And yet there are stripes!

Mathias Brust, Professor of Chemistry at the University of Liverpool

Mathias Brust

This is a guest post by Mathias Brust, Professor of Chemistry at the University of Liverpool

Followers of Raphaël Lévy’s blog on “stripy nanoparticles” have been waiting eagerly for the outcome of an experimental study conducted by independent experts to evaluate the veracity of Stellacci’s et al. original interpretation of their STM data as the presence of molecular stripes in certain thiolate-capped gold nanoparticles.

I assume that the two papers recently published in ACS Nano and Langmuir, respectively, both co-authored by recognized STM experts, detail the various findings of this study. Unfortunately, the authors shy away from making clear statements in relation to Stellacci’s previous work, published most prominently in 2004 in Nature Materials. The new images obtained with state of the art modern equipment, unlike the original ones from 2004, show hardly any discernible features. Indeed, in the absence of Stellacci’s previous claims, nobody would see anything special in these new data, which just look like rather well done STM images of gold nanoparticles, with some inevitable noise. Unfortunately, rather than stating that there is a clear discrepancy between the original data from 2004 and the new ones, the authors employ an arsenal of image analysis techniques to convince presumably themselves and evidently the referees that the now barely discernible ripples at the noise level represent indeed all the features Stellacci et al. had previously reported in much clearer images obtained with inferior instrumentation and without sophisticated image analysis. The new study thus implicitly admits interpretation errors in the original work, while explicitly aiming to corroborate it. A tacit assertion that it is hard to do these experiments, which are prone to imaging artifacts, is perhaps the only hint that maybe before we were not looking at the real thing at all.  I strongly advise everybody interested in this debate to look at the newly published images, and again at the ones from 2004, to make up their own mind on this topic. I suspect that an even more carefully conducted study in the future may reveal that the particles have no discernible features at the molecular level and that the stripes are “invisible” by STM. A new class of “stealth stripes” would make a nice cover for Nature Materials.


  1. Is it possible to get a comment from the STM scientists on this paper?

    I think this is the main discrepancy that a non-expert (like me) is unable to resolve. What do the expert STM scientists (who were not on the 2004 paper) think about the validity of the data and conclusions in the 2004 paper?

    I can understand the conflict for Stellacci, but not for them (not knowing much about their other work, but presuming that they were brought on to lend credibility to Stellacci’s perspective from their lack of obvious COI).


      1. Also would like to suggest you put up the images from the 2004 Nat. Materials, ACS Nano and Langmuir papers on this blog post (given the paywalls). Arranging them side by side would make sense, I think.


      1. Thanks for trying Raphael.
        What a shame that an author wouldn’t take the opportunity to write casually about the work they are involved in.

        I stil can’t reconcile why supposedly good STM scientists can’t agree on (what appear to be) a very fundamental aspect of a paper that clearly needs resolving. Your openness in promoting discussion goes a long way to suggesting what the correct interpretation of data is. I will keep looking out for your future posts and papers.


  2. Interestingly the Langmuir paper, which consists of most of the quantitative Fourier analysis, does apply the new analysis to four of the Nature 2004 images. So they don’t seem to be disregarding this old data.


    1. Yes, this is actually true. What I should have said is that I was missing a discussion as to why the new data do not show the features that are so clearly seen in the 2004 paper, actually much less so in the 2004 data used in the recent Langmuir paper. I would expect that STM data obtained in 2004, as long as they refer to actual features of the sample, can be reproduced at higher quality with more modern equipment. I miss a discussion of the fact that this is not the case here.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.