Figure 1.
Reanalysis of the data for Figure 1a) of Jackson et al. 2004 [23].
(a) The raw 157nm wide image collected for Jackson et al. (b) Zoom-in on a 37 nm wide area marked in green on(a). This image has been flattened using first order plane subtraction. It is clear that the ripples extend between the particles. (c) Figure 1a) from Jackson et al.(2004) Spontaneous assembly of subnanometre-ordered domains in the ligand shell of monolayer-protected nanoparticles. Nature materials 3: 330–6. (DOI:10.1038/nmat1116) Reproduced by permission of Nature Publishing Group. Note that the choice of contrast obscures the ripples between the particles. Scale bar 10nm. (d) Fourier transform high-pass filter of (b), removing spatial frequencies below 0.33 × 109 m−1. (e) Simultaneous current image of (b). Note the (inverted) similarity to (d). The colour ranges for (a), (b) and (d) are set to run linearly from the highest to the lowest pixel. For (e) the colour range is set to run linearly for the centre 99.6% of pixels, as extreme pixels mask much of the contrast (For this section of the image the tunnel current spans a range from -51.2 nA to 2.83 nA). Colour bar shows recorded current values, the setpoint current is +838 pA.
Figure 2.
Comparison of STM image of nanoparticle “stripes” with simulated STM feedback results.
(a) Image from [48] Uzun et al. (2008) Water-soluble amphiphilic gold nanoparticles with structured ligand shells. Chemical communications (Cambridge, England): 196–8. (DOI: 10.1039/B713143G [48]), reproduced by permission of The Royal Society of Chemistry. The image shows features which can be reproduced by simulated SPM feedback (scale bar 5 nm). (b) Surface topography used in all numerical simulations. (c) Numerically simulated image with appropriate parameters Kp = 500 and Ki = 100. (d–h)The same simulation with Kp = 50 and Ki = 8000, 5000, 3000, 2000, and 1000 respectively. Image (i) is the retrace image recorded while recording image (f) presented directly above.
Figure 3.
Imaging of unfunctionalised Ag nanoparticles with varying scan parameters.
Top two rows: The top left image was recorded with a gain of 5%. For each consecutive image (i.e. moving along the rows from left to right), the gain was incremented by 1%. Each image is 8 nm wide, and all were recorded with a tip speed of 38 nm/s. Bottom two rows: Trace (third row) and retrace (bottom row) image of Ag nanoparticles, upwards scan direction. At the point marked by an arrow in both images, the scan speed was reduced from 514 nm/s to 195 nm/s, causing a significant reduction in stripe width (indicated with red arrows; these arrows also indicate scan direction). Soon after, the gain was reduced from 22% to 10% and the stripes disappear (gradually decreased in the lines marked by the green double-headed arrow). Both images have a width of 50 nm.
Figure 4.
Reanalysis of the data for Figure 3 of Jackson et al. 2006 [24].
The black squares represent the peak frequency in the Fourier spectrum of the tunnel current images, while the grey area represents the full width at half-maximum (FWHM) of the peak in Fourier space. Red circles are digitised data from the noise spacings presented in Figure 3b of Jackson et al. 2006 [24]. Green diamonds and blue triangles are digitised data from the ripple spacings presented in Figure 3(b) of Jackson et al. 2006. All ripple spacings fall inside the spatial frequency band of the error signal. The first and last point represent images archived by Stellacci et al. along with the data for Figure 3 of Jackson et al. 2006 [24], but which were not analysed in Jackson et al. 2006. The full method and code used to generate this figure are given in the Supplementary Information.
Figure 5.
Arithmetic addition of images from Yu and Stellacci [42].
(a)-(j) Images of the same set of nanoparticles taken from each of the five trace and five retrace images provided by Yu and Stellacci. (a,c,e,g,i) are the trace images, while (b,d,f,h,j), respectively, are the corresponding retrace images. (k) Arithmetic addition of all 10 images. Note that the particles in the summed image appear entirely smooth, indicating that the features designated as stripes by Yu and Stellacci arise from noise and not real topographic structure on the nanoparticles. All images are 20 nm wide.
Figure 6.
Reanalysis of data from Yu and Stellacci [42].
Panels a, c and f reproduce images from Yu and Stellacci (2012) Response to stripy nanoparticles revisited. Small 8: 3720–3726 (DOI: 10.1002/smll.201202322) - reproduced by permission of John Wiley & Sons. (a) Image as presented in Yu and Stellacci (b) A 205 × 205 pixel section of the raw data which has been processed with second order background subtraction, the colour range reduced to just 40% of the original range, and the number of pixels interpolated to best match the image shown in (a); (c) Enlargement of region highlighted by a blue square in (a); (d) Zoom of a section of the image shown in (b) taken after interpolation and colour saturation; (e) Retrace image acquired simultaneously with (d); (f) Image shown in (c) but with the stripes identified by Yu and Stellacci highlighted using dashed lines; (g) Uninterpolated zoom of the raw data showing the true pixelation. (h) Retrace image acquired simultaneously with (g). The “stripes” in (f) not only arise from a very small number of fortuitously aligned pixels, but they are not present in the retrace images shown in (e) and (h).
Figure 7.
Representative data from Ong et al. [40].
Reproduced from Ong et al. (2013) High-resolution scanning tunneling microscopy characterization of mixed monolayer protected gold nanoparticles. ACS nano 7: 8529–39 (DOI: 10.1021/nn402414b) - reproduced by permission of the American Chemical Society (a) High-resolution STM image of an Au nanoparticle with a coating of 11-mercapto-1-undecanol and 4-mercapto-1-butanol, taken in UHV conditions at 77K. (b) High-resolution STM image of an Au nanoparticle with a coating of OT:MPA used in the original striped morphology paper (Jackson et al. 2004 [23]). (c) High-resolution STM image of homoligand nanoparticle with an OT coating. (a) and (b) allegedly show stripe-like domains while (c) does not. (d–f) Radially averaged PSDs from STM images of the same type of particles shown in (a)–(c) respectively.
Figure 8.
The persistence of tip induced features on bare Ag nanoparticles.
Four successive images (a–d) with black arrows showing the direction of the slow scan. Tip change events, marked by a red arrow, change the apparent sub-particle structure of the bare nanoparticles. Note the persistence of the artefacts throughout the images. The tip state shown in (d) was persistent over many consecutive scans. The green circle identifies the same particle in subsequent images and (e–h) show offline (and interpolated) zooms of this particle from each of the images (a–d). Blue circles mark the same features in all images as a reference point to show the scan area is consistent. All scale bars in (a–d) are 30 nm. Minor contrast adjustment has been applied to images (a,b,d,e,f).
Figure 9.
Digitised position of ligand head groups and stripes identified by Ong et al. [40] as compared to eight sets of randomly distributed ‘head groups’.
The top panel is reproduced from Ong et al. (2013) High-resolution scanning tunneling microscopy characterization of mixed monolayer protected gold nanoparticles. ACS nano 7: 8529–39 (DOI: 10.1021/nn402414b) - by permission of the American Chemical Society. The top row shows the image in question from Ong et al. (upper right corner) along with a version of that image where we have superimposed a semi-transparent square and highlighted the ‘stripes’ identified by Ong et al. using green dashed lines. The original blue circles (right) are visible through the digitised head groups. The positions of the head-group features within that square, and the corresponding dashed lines highlighting the ‘stripes’, are then reproduced on a featureless background, as indicated by the red double-headed arrow. The other eight images in the figure for comparison show randomly distributed features. By either assigning straight lines (green), curved lines (blue), or stripe-like domains (purple) it is possible to guide the reader’s eye to clustering in random features.
Figure 10.
1D PSDs of simulated nanoparticle substrates.
(a) 1D PSD for nanoparticles for simulated stripes of increasing amplitude (see simulated images shown in c–f); (b) Equivalent to (a) but in this case for simulated nanoparticles covered in randomly positioned speckles (ligand head groups (see images (h–j)). The speckled images simulating a random distribution of head-groups yield the plateau and shoulder observed by Biscarini et al. [41] which were inadvertently assumed to represent the signature of a striped morphology.; (c) Simulated flat surface with 10nm diameter spherical nanoparticles. (d–f) 1nm wide sinusoidal stripes are added to the surface of the nanoparticles (thus, they reduce in width at the edge) with peak-peak amplitudes of 2, 4 and 6 nm respectively; (g) Randomly distributed ‘speckled’ pattern of features 0.8 nm in diameter. (h–j) Images of simulated nanoparticles where the speckles in (g) have been added to (c) with heights of 0.8, 1.6 and 2.4 nm respectively. (c–f and h–j) have had identical white noise added for consistency and for a fair, unbiased comparison.
Figure 11.
TEM data of OT:MPA-coated nanoparticles from Figure S2 of Jackson et al. [23] - Jackson et al. (2004) Spontaneous assembly of subnanometre-ordered domains in the ligand shell of monolayer-protected nanoparticles.
Nature materials 3: 330–6. (DOI:10.1038/nmat1116) Reproduced by permission of Nature Publishing Group. (a) Red arrows indicate dark features surrounding the nanoparticle which have been interpreted as MPA head groups. Such features commonly arise from TEM defocus, and even if real are not arranged in striped domains. (b) Dark-field TEM image with inset power spectrum. (c) Further TEM image of mixed-ligand nanoparticles with red arrows supposedly indicating sinusoidal features. Neither (b) nor (c) show any evidence for an ordered striped morphology.