Wikipedia:Reference desk/Archives/Science/2020 April 9

= April 9 =

Getting group 12 elements in oxidation states beyond +2
How likely is it?

Now, of course, I know that no such compounds are experimentally known for sure (mercury(IV) fluoride has been claimed, and disputed). For the relativistic copernicium, it's predicted to be possible, but that is because of 6d destabilisation. So what are the chances for less relativistic Zn, Cd, and Hg to surpass +2?

I know one paper suggested Zn(AuF6)3, but it's been pointed out that this is probably not thermochemically stable and anyway Zn(II) with oxidised ligands.

Does anyone know of any other predictions in the literature? Given that such things as helium and argon compounds have been targeted in predictions, it strikes me that something like this should have been, too. That the d electrons are involved in strengthening the bonding of group 12 compounds is not really in doubt (just look at the bond lengths, you'll see that 3d with its radius must have significant overlap in Zn compounds); but can we get higher oxidation states (since we can with difficulty get CuIII and AgIII)?

(I asked this of Droog Andrey already – he's a computational chemist – but curiosity overwhelmed and hence I brought it here too.) Double sharp (talk) 05:04, 9 April 2020 (UTC)
 * Look at the fourth ionization energy Molar ionization energies of the elements for zinc, and you will see it is 5731 kJ mol−1. This is just a little more than for copper, which can have a +4 in a ternary fluoride. Nothing with higher energy forms ionic compounds, but carbon can form 4 bonds, presumably the energy formed per bond is very high due to the small size of the carbon atom. We have no data for cadmium or mercury. Copernicium crystalline substances will be imaginary as they will not be stable enough to observe a solid even if enough can be created. Molecules may be possible. Graeme Bartlett (talk) 04:17, 11 April 2020 (UTC)
 * Interesting! Atomistry gives 5300 kJ/mol and 4400 kJ/mol for the 4th ionisation energies of Cd and Hg respectively (they pretty much agree with our Zn value). So it looks like the possibility of ZnIV, CdIV, and HgIV is not that far-fetched, even if no one has managed it yet. Double sharp (talk) 06:11, 11 April 2020 (UTC)

OK, I found another paper suggesting that oxidation states beyond +2 for these elements may be stabilised in anionic complexes, suggesting [MF4]− or [MF5]2− as serious possibilities for the +3 oxidation state (so, something like how TlIII is stabilised in anionic complexes like TlI4−). So, it looks like there may be a bright future for such things and a chance to destroy the paradigm that group 12 is strictly a post-transition group. ;) Double sharp (talk) 10:11, 11 April 2020 (UTC)
 * That said, as these are soft cations, probably Cl is a better choice than F for this. Double sharp (talk) 06:56, 8 December 2023 (UTC)

More possibilities for HgIV! Hg[OTeF5]4 and Hg[AsF6]4 seem to be particularly promising. Double sharp (talk) 10:23, 11 April 2020 (UTC)
 * 10.1021/ja002360s discusses [Hg(CO)4]4+ computationally. As a general lead ref (now 25 years old), I found 10.1021/ic00088a012, which has one of the same authors as your 2008 10.1021/ic702384y. As a search annoyance, CdO2 is "peroxo" Cd(II) not "dioxide" Cd(IV). DMacks (talk) 03:31, 13 April 2020 (UTC)

The next year, a computational paper came out suggesting indeed the possibility of Zn(III), when some super-electrophilic anions are involved. :) Double sharp (talk) 22:31, 16 August 2022 (UTC)

Is possible to oxidate ethanol with hydrogen peroxide?
(Google translate helped me for this message) Hello! I don't know in the other countries, but in Italy you can find and buy 96% ethanol for food use in every supermarket. If I mixed a little amount of 96% ethanol with hydrogen peroxide (wich I can find in every supermarket, 3% for medical use or 30% for hairstyling use), can I get acetaldehyde or even acetic acid? Thank you--87.17.174.42 (talk) 12:18, 9 April 2020 (UTC)


 * This article suggests that you can react them with a suitable catalyst (ferric iron) to produce acetaldehyde and diethylacetal, although I have no idea whether that would work at the concentrations available. I can only view the abstract. Why would you want to do that anyway? Mikenorton (talk) 14:48, 9 April 2020 (UTC)


 * As the whole world has suddenly pivoted to taking safety advice from CDC (after ignoring the very same advice for many years), may I recommend the CDC's School Chemistry Laboratory Safety Guide? Remember, what we're trying to do here is use real scientific knowledge to lower the statistical likelihood for the entire group so that any individual outlier will avoid harm.
 * If you're doing experimental chemistry in the classroom or the kitchen, you should probably be a little bit cautious: if you don't know exactly what your reaction will do, ... maybe you shouldn't do it. It is very easy to create really dangerous hazards - fires, chemical burns, poisons, and toxic gases - using the ordinary household chemicals that you can easily buy in the supermarket.  Don't be stupid, don't mix things unless you know what it's going to do; make sure you're in a safe area that won't easily burn, that has adequate ventilation, and ... maybe consider reading introductory chemistry textbooks for a few years before you start doing home lab experiments?
 * Here are some Material Safety Data Sheets from Fisher:
 * Anhydrous ethanol
 * Hydrogen peroxide
 * You can find the exact MSDS for the exact product you own from the manufacturer who made it. If you're really doing kitchen-chemistry, you should read those data sheets.  Pay real close attention to the ethanol - lots of people keep dying because they don't understand the difference between subtly-different types of commercially-available extremely poisonous ethanol mixtures!  Things get really confusing and troublesome when you cross international boundaries, and the rules change about what chemical may- and may-not be called "alcohol" on its packaging.
 * Even something as innocuous as grocery-store ethanol is marked as a "hazardous chemical" on its MSDS. It's pretty unlikely you'll be hurt by these things - but even if there's a four-per-hundred-thousand chance of serious injury, the whole world will shut down.
 * From the American Federal Emergency Management Agency: Keep Your Family Safe From Household Chemicals.
 * For our Italian-speaking friends: si prega — essere sicuro e non sperimentare con sostanze chimiche in casa domestica:
 * Sicurezza chimica from the Ministero della Salute
 * Nimur (talk) 16:32, 9 April 2020 (UTC)
 * Ti rispondo in italiano, sia perché lo sono pure io sia perché già il mio inglese è una vergogna nei dialoghi normali, figurati in quelli tecnici. Al di la del lungo discorso sulla sicurezza in casa, che da tutto tranne che una risposta alla tua domanda, l'acqua ossigenata non è un agente ossidante sufficientemente forte da riuscire ad ossidare l'etanolo, per lo meno in assenza di catalizzatori. Infatti solitamente si indicano ossidanti più potenti, come il cromo esavalente o il permanganato, per fare la reazione di cui parli. Anche la candeggina riesce a farlo (con sottoprodotti anche tossici). Infine ricordo che l'acqua ossigenata e l'etanolo rientrano nella ricetta che l'organizzazione mondiale per la sanità consiglia di preparare come igienizzante per le mani per chi deve operare chirurgicamente in paesi dove i normali igienizzanti per le mani non sono disponibili. Quindi etanolo e perossido di idrogeno non reagiscono. Ricordo comunque (in modo meno drammatico di Nimur) di evitare mischiotti strani in casa. Ciao--93.38.168.239 (talk) 17:37, 9 April 2020 (UTC)
 * This WHO guidance (from 2010) is for pharmaceutical professionals, and provides special guidance about handling raw materials - and also provides extra cautions about international variations in the source materials. They also provide specific guidance about flammability and corrosive hazards of hydrogen peroxide: "H2O2 adds an important safety aspect, however the use of 3–6% for the production might be complicated by its corrosive nature and by difficult procurement in some countries.  Difficulties sourcing satisfactory H2O2 resulted in the need to import..." (instead of producing hand sanitizers locally).
 * Nimur (talk) 17:46, 9 April 2020 (UTC)
 * Nimur, everything you are saying has nothing to do with the question. You only are creating confusion with unnecessary articles about the risks and dangers of hydrogen peroxide. The question is "can the hydrogen peroxyde oxidate the ethilic alcool?" and the answer in "no, unless you use catalysts, because it is not an oxidizer strong enough to do this". Stop--93.38.168.239 (talk) 18:10, 9 April 2020 (UTC)
 * Well, that's not entirely accurate either, as that is not how catalysts work. Catalysts can increase the rate of a reaction (oftentimes dramatically so), but they cannot cause a reaction to occur that otherwise would not occur without the catalyst. If hydrogen peroxide does not have a reduction potential suitable to oxidize ethanol, no catalyst will cause an oxidation reaction between them to take place. However, if it does have a reduction potential capable of oxidizing ethanol, but only very slowly (to the point that the reaction appears to not take place at all), the presence of a catalyst can increase the rate of reaction to something functionally useful. --OuroborosCobra (talk) 18:42, 9 April 2020 (UTC)
 * There are a bunch of literature refs for various catalysts for using the oxidizing ability of H2O2 to oxidize ethanol and similar compounds. Iron ions seem popular, some copper, and palladium examples also. They seem to stop at the carbonyl state (either aldehyde, or "diethyl acetal" deriative thereof that can be easily converted back to it). Heat, light, and gamma-radiation can also accomplish it to a measurable rate (without 2).
 * Here I'm headed beyond what the OP asked, but it leads to an on-topic additional aspect of rate vs ability... Tungsten catalysts seem to go to the carboxylic acid, therefore both of:
 * [alcohol] &rarr; [aldehyde]
 * [alcohol] &rarr; [acid]
 * are within the "oxidizing power" of H2O2. And also:
 * [2 aldehydes] [1 alcohol] + [1 acid]
 * is often a favorable equilibrium (Cannizzaro reaction), therefore
 * [2 alcohol] &rarr; [1 alcohol] + [1 acid]
 * is more favorable than
 * [2 alcohol] &rarr; [2 aldehyde]
 * However, some reactions do that latter successfully, so the first oxidation step (alcohol&rarr;aldehyde) must be pecifically accelerated compared to the second (aldehyde&rarr;acid).
 * By the by, the standard safety pages are now called SDS (Safety data sheet) not MSDS, as of 2012ish. They're still often ridiculously useless in many contexts, but the WP refdesks aren't a place to provide medical or legal info regardless. DMacks (talk) 03:48, 10 April 2020 (UTC)
 * Is that alcohol actually for consumption, meaning, not denatured alcohol? Important if you envision using any resulting product in food (acetic acid is the stuff that makes vinegar vinegary). And being an acid, concentrated solutions are dangerous and need careful handling (typical vinegar is around 3% acetic acid). --47.146.63.87 (talk) 02:12, 10 April 2020 (UTC)
 * Good point. But iron and hydrogen peroxide might take care of whatever's used for distillation/denaturing anyway. Pelirojopajaro (talk) 14:30, 16 April 2020 (UTC)
 * Is that alcohol actually for consumption, meaning, not denatured alcohol? Important if you envision using any resulting product in food (acetic acid is the stuff that makes vinegar vinegary). And being an acid, concentrated solutions are dangerous and need careful handling (typical vinegar is around 3% acetic acid). --47.146.63.87 (talk) 02:12, 10 April 2020 (UTC)
 * Good point. But iron and hydrogen peroxide might take care of whatever's used for distillation/denaturing anyway. Pelirojopajaro (talk) 14:30, 16 April 2020 (UTC)


 * I am revisiting this topic after it's been in the archive for a few days ... because this week's CDC Morbidity and Mortality Weekly Report (Volume 69, April 20, 2020) has a special section on Cleaning and Disinfectant Chemical Exposures and Temporal Associations with COVID-19.
 * Nationwide, poison control centers are reporting a significant jump in people being poisoned by household chemicals, "representing overall increases of 20.4%"... "there appears to be a clear temporal association with increased use of these products" for COVID-19-associated disinfection.
 * In particular, the report calls out two case-studies: in the first, a home-made cleaner made from household chemicals resulted in emergency hospitalization. In the second case study, a child required hospitalization after contact with an ethanol-based hand-sanitizer.
 * These case-studies are just two examples: "daily number of calls to poison centers increased sharply at the beginning of March 2020 for exposures to both cleaners and disinfectants..."
 * Nimur (talk) 18:05, 21 April 2020 (UTC)

Effects of DRD4 7R polymorphism
It's well known that the DRD4 protein has a polymorphism known as "7R" and that this polymorphism has been related to many psychosocial things. But are its biological effects on DRD4 known? Say, does it cause an overproduction/underproduction of DRD4 protein? I've been looking for evidence but didn't find anything. JoJo Eumerus mobile (main talk) 16:47, 9 April 2020 (UTC)
 * Our article says which is somewhat confusing since the first paragraph seems to almost contradict the last one, but maybe the sources will explain better. (It's possible that 6-10 are higher than 4, but 7 isn't as strong as the other 6-10.) Nil Einne (talk) 19:44, 9 April 2020 (UTC)
 * One source says 7R reacts stronger to dopamine than 4R and 2R, the other says the opposite, but one used frog oocites and measured the current in "G protein-regulated inwardly rectifying potassium channels", the other used hamster ovaries cells and measured the "concentration‐dependent inhibition of the forskolin‐stimulated cyclic AMP levels". Beside that maybe does the same receptor work differently in different animals.
 * The promoter of DRD4 shows also polymorphism, but this seems not to influence its activity, see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4833125 . Therefore it could be that only the reactivity to dopamine varies, not the number of DRD4 molecules. 2003:F5:6F05:8A00:F406:AA94:A204:5533 (talk) 15:17, 10 April 2020 (UTC) Marco PB
 * So I take that it is not known with certainty what the biological effects of this polymorphism are. Jo-Jo Eumerus (talk) 12:30, 13 April 2020 (UTC)

? per ? example
I have a lesson in which I compare Euclidean distance to cosine distance. I currently plot three Miles per Gallon points with miles on the X axis and gallons on the Y axis. I show that while 300 miles to 10 gallons is closer to 290 miles to 12 gallons, it is on the same vector as 240 miles to 8 gallons. The problem: The graph needs to include the origin to demonstrate cosine similarity and the plot points are very very far from the origin. So, what can I use instead of miles per gallon? What ? per ? is common with both the numbers being between 1 and 20? 97.82.165.112 (talk) 18:49, 9 April 2020 (UTC)
 * Why not just label the X axis from 0 to 400 miles with tick marks every 20 miles? --Khajidha (talk) 19:21, 9 April 2020 (UTC)


 * Why call it cosine distance? Why not tangent distance or else sine distance? What about: You can buy 8 oz bottles of pomegranate juice for $15.99, or 250 mL (about 8.45 oz) bottles for $16.99. Which is the better deal? --Lambiam 20:49, 9 April 2020 (UTC)
 * Because that is what the homework assignment specified? Cosine similarity might be relevant. 107.15.157.44 (talk) 21:07, 9 April 2020 (UTC)
 * OK, thanks. If the scales represent quantities in different units, then both Euclidean distance and cosine similarity are meaningless measures. I mean "meaningless" in the sense of "nonsensical", "devoid of sense", "pointless", "absurd". Comparing them is a waste of anybody's time and intellect. If this is a lesson you are developing in order to teach it, please do not expose your students to this. Use equal units, for example wealth before vs. after, and two (or more) investors following different strategies. Even then I don't get the point, since one is a similarity measure on pairs of points and the other on pairs of vectors, so it is like comparing apples and aardvarks. --Lambiam 05:44, 10 April 2020 (UTC)
 * So, your claim is that in knn clustering, any and all attributes use the same unit or it is all nonsense. It is complese nonsense to cluster patients using blood pressure, measured in mmHg, along with cholesterol, measured in mg/dL, along with hbA1c, measured in %. According this this claim, we need to toss out over 50 years of medical research.
 * The lesson is: When do you consider Euclidean distance to be a better form of measuring distance between two points in Euclidean space and when do you consider the angle between the vectors from the orrigin to the points to be a better measure of distance between the points? As with all tools, if you use the wrong tool for the job, you won't get what you want in the end. The problem in the example is that when projecting a graph on the screen where the points are at an extreme distance from the origin, it is very hard to see those tiny little dots. The origin is necessary to draw a vector from the origin to the points. You can't simply compress the axis because then all points end up on top of one another and you can't show distance between them. So, the goal is to have two measures that are in the same range, one on the Y axis and one on the X axis. 97.82.165.112 (talk) 10:49, 10 April 2020 (UTC)
 * In such applications the variables tend to have a Gaussian distribution, and one uses the standard variation of each variable as the unit for its scale, making the scales comparable. If you have just two data points, using their standard variations as yardstick makes no sense. --Lambiam 14:04, 10 April 2020 (UTC)
 * I get plenty of separation for the 3 points you gave us. Unless your graph is extremely tiny, compressing the axis shouldn't cause those points to "end up on top of one another". --Khajidha (talk) 12:41, 10 April 2020 (UTC)
 * Suggested from my wife: I went with Y axis having inch-squared area of a pepper that has capcasin on it and X axis having inch-squared area of the pepper. My three data points are A:(3.3,6.6) B:(3.4,6.5) and C:(4.2,8.4). If I ask which two are similar, the distance between A and B is much less than A-C or B-C. But, that really just looking at the size of the pepper. If I wanted to ask about spiciness, that is percent of capcasin, so I look at Y/X, which is a vector. A and C sit on the exact same vector, so they have the same Y/X ratio. In the end, "similar" is dependent on what is being asked. How you measure it in Euclidean space is, as expected, based on what is being asked. Now, I can have a graph with huge dots and thick lines so the blind kids who always sit in the back row will see it and have a feeling of confidence before I get to set similarity. 97.82.165.112 (talk) 17:07, 10 April 2020 (UTC)
 * I think you mean “capsaicin”. --Lambiam 18:20, 10 April 2020 (UTC)
 * It's a good thing I labeled it "C%" 97.82.165.112 (talk) 19:21, 10 April 2020 (UTC)
 * Except that C% could easily be misunderstood as "percentage of carbon".--Khajidha (talk) 20:05, 10 April 2020 (UTC)