Reactions

Letters to the editor

Incentives for science

I am writing to bring to your attention a jarring inconsistency in the Feb. 25 issue of C&EN. On page 3, the editorial, titled “Wrong Incentives,” encourages us to reevaluate how publishing is incentivized, a welcome conversation starter.

This contrasts with the ACS Central Science advertisement on pages 20 and 21. This advisement boasts of the journal’s time to publication along with the article views and altmetrics of a recent paper.

As a frequent ACS author and peer reviewer, I find this highly disturbing. For starters, article views tell nothing about the paper’s quality. An infamous 2009 communication in the Journal of the American Chemical Society titled “Reductive and Transition-Metal-Free: Oxidation of Secondary Alcohols by Sodium Hydride” was, for a period, one of the most viewed articles published in that journal (2009, DOI: 10.1021/ja910615z). This paper has been “withdrawn for scientific reasons,” and the high article views were due to interest in the erroneous paper, not quality science.

Second, altmetrics are inherently flawed as a measure of scientific impact. Altmetrics use a combination of tweets and Facebook posts, among other factors, to determine a score. It should be noted that for-profit services are available to promote materials on social media, including Twitter’s Quick Promote feature. The use of altmetrics incentivizes using advertisement tools to garner scientific impact. This is particularly dangerous with the advent of preprint servers such as ChemRxiv. The article featured in the page 20–21 advertisement was published to ChemRxiv the same day the article was submitted to ACS Central Science (Oct. 17, 2018). This means that, in principle, peer reviewers or editors could have their scientific judgment biased by the speed at which altmetrics accumulate for a preprint article. This is a process potentially accelerated by the use of a for-profit advertisement service. This is a very perilous situation that promotes short-term buzz over accuracy, reproducibility, and reliability. I think ACS Publications should consider if altmetrics and article views are the wrong incentives to be promoting.

Joseph Topczewski
Minneapolis

Non-BPA epoxy coatings

Though Sherwin-Williams may be commended for replacing bisphenol A (BPA) diglycidylether with tetramethyl bisphenol F diglycidylether (C&EN, March 11, page 32), that is only half the equation. What was used to replace BPA as the epoxy extender? Was this compound found in Ana M. Soto’s extraction tests? If so, at what levels? Are there different health concerns associated with this epoxy extender? Sherwin-Williams has been silent on this subject, and Soto’s estrogenicity assessment does not mention an epoxy extender.

Patent US9409219B2 seems to cover the composition of this epoxy coating, which claims the use of “catechol, hydroquinone, or resorcinol” as the epoxy extender (claims 11 and 56). This is affirmed in the body of the patent: “Some specific examples of suitable upgrade dihydric phenols include hydroquinone, catechol, p-tert-butyl catechol, resorcinol, substituted variants thereof, or a mixture thereof. Hydroquinone is a presently preferred compound.” Assuming that hydroquinone is the epoxy extender, and this is only an assumption, what are the health hazards of hydroquinone compared with BPA?

A safety data sheet for hydroquinone states, under section 2.2, “Harmful if swallowed”; “May cause an allergic skin reaction”; “Causes serious eye damage”; “Suspected of causing genetic defects”; “Suspected of causing cancer”; and “Very toxic to aquatic life with long lasting effects.”

A BPA safety data sheet states, “May cause an allergic skin reaction”; “Causes serious eye damage”; “May cause respiratory irritation”; “Suspected of damaging fertility or the unborn child”; and “Toxic to aquatic life with long lasting effects.”

Because of this information, and in the spirit of openness and transparency for which Sherwin-Williams has been praised, I ask that it take this one step further and disclose the identity of the epoxy extender used in its new epoxy can coatings. Along with this, any extraction test results for this compound and whether there is reason for additional tests, besides estrogenicity, based on the (possibly) different health hazards associated with the non-BPA epoxy extender.

Dan Hartinger
Mill Hall, Pennsylvania