Stripy revisited posts; where to start…

Sorry: this page is quite a bit out of date… the good news though is that you can go through the stripy revisited posts by browsing this link.

I have added a Timeline post which is also a useful place to get an overview of the story.

If you are new to the story, this guide might help you to navigate:

Start with the first post which introduces the topic and contains the references (hyperlinks) to the 23 articles which are based on the “stripy” hypothesis/artifact. That post has many interesting comments; why not add your own?

The second post is “responding to the response”. We had made a simple geometric argument that proves that the stripes cannot be a real feature at the surface of nanoparticles. That post addresses the shortcomings of Stellacci’s response – go there if you are interested by how scanning tunneling microscopy (STM) works (or if you are an expert and would like to comment).

The third post is Philip Moriarty extremely important contribution (guest post); he is an STM expert but his post is written for the non-expert too.

Beyond the mis-interpretation of a classical well-known artifact, one very disturbing aspect of this story is the publication in respected journals of a number of claims with literally no experimental evidence at all, or extremely little evidence. I give specific examples from three different articles in these 4 posts: 1) electron microscopy (Nature Materials 2004); 2) water-soluble stripy nanoparticles (Nature Materials 2008 and Chem Comm 2008; includes duplication of data); 3) non-specific interactions of proteins (Nature Materials 2004; claim repeated multiple times in following papers); 4) mercury-capturing stripy nanoparticles (Nature Materials 2012).

Seven years of imaging artifacts: What gives? Predrag extremely strong demonstration that stripes are an artifact… demonstration which dates back from 2005 when he was a graduate student in Stellacci’s lab.

Scientific controversy is healthy is a reflection inspired by an update on Douglas Natelson blog (see link below).

I have also published the letter to the Editor that accompanied the submission to Nature Materials in 2009. It included specific recommendation to ensure a fair reviewing process, including the request that the article should be sent to STM experts (and other relevant experts) outside of the nanoparticle field.

Francesco Stellacci has been invited (and the invitation remains valid) to comment here and even to write a guest post addressing the points made. For the moment, he has made the choice to ignore the controversy.

Comments (even anonymous) on the scientific debate as well as the broader issues are extremely welcome. You will be in good company too: thank you to Mathias Brust, Simon Higgins, Philip Moriarty,Vlad Tsuruk and all the others for sharing their views here.

Pep Pàmies, Editor at Nature Materials commented here prolifically as Pep; see my comment here as well as the robust discussion that followed; see also Ben Goldacre take on this episode, as well as Pep Pàmies response entitled “On my comments on Lévy’s blog“. Dave Fernig has responded to Pep Pàmies apologia for reuse of data on his blog (here) and Philip Moriarty has provided a comprehensive response as a guest post here.

Outside of Rapha-z-lab (let me know if I am missing something):

1. Chemistry World covers stripes controversy
2. Alan Dove comments on the web and peer review in his post entitled Do These Stripes Make My Nanoparticles Look Weird?
3. Dave Fernig has a growing list of posts, all worth reading (what else can I say, he is my boss); the “over the line” series relate to the issue of data duplication/self-plagiarism.

4. Douglas Natelson, professor of Physics at Rice University, discusses “a Nano Controversy” on his blog

5. Peculiar activity at Nature Materials… Ben Goldacre

6. On my comments on Lévy’s blog [Pep Pàmies, Editor at Nature Materials, in personal capacity]

7. Times Higher Education; Slow is no way to go, argues researcher, Paul Jump

8.; Have nanoparticles lost their stripes? – David Bradley

Cesbron, Y., Shaw, C., Birchall, J., Free, P., & Lévy, R. (2012). Stripy Nanoparticles Revisited Small DOI: 10.1002/smll.201001465

Yu, M., & Stellacci, F. (2012). Response to “Stripy Nanoparticles Revisited” Small DOI: 10.1002/smll.201202322



  1. Raphaël, may I ask you that you also explicitly cover on this blog the evidence for entropy-driven formation of stripes from coarse-grained and atomistic simulations?

    This essential part of the published research related to striped nanoparticles is clearly missing here.


  2. Pep – I am not qualified to judge the simulations per se. I will make some comments on the underlying assumptions as well as the interpretation of the results. In addition, I would happily publish a guest post on this topic if you wish to write it.


  3. Thanks. I am qualified to judge the simulations, and there is certainly sufficient support for the entropic effect that leads to the stripes. Also, similar entropic-driven phase-separation effects have been seen in other confined bicomponent polymer systems.

    I have not participated in the work related to the simulations of the striped nanoparticles, so it would be more appropriate if a researcher connected to such studies would write a guest post.

    Nevertheless, I feel that you should try to reconcile your criticisms with the simulation work, which can’t be dismissed. In fact, any expert would tell you that they offer strong support for the existence of the stripes.


    1. It is important that external experts, especially those not involved in the work (and therefore independent) participate in the debate.
      I welcome your contribution and I reiterate my invitation to publish an extensive critique (in the most general sense of that word, i.e. the critique could be entirely positive) from you as a base for a more general discussion. The latter would have to be signed with your own name though.

      Let me stress again that I do not question the possibility of phase-separation (entropy driven or otherwise). Here is a quote from the conclusion of our stripy revisited article: “While our analysis suggests that the evidence for stripy nanoparticles is inconclusive, it certainly does not rule out the possibility of self-organization in mixed self-assembled monolayers.” Whether self-organization into stripes, patches, or otherwise occurs is extremely interesting to me and other specialists, but largely irrelevant to the discussion of whether the evidence shown in the stripy papers is convincing or not.

      The simulation article we are discussing is

      As you note, entropic separation is rather common in polymers and surfactant system. The simulation in the above paper is essentially a simulation of a binary mixture of surfactants on a sphere. One difficulty is how representative of the physical chemistry of nanoparticles is this?

      To have self organization, you need dynamics, i.e. you need the possibility for the molecules to move. While diffusion has been observed on flat surfaces at low coverage, in dense SAMs it is not obvious that diffusion can occur on reasonable experimental timescales (and experimental results by the Chechik group suggest they don’t). Dynamics in SAMs through ligand exchange (exchange with excess free molecules in solution) is well documented…. but according to the latest Nature Materials paper (2012), ligand exchange does not lead to stripes although there is strictly no experimental evidence for that claim:


      1. Indeed, I am glad to hear that you do not question the entropy-driven self-organization of nanoparticles coated with self-assembled monolayers into stripes in general, as evidenced by simulation.

        It seems that you only question the experimental evidence for stripes on Stellacci’s nanoparticles. I have now read your Correspondence published in Small, and I am surprised to see that you do not mention at all the evidence for stripes from NMR (, AFM (Fig. 2 in or TEM (Fig. S3 in the original paper; Why did you decide to ignore such evidence?

        I appreciate that when doing STM one has to be careful not to run into the various possible artefacts (the focus of your Commentary and the series of blog posts here), but I can’t understand why you do not discuss at all in your Correspondence or in this blog the evidence form other methods such as NMR, AFM and TEM published in various papers.

        This strikes me as unfair to your readers.


    2. I don’t think anyone has a problem with the possibility of phase separation and stripes or simulation. The problem is that theory has to be backed up by evidence. It was a few years between theory predicting the Higgs and measurement that supported the possibility of the Higgs boson’s existence. At present, I see no evidence for ligands organising into stripes on nanoparticles. That does not mean that these ligands do not form stripes. There is just no evidence.


    3. @pep: Stellacci STM data are artifacts. Read more about it at my guest post. Now, lets look at another SPM — AFM data from “The effect of nanometre-scale structure on interfacial energy”. Figure 2b and 2c show molecular resolution AFM scans. Now these measurements are damn difficult to do, and as far as I know, they have to be performed in non-contact mode (NC-AFM). Figures 2b and 2c are digital zooms from large scans and they show absolutely ***nothing***—the same problem encountered in reported STM imaging as well. I quickly scanned through the paper, but there is no mentioning of the tip and scanning conditions they used to obtain atomic resolution AFM imaging. According to the paper, they employed ordinary atmospheric contact AFM. There is no way of obtaining atomic resolution with the toy they used.

      Again, the same problem and the focus of this blog: conclusions are not supported by veritable, robust and reproducible experimental measurements. From the experimental evidence in Stellacci papers, this is becoming a trend, rather then exception.

      I encourage AFM experienced people to correct me if I am wrong.


      1. A short note to back up Pedja’s comments above. Our group has done a considerable amount of non-contact AFM in ambient, UHV, and cryogenic environments so while I can’t comment authoritatively on the TEM data (which, however, looks rather unconvincing to me), I’m on firmer ground when discussing the AFM results.

        In addition to Pedja’s points, the central issue for me yet again is *reproducibility*. Have Francesco et al. successfully reproduced the same ‘stripe’ patterns in the same nanoparticle in consecutive AFM scans? No, they haven’t (as I know from correspondence with Francesco). This lack of reproducibility is damning because it is the minimum we would expect from AFM novices, let alone supposed experts.

        Li Jinfeng is absolutely correct when he implores, in a comment below, “Where were the referees?” Were the papers sent to SPM experts for review? If not, why not?


    4. Pep, I know that you are not contributing here any more but I hope that you might still be reading Raphael’s blog. I’m responding to this particular comment of yours because it seemed the most appropriate place to post the following relevant (and timely) links:

      Particle Traps, Philip Ball, Nature Materials 12 7 (2013)

      The extremely interesting article to which Ball refers is Simulations: The Dark Side, D. Frenkel (University of Cambridge); > cond-mat > arXiv:1211.4440)

      Best wishes,



  4. Thank you Pep for bringing some fairness to this blog!

    The correspondence was submitted 3 years ago (3 Dec 2009) so I think we can be excused for not discussing the NMR results of the 2012 paper. I have approached a number of international leading experts and I hope that they will be able to provide an independent opinion.

    The TEM is discussed in Stripy Revisited albeit in the supporting information (the TEM was also in the supporting information of Jackson et al). They are also discussed on this blog:

    You are right we could have – just – discussed the AFM in the Nature Materials paper which had been published 3 months before our submission to Small. Here is my take on it, in very few words: the only stripes I can see are those drawn on the figure by the authors, and those on the cartoons which have been pasted onto the figure. There is no quantitative analysis and I suspect (but Philip will maybe contradict me) that the pixel effects he discussed may well also be an issue there.


    1. Thanks for your swift reply and for acknowledging not having yet discussed the simulation, NMR and AFM evidence for stripes. You have now the opportunity to amend this in your blog, and I would also advise you to contact the journal Small and ask them whether you could publish an addendum to your Correspondence where the new evidence is at least briefly mentioned. It is important that you are fair to your readers.

      As for the TEM evidence, in the Supporting Information to your Correspondence you wrote the following: “The published results of the TEM study are however limited to one high-resolution image of one nanoparticle.”
      However, as I mention above in my previous comment, Fig. S3 original article by Stellacci ( also shows a TEM image where I can clearly see the signature of stripes in many nanoparticles.

      As for the AFM results, I can clearly see the stripes on a few nanoparticles in Fig. 2b in (in particular on the phase image).


      1. Pep, thank you for your suggestion and for your efforts in ensuring I am fair to my readers.

        Pep, thank you for pointing my readers to Fig S3. Supporting information is not behind a paywall so everybody can look here:

        Pep sees stripes in Fig S3. I don’t. What others can see? How can we go beyond subjective evidence and do an analysis that would lead to an objective conclusion? We would need for example a control sample to compare to and we don’t have one unfortunately.

        In any case, the authors of the original paper do not use Fig S3 as evidence for stripes. Fig S3 is at a lower mag than S2 and it would be extraordinary if those EM pictures could reveal the organization of the monolayer. Jackson et al actually forgot to reference to fig S3 in the main text, but it is clear that it is there to support this argument:
        “Additionally, by comparing the TEM images of the metallic cores of MPMNs with the STM images of the ligand shells of the same particles, we conclude that the crystallographic nature of the metal is not a determining factor for ripple formation. In fact, 15% of the particles have cores that are twinned, whereas we have never observed ‘twinned’ domains on the particles’ ligand shells”. In fig S3, the arrows point at twinned nanoparticles, not at stripes.

        As I have noted in response to an interesting comment by Elias, there is also the serious question of what would be the origin of the contrast (if there was any); see discussion of the TEM Weinstock group papers here:


  5. Of course, the arrows in Fig. S3 point at twinned nanoparticles, as stated in the caption of the figure. What I meant by ‘the signature of stripes’ is that it appears to me that one could see fragments of rings around the cores of the nanoparticles, as with the high-resolution image, and as better shown in But I am not a TEM expert, and I appreciate your point that TEM images alone are probably not sufficient to distinguish stripes from other geometries such as islands. My point above was only that there is more TEM evidence than just that on one nanoparticle.

    But TEM in addition to STM, AFM, NMR and simulations, as well as the microscopic and macroscopic effects that have been shown by full-atom simulations to result from a structured morphology on the surface of the nanoparticles (see for instance, it is really hard to argue that there is not sufficient evidence to support the claim that the nanoparticles do not have a structured, striped configuration. Your Correspondence in Small concentrates on the STM results, and this may mislead readers to believe that there is no other evidence. I really hope you will amend this.

    And I look forward to read your take about the AFM, NMR and simulation evidence, hopefully in a new post, which will have more visibility than this string of comments. Thanks in advance.


    1. Pep; I look forward to your acceptance of my invitation to write a guest post in your own name. I promise I will stick it in the front of this blog for a full week which will give it a very high visibility.

      In the interest of fairness, it would be useful to have the authors of the stripy papers arguing their points here. Francesco Stellacci is clearly not interested. I should ask other authors. Maybe Molly Stevens will accept. I’ll drop her an email..

      Pep; you write “TEM in addition to STM, AFM, NMR and simulations,” but the STM has been shown to be meaningless, there is no TEM evidence (only your belief that you see something but no control, no quantification, no understanding of the origin of the contrast), and the AFM is simply not convincing (and it adds nothing to the Jackson et al since it is a different SAM).


      1. As for your sentence ‘the AFM is simply not convincing’, you will of course have to elaborate in detail for people to judge your opinion. As for the TEM, as far as I understand, I see controls and quantification in

        And thanks for your offer, but again I am not an expert in imaging techniques, and I have not worked on striped nanoparticles. My main point in this thread is to highlight that you have neglected to discuss (or even mention), in this blog and in your just-published Correspondence, substantive peer-reviewed evidence for the formation of the stripes. I believe that as a scientist it is your duty to fix this serious shortcoming in the posts here in this blog and in your Correspondence.

        Thanks again for allowing me to contribute to your blog.


  6. Pep, I read your posts. I really don’t think that Raphael is expected to comment on any papers describing alternative methods suggesting the existence of the stripes if these papers appear to have been reviewed rigorously (unfortunately not often the case for Stellacci’s papers), and that do not raise any doubts. Raphael has invited you to write a guest post, preferably under your full name, which you could write on that.

    But how much would such a post contribute to the debate? – I think Raphael has made a good effort trying to explain what this discussion is about: no one here states that the stripes do not exist; a point is made that one simply cannot conclude it from the “evidence” that has been published over the past several years in the form of misinterpreted (intentionally?? see Predrag’s post) STM, AFM, TEM, etc. images.

    Having said that, let’s assume the recent NMR paper does indeed point at the existence of the stripies. Good for Stellacci! After all to publish papers with convincing evidence behind the claims is what all we scientists are expected to do (and Stellacci is no exception).

    I appreciate that you participate in this debate as a simulations expert. I am not qualified to judge simulations, but I do have very extensive (years of hands-on experience) experience in surface probe and electron microscopies. Having said that I simply can’t imagine how on earth the referees let this kind of interpretation of STM and TEM images go. 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! The best joke I had heard in a long time! I understand techniques such as TEM and STM might have been new for Stellacci as he started his independent career; he was surely under enormous pressure as an Assistant Professor at MIT; but, for God’s sake, where were the referees??

    I hope you now get my point that I really don’t think that Raphael, in order to be “fair to his readers”, to use your own words, is expected to comment on every single paper Stellacci is publishing; I would rather say that it is a good time for Francesco – to be fair to the community – to confess to having misinterpreted experimental data in the form of corrigenda published in respective (and respected!) journals.


    1. Li, I am afraid I have to disagree.

      Raphaël has written, in this blog (,

      “Stripy Nanoparticles Revisited shows that 23 peer reviewed articles published over the course of 8 years in prestigious journals are based on a simple microscopy artefact.”

      and in his Commentary,

      “This series of articles and the corresponding structure–property relationships are important because of their direct impact on our understanding of several of the key contemporary problems in the field of nanoscience. The latter include the characterization of nanomaterials with sub-nanometer resolution,[3] the possibility of controlling the self-organization of ligands on gold nanoparticles,[4] the understanding of nanoparticle–biomolecule and nanoparticle–cell interactions,[5] and the intracellular delivery of nanoparticles.[6] The proposed stripy structure is based on scanning tunneling microscopy (STM) images which have not yet been reproduced by other groups to date.”

      In both places Raphaël and co-authors mention and cite many papers from Stellacci and co-authors, but then when discussing the evidence they only refer to the STM results, and the TEM results on passing. The rest of the results supporting the appearance of stripes provided in the series of articles by Stellacci that Raphaël refers to and cites is missing in the Commentary and this blog (apart from a few late comments on this page after I pointed out this issue) .

      Clearly, this is unfair to readers, and this is why I decided to write comments on this blog.


      1. Pep; thank you for acknowledging that the TEM results were indeed discussed, not in passing, but with the same level of importance as it was given in Jackson et al. Those results do not support the appearance of stripes.

        Pep, the stripy hypothesis originates in Jackson 2004. It is based on a scanning artefact. The TEM that backs it up is, to quote Li above, a joke. The article also includes a claim that those nanoparticles don’t interact with proteins backed up only with a cartoon. You seem to be a highly principled scientist (if anonymous); are you concerned at all by this state of affairs, or only by the shortcomings of my blog?


      2. Well, well, well. So this is how an Editor at Nature Materials spends his time. Pep – I am stunned.

        I mean, seriously. I’ve been reading your posts and strongly believed / hoped they were by someone outside of the field who happened to be interested in the subject. Perhaps a high school or an undergraduate student. An Editor at Nature Materials? Wow. I mean, WOW. You know, bloggers here spent a LOT of time trying to politely, meticulously explain all the nitty-gritty details that you were coming up with. Shame on you, Pep. Why didn’t you tell upfront who you are and that you are not accepting any expert scientific explanation? Are you really as disillusioned as Stellacci himself?

        If your comments are in a personal capacity only, as you state, then you are seriously embarrassing yourself. If, however, this is your way to protect the good (?) name of Nature Materials, you are really discrediting yourself.

        On another note, let me add that I attended Stellacci’s talks three times during the period this debate covers, including very recently at a major conference. And you know what? I am feeling bad now. I feel like someone has cheated me and everyone else in the audience.


  7. @Pep.

    I agree entirely with Raphael and Pedja, the AFM data are just as unconvincing as the STM results. If I had been sent that paper for review to say I would have been sceptical is an immense understatement.

    Just as Pedja showed in his post how entirely unfunctionalised surfaces containing no nanoparticles can produce stripes in STM images, the AFM data shown in Nature Materials 8 837 (2009) could easily be produced via a variety of noise sources. The point is not whether you (or Francesco) can see stripes or not – as I said in my “Seeing is Believing? Not Always…” post even entirely uncorrelated (i.e. Poisson-distributed) objects can often appear to be correlated to the eye.

    A recurring theme in these papers appears to be the strong influence of observer bias in not only picking out stripes from noise but, remarkably, making sub-Angstrom precision measurements of the ‘periodicity’ of those stripes.

    The least one would expect of a referee would be to ask for evidence that the features are reproduced in consecutive scans.


  8. 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


  9. @ferniglab:

    “At present, I see no evidence for ligands organising into stripes on nanoparticles. That does not mean that these ligands do not form stripes. There is just no evidence.”

    This is an opinion. I am sure you will agree that for this opinion to turn into supported criticism you will have to elaborate in detail.


    I read your post, and I also read Stellacci and co-authors’ Response in Small ( Could you also discuss this response, and in particular Figs. 3 and 4, in your entry in this blog? I would appreciate your opinion on it. Thanks.


    As for your central issue on reproducibility, in Stellacci’s Response I read: “Here in Figure 3 and Figure 4, we also show that striped nanoparticle images qualitatively and quantitatively comparable to the ones we have presented through the years (imaged in air with a Veeco Multimode IIIa microscope) can also be obtained in vacuum on an Omicron Vakuumphysik ‘micro-STM’, exhibiting a stripe width consistent with similar particles presented four years ago.[16] Figure 3 and Figure 4 also show that the stripes of octanethiol:methylbenzylthiol 2:1 nanoparticles are invariant with tip speed and rotate with scan direction.”


    As a non-specialist in imaging, I would appreciate it if instead on keeping your comments focused on a small set of images and/or techniques, you all would consider the whole evidence provided over the years, and in particular the Response by Stellacci ( published alongside Raphaël’s Commentary in Small. I am sure you all will agree that this is the right thing to do.


    1. @pep.

      All of the STM data I have seen thus far has been entirely irreproducible – the same stripe pattern has not been seen in the same nanoparticle in two consecutive scans. This is true not only for a small set of images as you put it – it’s true for the entirety of the STM data.


    2. @pep:
      I definitely agree that some surface features in Figures 8 and 9 are oriented at ~45 degrees. Now, take a look at the left image
      This is an image of unfunctionalized gold foil. Can you see ripples at an angle on this picture? I can certainly identify them for you, but I will allow you to look carefully and find them.
      So the bottom line is that these features are feedback oscillations. When you superimpose feedback oscillations on a shape ( imagine elongated semicylinder on the surface), while scanning, the beats ( dots in the image) will be shifted, scan line to scan line — and that gives an impression of an angle. How do I know that these are feedback oscillations? I have seen them, worked with them, so did every STM/AFM experienced scientist. Fortunately, we don’t have to speculate that these are feedback oscillation — there is a proof of that. If you request data from Figures 8 and 9 from Stellacci et al. and look at the tunneling current reading, you will see that the tunneling current over the “ripples” peaks to 10-100 times the set value and begins to oscillate. Again,and again, and again, this is very common in STM/AFM and it is not an esoteric rare feature. The fact that Stellacci et al. show the topography without the tunneling current, which is supposed to be constant, is dishonest and misleading. Every AFM/STM expert will confirm this. I hope I clarified enough. If you would like to see other images showing the same phenomenon, I have plenty of them on my computer.


  10. @pep. No. Sorry, but no. The data in Yu and Stellacci’s response are utterly unconvincing. See my previous post on this and my comment here .

    Oh, and here aswell.

    Those ‘data’ are nothing but cherry-picked collections of noisy pixels.

    The key aspect of this debate is that even if high resolution NC-AFM in UHV by other groups (including our own here in Nottingham) subsequently detects stripes on nanoparticle surfaces, this does not in any way vindicate the approach of the Stellacci group to acquiring and ‘analysing’ STM data.


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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s