectrum, afaik).

I have uploaded the files and linked them at the Template:Infobox element/symbol-to-spectral-lines-image (diff). I'm sorry I don't have time to comment more thoughtfully, but I largely agree with the sentiments raised above and I especially find the scale bar unhelpful (I'm kicking myself that I didn't at least give the range in the file descriptions). As for style there's a long way to go and I think it's regrettable that there are not many photos of actual emission spectra on Commons. ⇌ Synpath 03:30, 21 January 2025 (UTC)

File-naming convention

Regardless of whether we use custom-generated SVG from the data or downloaded PNG from NIST, please consider naming the files in a way that is not redundant and easiest to use in templates. For example, "File:1 (H I) NIST ASD emission spectrum.png" has both the atomic number and the element symbol. That means a caller must have both of those pieces of data, even though they are each unique and a 1:1 correspondence. So, when {{Infobox element/symbol-to-spectral-lines-image}} is passed the symbol as a parameter, it needs a giant {{#switch}} to choose among a list of all of the full hardcoded filenames because it also needs the number. If the file was "File:NIST ASD emission spectrum (H I).png", it could just be a simple concatenation with whatever symbol is passed. DMacks (talk) 16:42, 21 January 2025 (UTC)

I definitely wasn't thinking about whether the template could be implemented more elegantly when deciding on a file name, I just was following what was already there and figured that adding the atomic number makes category sorting easier. As I see it the primary benefit to this brute-force switch statement is that any particular spectra can be easily and intuitively over-written when/if an image of an element's emission spectral lines are uploaded to Commons.

The template can also be refactored so it isn't brute-force and the #switch statement is only used to provide non-standard filenames and defaults to using the NIST ASD simulated em spectra. This would require passing the {{{Z}}} parameter to the refactored version or generating it from the symbol. The change is not difficult to do, but I think it's mostly a style choice with maybe a marginal benefit in readability to the current implementation. ⇌ Synpath 22:31, 21 January 2025 (UTC)

Certainly "reasonable reasons". As long as there's a reason for *some* pattern, I'm fine with it. DMacks (talk) 22:45, 21 January 2025 (UTC)

Categorization

[splitting off to revive subtopic]

This uniform new NIST set are all in c:Category:Simulated atomic emission spectra. But there are also other files that meet that description, including the disputed old set, and possibly others. I think it would be useful to have this canonical NIST set in its own category, which will among other things help clean up the chaos of c:Category:Atomic spectra and c:Category:Emission spectra. Any objection to c:Category:NIST simulated atomic emission spectra for this set and...I guess c:Matlab-spectra1.0 simulated atomic emission spectra for the old ones? And yes, I'm volunteering to do the recat'ing. DMacks (talk) 04:31, 24 January 2025 (UTC)

It was my intention to move all the previous Matlab generated images to c:Category:Simulated atomic emission spectra, but it slipped my mind after finishing up the initial upload. Further sub-categorizing makes sense to me as well. If its any trouble, I can take this on sometime this weekend. ⇌ Synpath 22:10, 24 January 2025 (UTC)

Looks good. Glad you know about cat-a-lot:) DMacks (talk) 23:45, 25 January 2025 (UTC)

IMO the old set should be placed in their own category, with a category name that explicitly states that they've been superseded, that their accuracy is dubious and may be entirely spurious. That should be the only category they appear in -- they should be retained for archival purposes, since they were used for so long, but going forward we don't want to risk people using them again. — kwami (talk) 17:42, 27 January 2025 (UTC)

Etymology entry on chloroform/iodoform

Both the pages for chloroform and iodoform reference a specific etymology as halides of "formyle", which is purported to be the archaic radical theory name for CH. Neither entry is cited - I suspect the appropriate citation would be Dumas's original writings on the discovery of chloroform. My 19th-century German isn't up to snuff to see if he actually references that specific "radical" in his naming, as opposed to simply relating its formula and composition to formic acid/formaldehyde. Some quick Googling only turns up references to "formyle" as a surname or the French translation of "formyl".

Can someone crosscheck if this is a claim that can actually be supported by Dumas's writings? There's some copyediting to be done regardless, since the articles mix the definitions of "radical" from then and now, but I'm loathe to prune the information if the specific history is there. Fishsicles (talk) 16:35, 6 February 2025 (UTC)

"Iupac nomenclature" listed at Redirects for discussion

The redirect Iupac nomenclature has been listed at redirects for discussion to determine whether its use and function meets the redirect guidelines. Readers of this page are welcome to comment on this redirect at Wikipedia:Redirects for discussion/Log/2025 February 6 § Iupac nomenclature until a consensus is reached. --MYCETEAE 🍄‍🟫—talk 22:10, 6 February 2025 (UTC)

Widely Retracted Author's Older Work

I frequently like to check Category:Articles citing retracted publications. Retraction Watch recently published an article [7] regarding Suman L. Jain. Despite her more recent work being frequently cited, I noticed that the category did not actually grow (indicating that the bot had not updated it yet to reflect the retracted papers). I searched the author's name manually and only came across two references on Wikipedia, both to an older work in Synlett - a journal with an impact score of ~2. This 2006 work was done through the same institute as the later retractions, and she is listed as the primary author as well. I am nowhere near qualified to check it, and I am unaware of what best practice would be to do if there are reasons to be concerned as well, and thus I would like to ask for wikiproject chemistry participants to look it over.

The two pages are: Methylrhenium trioxide and Pyridine-N-oxide Relm (talk) 08:40, 7 February 2025 (UTC)

References