Three months of stripy nanoparticles controversy

In a post at MaterialsToday.com, David Bradley asked whether nanoparticles have lost their stripes?  and concluded:

There are now at least a couple of dozen comments on the THE article itself. If only there were some centralised system for pulling all the arguments together and perhaps tying them to the original papers from Stellacci and from Lévy. Perhaps we will one day see such a development in web 3.0. Meanwhile, we still don’t know for sure whether those gold nanoparticles are stripy or not!

In the absence of web 3.0, here is an attempt at providing a current picture of the controversy, focusing on the scientific arguments. Ethical issues such as data re-use and refusal to provide raw data are covered elsewhere.

First a quick reminder if you have not been following: the stripy nanoparticle hypothesis was first proposed in Nature Materials in 2004 by the group of Professor Stellacci (then at MIT and now at the EPFL). This hypothesis now forms the basis of 26 articles by the same group, mostly published in high impact journals including Nature Materials, Nature Nanotechnology, Nature Communications, Science, Journal of the American Chemical Society, Small, etc:  -1 , 0, 1, 2, 3, 4, 5, 6, 7 , 8, 9, 10, 11, 12, 13, 13a, 14, 15, 16, 17, 18, 19, 20, 21, 22 and 23. {apologies for the strange numbering}

ENGAGE? 

A number of technical questions have been discussed not just at Times Higher Education (25 comments) but also here at rapha-z-lab (over 100 comments), on the blog of Doug Natelson (23 comments) and in a few other places. One notable fact is the very unfortunate refusal of Francesco Stellacci and his co-authors to engage in the online discussion (with the exception of one comment by Sharon Glotzer on the article at Chemistry World). This can be contrasted with the way Phil Baran engaged with Blog Syn assessment of one of its papers as ‘difficult to reproduce‘ ; that latter case nicely demonstrates how post publication peer review combined with engagement from the criticized authors can lead to better science. Stuart Cantrill puts it succintly (referring not to this blog but to Blog Syn: “ENGAGE (and do it nicely). This is not a witch hunt, it’s for the good of science.”).

Scanning Tunneling Microscopy (STM)

The primary evidence for the existence of stripes is STM. In Stripy Nanoparticles Revisited (open access), we argue that the observed stripes are a scanning artefact rather than a feature of the particles. We base our conclusion on several argument: 1) a simple geometric consideration about projection from a sphere (the particle) to 2D (the image), 2) the direction of the stripes (always perpendicular to scanning direction), 3) the unlikely correlation between particles (see this video to understand the problem), and Fast Fourier Transform analysis. Stellacci and Yu response was published in Small at the same time as our article.

After the publications, Predrag Djuranovic contacted me. He had already demonstrated that the stripes were an artefact several years ago. He published on my blog results which demonstrate that the stripes could be obtained in the absence of nanoparticles and Matlab simulations which show how the STM feedback mechanisms  generate those patterns. Predrag was a graduate student in Stellacci’s group at MIT in 2005 when he generated these data… but they had not been published or communicated outside of MIT.

Philippe Moriarty, Professor of Physics at Nottingham and leading STM expert, confirmed our interpretation of the stripes as a scanning artefact, and, discussed the response of Stellacci and Yu, showing that the ‘stripes’ presented in the new images are ‘a fortuituous alignment of random noise‘.

At Doug Natelson’s blog,  SPMer said:

As somebody who has worked on high resolution SPM for many years, the first instinct on seeing the images in the Nature Materials paper mentioned is “those are crappy scans”. […] it should never have been published. It is a disservice both to the field and to the authors of the paper themselves.”

After having looked at Stellacci and Yu’s response, SPMer was disheartened:

[…] as an experimentalist I looked closely at figures 3 and 4 […] Two images are shown, and the authors pick at random two particles and claim that they see lines that match on the two particles. However, if we look at other particles in the figure, there appears to be absolutely no correlation in the two rotated images for the dimples or stripes on the particles. What about an autocorrelation analysis? […]

Still at Doug Natelson’s blog, another STM expert said:

I am an STM-er. For 10 years. Atomic resolution spectroscopy is the only thing I do. Granted I hunt for flat areas to do my work, but I do encounter a lot of nm scale “mountains”. My judgment (without it being worth anything here as I prefer to remain anonymous) is that the STM images in the papers by Stellaci are not proof of ligand ordering. I believe (but can’t prove that without actually repeating these measurements myself) that they are indeed feedback-loop ringing. […]

It is unfortunate that SPMer and the above STM expert chose to remain anonymous (Rich Apodaca’s thoughts on the choice of anonymous science blogging here).

At Rapha-z-lab, AFMhelp, a.k.a Peter Eaton, was one of the first to post a comment; he said:

Good work. The first time I saw those images, I was very very doubtful about them. I think it would be very easy to produce such images, by having some periodic noise in a scanning image of “normal” nanoparticles.

and later (in a discussion of this article):

The AFM data in the Nature Materials paper are nowhere near to being “proof”. Phase imaging of heterogeneity at the (small) molecular level on non-flat surfaces is extremely difficult. There would need to be more images than that shown. The stripes are interesting, but occur only on one or two particles…they do seem to be digital zooms of larger images, since many of the “features” seem to be single pixel in size.The AFM data is basically hard to interpret and more data should have been got before publication. It is also rather confusing in its presentation, and I think it would have been more fair to show the data without the cartoons drawn on top of the data, or at least include this data in the SI. This is in contrast to the original STM data which as discussed previously (and as was pretty much proven by Predrag) is completely artifactual.
Peter Eaton

Pep

Some other aspects of the STM evidence were discussed in minute details in a series of exchanges involving anonymous commentator Pep. Predrag Djuranovic, Philip Moriarty, Li Jinfeng, I, and others spent a considerable amount of time answering carefully the 20 or so comments that Pep left here over a period of a few days. Apects of that discussion were genuinely interesting and led for example to a refined understanding of the ‘projection argument’ (using a suggestion from a comment at Doug Natelson’s blog: positive post peer review in action!), but, as Dmitry Baranov noted  very early in the discussion (on Twitter) “@raphavisses and yeah, looks like Pep got more than just a quest for fairness there.” For this reason, I won’t attempt to summarize those long-winded arguments in this post (they can be found in particular under these two posts). It turned out that Pep was Pep Pàmies, Editor at Nature Materials, the journal which has published 4 of the stripy articles including the inaugural one (and which rejected a first version of Stripy Revisited; see my Letter to the Editors here). My comment identifying Pep is here as well as the robust discussion that followed; see also Ben Goldacre take on this episode, as well as Pep Pàmies note entitled “On my comments on Lévy’s blog“ [update 404: Pep has now removed the note from the web!]. Dave Fernig has responded to Pep Pàmies apologia for reuse of data on his blog and Philip Moriarty has provided a comprehensive response as a guest post.

Transmission Electron Microscopy (TEM)

In addition to scanning probe microscopy, it is claimed in the original 2004 article paper that the existence of stripes is backed up by TEM (and XRD). In stripy revisited (supporting information, section 2), we show that “no conclusion regarding the structure of the capping layer can be drawn from this image”. In their response (and at Chemistry World), the authors repeat that TEM backs up the existence of stripes but do not address our criticism. Online post publication, the TEM evidence has been brought up by Pep too (see above) and others. In a strongly worded comment (a response to Pep), Li Jinfeng wrote:

To claim the existence of stripes made of alkyl thiolate molecules based on a few dark spots in rather bad-quality TEM images…? Unheard of in the TEM community!

Elias wrote (in another thread):

Here is more TEM evidence for monolayer segregation that you seemed to have ignored: http://pubs.acs.org/doi/abs/10.1021/nn204078w

The reference here is to really interesting paper, but it is not about stripes nor does it claim to show the existence of stripes, as I responded to Elias here.

The last (?) word on the TEM evidence for stripes goes to David A Muller, Professor of Physics at Cornell. He is a leading electron microscopy expert and the motto of his group his ‘Understanding Materials, Atom by Atom“. At Doug Natelson’s blog, he says:

Funny thing is the TEM image from their appendix, Fig S2a, that is cited as independent confirmation is also an instrumental artifact. The ring of black dots is the out of focus point spread function (basically a Fresnel fringe). This is a very common problem for casual users of a TEM who are looking for core-shell nanoparticle structures. By changing focus, either a dark or bright ring can be created. Going in to focus will make it go away. The focus of the image can be determined from a quick FFT of the amorphous background, and sure enough, the passband is about a factor of two off from the optimal defocus.

NMR

While the stripy controversy was about to become public, an article ‘confirming’ the existence of stripes by NMR was published in Nature Communications. No independent post publication peer review yet. Do you know an NMR expert who could comment on the Nature Communication paper? The invitation for a guest post is open.