Wikipedia:Reference desk/Archives/Science/2022 November 9

= November 9 =

Arsenic compound used in WWI weapons
Zone Rouge says:

Some areas where 99% of all plants still die remain off limits (for example, two small pieces of land close to Ypres and Woëvre), as arsenic constitutes up to 175,907 mg/kg of soil samples.

I found that the source of this arsenic was from detonators:

A study, published in 2007, claimed the levels of arsenic, used in the detonators, were between 1,000 and 10,000 times the level usually found in the ground.

What was the chemical compound containing arsenic that was used in WWI-era detonators?

Helian James (talk) 00:12, 9 November 2022 (UTC)
 * That seems to be a typo or misunderstanding. It should have been detonations, not detonators.  Recovered chemical weapons in World War I were detonated there after the war.  See:
 * --136.56.52.157 (talk) 04:04, 9 November 2022 (UTC)
 * The chemicals include diphenyl-chloroarsine and diphenylcyano-arsine
 * --136.56.52.157 (talk) 04:33, 9 November 2022 (UTC)
 * This publication is the source of the value 175,907 mg/kg. The number seems ridiculously precise to me. --Lambiam 10:28, 9 November 2022 (UTC)
 * Why question their results when the precision of the measured data is lab dependent and any less precise value(s) than they report are not maximums of their empirical data sets? Modocc (talk) 14:26, 9 November 2022 (UTC)
 * See False precision --136.56.52.157 (talk) 15:01, 9 November 2022 (UTC)
 * For more on that subject, see today's xkcd. --174.89.144.126 (talk) 23:11, 9 November 2022 (UTC)
 * The question/possibility of false precision was raised by Lambiam, but he did not substantiate why it seems to be the case with the published source. Modocc (talk) 15:08, 9 November 2022 (UTC)
 * I'll let respond specifically; in the meantime read False precision. Regardless of the source, there are many variables that would affect the data (minute changes from day to day, variability of sample locations, etc.). At best, one can obtain an average or mean with a margin of error. The entire volume of soil in the study area surely isn't exactly 175,907 mg/kg without variation. 136.56.52.157 (talk) 16:38, 9 November 2022 (UTC)
 * Yes, I know that empirical data sets that are produced are affected by circumstances and methodologies (and subject to peer-review) and are reported on accordingly. With sampling, it's likely that localized contamination is higher than what was sampled, but the authors are reporting on what was found. Not what could be found. Modocc (talk) 17:07, 9 November 2022 (UTC)
 * A specific measurement on a specific sample taken on a specific date might be "accurate" but meaningless. 136.56.52.157 (talk) 17:17, 9 November 2022 (UTC)
 * Empirical recorded maximums are typically not meaningless though. Modocc (talk) 17:30, 9 November 2022 (UTC)
 * That data might make sense in its original context, but journalists using it for a general media article is nonsensical. 136.56.52.157 (talk) 19:03, 9 November 2022 (UTC)
 * The cited 2007 article is behind a paywall (and the DailyMail is not a reliable source the last time I checked and it's not cited). Modocc (talk) 19:49, 9 November 2022 (UTC)
 * If they had divided the soil sample into three equal parts after thoroughly mixing it and then had tested these parts separately, they might have found three values like {175,837 mg/kg, 175,893 mg/kg, 175,921 mg/kg}. There is no way the spread would have been in the order of 1 mg/kg. Copying the number as read off a measuring device in a table as one of the entries among the 14 samples tested is one thing; reporting it that way in the abstract is something else. --Lambiam 19:33, 9 November 2022 (UTC)
 * Suppose the mixing you suggest is inadequate and its all found only in one subsample (in 1/3 the sample), the result for the entire sample is still the same as reported. The question regarding its meaningfulness depends on its distribution in the soil, how many samples per area were measured, and whether or not any of the data can be considered as outliers, none of which is in the paper's abstract and which is the only part available to me. I have not nor plan to review their results. Modocc (talk) 20:05, 9 November 2022 (UTC)
 * Suppose you want to examine how fast a snail can travel. You do a number of experiments, and then report in the abstract, "in one run, the high speed of 13,914 μm/s was recorded". Perhaps you think this is normal, but I think it would make the publication a candidate for the Ig Nobel Prize. --Lambiam 13:34, 10 November 2022 (UTC)
 * Perhaps the snail needs microscopic brain surgery while not being immobilized... Yes, precision is often insignificant, furthermore I'm not a soil scientist and I have limited knowledge of the field and enough respect for it to not ridicule a published paper I've not read. I expect there are established practices that are followed that given their sampling precision are quite capable of mapping and quantifying very sparse contaminants. That arsenic is relatively abundant and ubiquitous ought not matter. Modocc (talk) 14:15, 10 November 2022 (UTC)
 * Suppose you want to examine how fast a snail can travel. You do a number of experiments, and then report in the abstract, "in one run, the high speed of 13,914 μm/s was recorded". Perhaps you think this is normal, but I think it would make the publication a candidate for the Ig Nobel Prize. --Lambiam 13:34, 10 November 2022 (UTC)
 * Perhaps the snail needs microscopic brain surgery while not being immobilized... Yes, precision is often insignificant, furthermore I'm not a soil scientist and I have limited knowledge of the field and enough respect for it to not ridicule a published paper I've not read. I expect there are established practices that are followed that given their sampling precision are quite capable of mapping and quantifying very sparse contaminants. That arsenic is relatively abundant and ubiquitous ought not matter. Modocc (talk) 14:15, 10 November 2022 (UTC)


 * As to how arsenic was used in munitions, Chemical Warfare and Medical Response During World War I says:
 * As the American war effort intensified, research expanded to include offensive weapons, resulting in numerous discoveries, including the creation of one of the conflict’s only new chemical weapons, an arsenic-based agent similar to mustard gas called lewisite (β-chlorovinyldichloroarsine). Synthesized in his laboratory by Wilfred Lee Lewis [sic], this deadly substance was soon mass-produced by the military under the direction of chemist and future Harvard president James B. Conant.
 * This may not be the only use, but its all I have found so far. Alansplodge (talk) 15:48, 10 November 2022 (UTC)
 * Weapons of War Litter the Ocean Floor says:
 * Military scientists had weaponized some stocks of sulfur mustard by adding arsenic oil and other chemicals. The additives made it stickier, more stable, and less likely to freeze on the battlefield.
 * Alansplodge (talk) 16:04, 10 November 2022 (UTC)
 * See also Adamsite. Alansplodge (talk) 16:09, 10 November 2022 (UTC)