4
$\begingroup$

I was reading an article about an independent pharmaceutical watchdog group called Valisure which recently released a report stating that they found Benzene, a toxic substance, in many bottles of consumer sunscreen. (Full report here: https://www.valisure.com/blog/valisure-news/valisure-detects-benzene-in-sunscreen/ )

Valisure LLC has tested and detected high levels of benzene, a known human carcinogen, in several brands and batches of sunscreen, which are considered drug products by the Food and Drug Administration (FDA), as well as in after-sun care products, which are generally regulated by FDA as cosmetics. Benzene is known to cause cancer in humans according to the U.S. Centers for Disease Control and Prevention, the U.S. Department of Health and Human Services, the World Health Organization, and other regulatory agencies

As a layman, it made me marvel at how they are able to do this. What is the process by which a chemist is able to start with a complicated multi-ingredient substance like sunscreen, and determine that it contains a particular molecule, like benzene?

Is there some kind of litmus paper that ONLY reacts with something containing benzene, and not ANY other substance? Or do they mix in substances known to react ONLY with the other sunscreen ingredients, leaving benzene behind? This sounds like it would be incredibly complicated.

I'd love some more insight about how this works

$\endgroup$
3
  • $\begingroup$ Analysis is also context based and, in particular, starts before the steps that ultimately lead to identification and quantification of a species. In particular a complex matrix (the ensemble that as be collected via sampling, the latter being already a critical step) can be treated as such to became simpler. For instance one can divide inorganic salts and ions from organic molecules, divide water soluble and water insoluble components, etc. Afterwards specific techniques are applied. The answer below goes pointing to one important example in analysing commodities, pharmaceuticals, etc. $\endgroup$ – Alchimista Jun 1 at 10:17
  • $\begingroup$ In case you are interested in reading the details of how the analysis proceeded, it starts on page 6 of this document: valisure.com/wp-content/uploads/… Note that there is a difference also between detect and quantify. Doing the latter accurately adds another layer of complexity. $\endgroup$ – Buck Thorn Jun 2 at 0:25
  • $\begingroup$ I suggest you narrow your question to how this particular analysis was performed, since that is interesting enough and sufficiently narrow in scope to be answered in a reasonably short post. Your question as it stands is the subject of analytical chemistry and the answer could fill one of more libraries. $\endgroup$ – Buck Thorn Jun 2 at 0:30
7
$\begingroup$

There are three common classes of methods:

  1. Separate the components of the mixture, then detect the amounts of the substances.
  2. Use a fancy method that can identify multiple substances in a mixture (such as training a dog to smell different substances, or nuclear magnetic resonance)
  3. Use a method that detects only your substance of interest, even when in a mixture (such as enzymes that only react with your substance of interest).

The last class of methods is your "fancy litmus paper". Examples of actual methods are a pregnancy test looking for a specific hormone in urine (class 3), a doping test looking for various illicit substances in a blood sample (class 1) or a chemist monitoring a reaction mixture to test for completeness of the intended reaction and for possible side reactions (class 2).

$\endgroup$
2
  • 2
    $\begingroup$ I don't know if I'd say using a trained dog is "common" (on a par with NMR) but ok in general #2 is common. Also, there is more to fancy spectroscopy than just NMR. Some people can look at an IR spectrum and say what's in a mixture, which I find even more amazing than a dog detecting benzene. My guess is the method applied to the sunscreen was GC-MS since benzene is volatile and the matrix is a mess that would certainly be unkind to NMR. Some kind of separation would probably be necessary for NMR detection. $\endgroup$ – Buck Thorn Jun 2 at 0:04
  • 1
    $\begingroup$ NMR cannot even detect benzene at parts per million level. Very insensitive to impurities. GC-MS is the way to go in the OP's case. $\endgroup$ – M. Farooq Jun 3 at 19:14
6
$\begingroup$

Most likely chromatography. Gas chromatography is often used for detecting volatile compounds. You vaporize the sample at the inlet of a very long, thin tube (the column). The sample then gets pushed along by a carrier gas (the mobile phase). The different components of the mixture interact differently with the stationary phase inside the column and so take different amounts of time to reach the other end. A detector at the other end registers how much stuff comes out as a function of time, and so you can tell how many dissimilar components you have in your mixture. It's common to use a mass spectrometer as the detector, in which case you can also look at the mass spectrum to identify each component, often by comparison to a library. For benzene this would be easy to match.

https://en.wikipedia.org/wiki/Gas_chromatography

$\endgroup$
3
$\begingroup$

Determination of residual solvents is a routine exercise in pharmaceutical industries and benzene is one of them. The technique is called gas chromatography. When you have a very complex mixture, such as sunscreen, one might use a technique called headspace gas chromatography. You would heat a sample in a very small tightly closed vial, at a desired temperature, and sample the "headspace" with the help of a syringe. The "air" in the syringe then injected into in a gas chromatograph, whose job is to separate and detect the components in the gas phase. Benzene is very volatile and easily detected by GC.

For legal purposes, a mass spectrometer is connected to the gas chromatograph or an HPLC , which confirms the exact identity of the molecule exiting the chromatography column.

Just keep in mind that, NMR is the last choice the detect impurities. It cannot tell you what components are present at parts per million or parts per billion level.

As others have hinted at, analytical chemists often do a targeted analysis. It will be a futile exercise to ask someone "Can you find all the components in this random sunscreen"? The answer is no. It will take a 6 months, to probably more to determine and identify all components from scratch and it will easily cost thousands of dollars.

$\endgroup$
4
  • $\begingroup$ So when presented with an unknown substance, for example a mysterious powder, chemists try multiple targeted analyses? How do they avoid running out of powder to test? $\endgroup$ – CodyBugstein Jun 3 at 17:20
  • 2
    $\begingroup$ Analytical chemist will not avoid an unknown, it just how much time and money is available along with the instruments. Identifying plant extract components may take 5 years of PhD work. There are many ways which may be used for non-destructive testing. $\endgroup$ – M. Farooq Jun 3 at 18:20
  • $\begingroup$ For that you still very expensive spectra libraries which contain spectra of thousands and thousands of known compounds. Infrared spectrum is the first way, once the spectrum of an unknown is collected, software tries to match it with the library. Second way is Raman spectrum, and finally liquid chromatography mass spectrometry, gas chromatography mass spectrometry. $\endgroup$ – M. Farooq Jun 3 at 18:20
  • 1
    $\begingroup$ @CodyBugstein if you invent a situation or a challenge, but even in practice, that is a possibility. One can run out of sample even when knowing exactly what it is but need to perform other characterisation. Your question tend to mix practice and principles. In principles we can do a lot. In many cases a melting point determination can already tell if we have a mixture, an impure sample, a pure compound... And so on. Of course quantification of everything depend very much on the level needed. $\endgroup$ – Alchimista Jun 4 at 9:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.