Toward giga-pixel nanoscopy on a chip: a computational wide-field look at the nano-scale without the use of lenses - PubMed (original) (raw)
Review
Toward giga-pixel nanoscopy on a chip: a computational wide-field look at the nano-scale without the use of lenses
Euan McLeod et al. Lab Chip. 2013.
Abstract
The development of lensfree on-chip microscopy in the past decade has opened up various new possibilities for biomedical imaging across ultra-large fields of view using compact, portable, and cost-effective devices. However, until recently, its ability to resolve fine features and detect ultra-small particles has not rivalled the capabilities of the more expensive and bulky laboratory-grade optical microscopes. In this Frontier Review, we highlight the developments over the last two years that have enabled computational lensfree holographic on-chip microscopy to compete with and, in some cases, surpass conventional bright-field microscopy in its ability to image nano-scale objects across large fields of view, yielding giga-pixel phase and amplitude images. Lensfree microscopy has now achieved a numerical aperture as high as 0.92, with a spatial resolution as small as 225 nm across a large field of view e.g., >20 mm(2). Furthermore, the combination of lensfree microscopy with self-assembled nanolenses, forming nano-catenoid minimal surfaces around individual nanoparticles has boosted the image contrast to levels high enough to permit bright-field imaging of individual particles smaller than 100 nm. These capabilities support a number of new applications, including, for example, the detection and sizing of individual virus particles using field-portable computational on-chip microscopes.
Figures
Fig. 1
Lensfree on-chip microscopy. (a) A schematic diagram of the experimental setup. A partially coherent light source (e.g., a light-emitting-diode) positioned at distance _z_1 above the sample plane allows for holographic imaging, while lateral shifts of the source enable pixel-super-resolution. The small sample-to-sensor distance (_z_2) provides unit magnification and an ultra-large FOV, _W_X × _W_Y that is equal to the active area of the opto-electronic sensor-chip. (b) Computational procedures for obtaining high-resolution on-chip images. Figure adapted from ref. .
Fig. 2
State-of-the-art resolution of lensfree on-chip microscopy. (a) The highest NA achieved to date for a lensfree on-chip microscope, showing the resolution of a grating with 300 nm half-pitch, corresponding to a NA of 0.92 (wavelength = 550 nm). (b) The smallest resolved feature size, corresponding to a grating with 225 nm half-pitch. The equivalent NA is lower than in (a) because of the reduced wavelength, 372 nm. (c) Lensfree nanoscopy applied to imaging single helical multi-walled carbon nanotubes. The shape of the nanotube is visible, with its width (500 nm) of the order of twice the half-pitch resolution (225 nm). All lensfree results use immersion oil between the sample and the sensor to improve spatial resolution. The SEM comparison image of the nanotube contains a metal coating, which was deposited after its lensfree imaging has been performed.
Fig. 3
Self-assembled nanolenses enable nanoparticle detection. (a–c) To-scale schematic representations of the catenoid nanolens shape surrounding a 200 nm particle for various substrate (_θ_s) and particle (_θ_p) contact angles. In (a), the equation gives the radial extent of the nanolens as a function of its height, where a and b are functions of the particle size, _θ_s, and _θ_p. (d) A SEM image of a large (1 μm) spherical particle with the remnants of a nanolens that was desiccated during SEM sample preparation, illustrating the extent of the nanolens. (e) A comparison SEM image of a polystyrene bead without a nanolens. Images adapted from ref. .
Fig. 4
Experimental detection of individual nanoparticles and viruses. (a) Lensfree pixel-super resolution imaging results of 95 nm polystyrene beads with and without self-assembled catenoid nano-lenses. 100× (NA = 1.25) oil-immersion objective images of the same samples are provided for comparison purposes and intensity cross-section curves of individual particles are shown in their insets due to the low contrast. (b) The reconstruction result (top-middle image) of a region-of-interest from a heterogeneous nano-bead sample. Corresponding SEM images (s1–s4) are in good agreement with the lensfree reconstruction. Red and blue arrows locate the ≤ 100 nm beads and the beads having diameters in the 100–150 nm range, respectively. (c) Lensfree holographic pixel-super resolution imaging of single adenoviruses and influenza A (H1N1) viruses, with corresponding SEM and 100× (NA = 1.25) oil-immersion objective images for verification. Red arrows are used to identify the particles in lensfree amplitude and phase reconstructions, as well as SEM images. Intensity cross-section curves of single viruses within the 100× objective image are shown in the inset due to the low contrast. Note also that the lensfree images are digitally cropped from a much larger FOV (i.e., 20.5 mm2 – the active imaging area of the CMOS sensor used in this work). Images adapted from ref. .
Fig. 5
Ultra-wide-field CCD-based lensfree imaging results of sub-150 nm particles. The FOV of this CCD sensor-chip (>18 cm2) is 90 times larger than the CMOS sensor used to generate the lensfree imaging results in Fig. 4. Note that only half of the CCD active area (>9 cm2) is shown in A. B was digitally cropped from A, which was also cropped from a much larger FOV (37 mm × 25 mm) on the left (where black spots were used for the registration of the FOVs of the comparison images). A high contrast and background-subtracted 60× objective lens-based image of the corresponding region-of-interest, as well as two SEM images (s1 and s2) of subregions are demonstrated for comparison purposes. Images adapted from ref. .
Fig. 6
Simulated impact of nanolens improvements on lensfree holographic microscopy. (a) In both amplitude and phase reconstructions, highly absorbing nanolenses around 50 nm particles enhance their contrast with respect to the background. (b) Reducing the contact angle with the substrate makes the nanolenses larger and more effective in detecting 50 nm particles. Insets show amplitude reconstructions for three points in the vicinity of the detection threshold. (c) Nanolenses composed of highly refractive fluids scatter light more effectively, enabling detection of 75 nm particles. All of the amplitude reconstructions without nanolenses in (c) show SNR < 2 dB. For all data in a given subfigure, an identical randomly-generated 1% Gaussian noise field is added at the detector plane. Unless a parameter is being varied explicitly, all simulations assume a purely real film refractive index of 1.35 and a substrate contact angle of 10°. Images adapted from ref. .
References
- Fernández-Suárez M, Ting AY. Nat Rev Mol Cell Biol. 2008;9:929–943. -PubMed
- Goldstein M, Naulleau P. Opt Express. 2012;20:15752–15768. -PubMed
- Cotte Y, Toy F, Jourdain P, Pavillon N, Boss D, Magistretti P, Marquet P, Depeursinge C. Nat Photonics. 2013;7:113–117.
- Midgley PA, Weyland M. Ultramicroscopy. 2003;96:413–431. -PubMed
- Hell SW. Science. 2007;316:1153–1158. -PubMed
Publication types
MeSH terms
LinkOut - more resources
Full Text Sources
Other Literature Sources
Miscellaneous