Scanning superlens microscopy for non-invasive large field-of-view visible light nanoscale imaging

Feifei Wang, Lianqing Liu, Haibo Yu, Yangdong Wen, Yu Peng, Zhu Liu, Yuechao Wang, Wen J. LiNature Communications 2016 · DOI

TL;DR

This paper introduces a novel "scanning superlens microscopy" (SSM) technique that uses microspheres to achieve nanoscale resolution over large fields of view with visible light. This non-invasive method significantly speeds up imaging compared to traditional AFM, opening new avenues for studying live cellular processes at the sub-diffraction limit.

What Problem Does This Paper Address?

Understanding complex biological processes often requires correlating structural information with the identification of specific molecules. While electron microscopy (EM) offers high resolution, its destructive nature limits its use on live samples. Super-resolution fluorescence microscopy can identify molecules but often sacrifices structural detail. Traditional optical microscopy is non-invasive and fast but limited by the diffraction limit. Existing super-resolution techniques, like scanning probe microscopy (SPM) and scanning electron microscopy (SEM), are often slow and can be invasive, making them unsuitable for observing dynamic cellular events over extended periods. The authors aim to bridge this gap by providing a non-invasive, high-speed, large-field-of-view, sub-diffraction-limited imaging technique.

Methods

The core innovation is Scanning Superlens Microscopy (SSM). It integrates a microsphere superlens with an Atomic Force Microscope (AFM) scanning system. Instead of a sharp AFM tip, a microsphere is attached to the AFM cantilever. This microsphere acts as a superlens, capable of focusing light beyond the diffraction limit.

Key aspects of the method include:

  • Microsphere Superlens: High-refractive-index microspheres (e.g., BaTiO3) are used to convert evanescent waves into propagating waves, enabling sub-diffraction-limited imaging.
  • AFM Integration: An AFM system provides precise control over the distance between the microsphere and the sample. This allows for two modes:
  • Contact Mode: The microsphere is in direct contact with the sample. This is suitable for imaging stiff, non-biological objects.
  • Constant-Height Mode (Non-invasive): The microsphere is held at a constant, small distance above the sample. This is crucial for imaging sensitive biological specimens like live cells without causing damage.
  • Large-Area Imaging: SSM achieves large field-of-view imaging by scanning the microsphere across the sample and then stitching together multiple recorded images. This "area-based" scanning, rather than point-based scanning, significantly improves efficiency.
  • Illumination Control: The paper explores different illumination strategies, including coaxial and partial/inclined illumination, to optimize imaging quality and the working distance (Az) between the microsphere and the sample.

Datasets

The paper utilizes several types of samples for demonstration and validation:

  • Non-biological:
    • Blu-ray disc surface (features ~100 nm)
    • Central Processing Unit (CPU) structures (~80 nm)
    • Silver nanowires (AgNWs) (~40 nm diameter)
  • Biological:
    • Mouse myoblast cells (C2C12)
    • Human breast cancer cells (MCF-7)

These datasets are used to showcase the resolution capabilities, non-invasiveness, and large-area imaging potential of SSM across different sample types. No new benchmark datasets were introduced.

Key Insights and Ideas

  • Bridging the Resolution-Functionality Gap: SSM offers a compelling solution for simultaneously acquiring high-resolution structural information and enabling non-invasive observation of live biological processes, a long-standing challenge.
  • Efficiency Through "Area" Scanning: By replacing the AFM tip's single point with the microsphere's imaging area, SSM dramatically increases imaging speed, achieving up to a ~200x improvement over traditional AFM.
  • Non-Invasive Imaging of Sub-Membrane Structures: The constant-height mode is a breakthrough for observing delicate cellular components without disruption, paving the way for studying intracellular events in real-time.
  • Correlative Imaging Potential: The ability to perform both white-light and fluorescent imaging with SSM opens doors for detailed correlative studies, allowing researchers to map molecular distributions onto precise structural landscapes.
  • Illumination Optimization: The study highlights how different illumination schemes (coaxial vs. partial/inclined) impact imaging performance and the achievable working distance, offering practical guidance for experimental setup.

Key Results

  • Sub-diffraction-limited Resolution: SSM achieved resolutions better than λ/3.8, with specific examples showing resolutions of ~25 nm and ~14 nm pixel size.
  • Imaging Efficiency: The SSM system demonstrated an acquisition efficiency improvement of approximately ~200 times compared to traditional AFM for large-area imaging. For example, imaging a 96x96 µm² area took ~3 minutes with SSM versus ~9.5 hours with AFM.
  • Large Field-of-View: The technique successfully imaged areas up to ~96x96 µm² by stitching together multiple scans.
  • Non-invasive Observation: Live cell imaging was demonstrated, showing improved surface morphology and sub-membrane structures without apparent damage.
  • Sub-surface Imaging: SSM could image structures beneath optically transparent films that blocked SEM and traditional optical microscopy.
  • Filament Identification: Correlative white-light and fluorescent imaging allowed for the identification of actin filaments within cells.

Limitations and Future Work

  • Image Stitching Artifacts: While effective, image stitching can introduce artifacts, and the authors mention that further improvements can be made using band-pass or recursive bilateral filtering algorithms.
  • Working Distance (Az) Sensitivity: The imaging quality and resolution are sensitive to the distance between the microsphere and the sample (Az). While optimized, maintaining this precise distance over large areas can be challenging. The paper notes that Az should be controlled to be less than half of the maximum achievable distance (Azmax/2) for high-quality imaging.
  • Illumination Dependence: The performance varies with illumination conditions, and further optimization of partial and inclined illumination is suggested.
  • Future Work: The authors suggest that further improvements in imaging efficiency can be achieved through more advanced algorithms for image stitching, higher-speed cameras, or parallel scanning techniques (e.g., parallel AFM or superlens arrays). The potential for simultaneous tracking of specific molecules with nanoscale resolution over complete cellular lifecycles is highlighted as a major future direction.

Summary updated

Abstract

Nanoscale correlation of structural information acquisition with specific-molecule identification provides new insight for studying rare subcellular events. To achieve this correlation, scanning electron microscopy has been combined with super-resolution fluorescent microscopy, despite its destructivity when acquiring biological structure information. Here we propose time-efficient non-invasive microsphere-based scanning superlens microscopy that enables the large-area observation of live-cell morphology or sub-membrane structures with sub-diffraction-limited resolution and is demonstrated by observing biological and non-biological objects. This microscopy operates in both non-invasive and contact modes with ∼200 times the acquisition efficiency of atomic force microscopy, which is achieved by replacing the point of an atomic force microscope tip with an imaging area of microspheres and stitching the areas recorded during scanning, enabling sub-diffraction-limited resolution. Our method marks a possible path to non-invasive cell imaging and simultaneous tracking of specific molecules with nanoscale resolution, facilitating the study of subcellular events over a total cell period.
BibTeX
@inproceedings{wang2016scanning,
  author = {Feifei Wang and Lianqing Liu and Haibo Yu and Yangdong Wen and Yu Peng and Zhu Liu and Yuechao Wang and Wen J. Li},
  title = {{Scanning superlens microscopy for non-invasive large field-of-view visible light nanoscale imaging}},
  booktitle = {Nature Communications},
  year = {2016},
  doi = {10.1038/ncomms13748},
}
Figure