Metamaterial assisted illumination nanoscopy via random super-resolution speckles

Structured illumination microscopy (SIM) is one of the most powerful and versatile optical super-resolution techniques. Compared with other super-resolution methods, SIM has shown its unique advantages in wide-field imaging with high temporal resolution and low photon damage. However, traditional SIM only has about 2 times spatial resolution improvement compared to the diffraction limit. In this work, we propose and experimentally demonstrate an easily-implemented, low-cost method to extend the resolution of SIM, named speckle metamaterial-assisted illumination nanoscopy (speckle-MAIN). A metamaterial structure is introduced to generate speckle-like sub-diffraction-limit illumination patterns in the near field with improved spatial frequency. Such patterns, similar to traditional SIM, are then used to excite objects on top of the surface. We demonstrate that speckle-MAIN can bring the resolution down to 40 nm and beyond. Speckle-MAIN represents a new route for super-resolution, which may lead to important applications in bio-imaging and surface characterization.

The key advantage of SIM, is the "structured illumination", and by knowing the illumination periodicity, an important prior knowledge in SIM, it enables SIM with only 9/15 images for 2D/3D SIM image. If 100 images are needed, intensity fluctuations methods would also perform, which will generate results similar as shown in this manusript (e.g. previous literature, on MUSICAL where 50 nm resolution has already being documented) Agarwal, K., Macháň, R. Multiple signal classification algorithm for super-resolution fluorescence microscopy. Nat Commun 7, 13752 (2016). https://doi.org/10.1038/ncomms13752 On using the speckle pattern for illumination, there is no prior information and thus loss of temporal resolution becomes evident. Author stated about pre-determined illuminations using speckle idea but no scientific information was provided, so it is difficult to foresee how this would work.
The study carried out in this manuscript is systematic and is more suitable for publication in Scientific Reports.

Reviewer #3 (Remarks to the Author):
From my point of view the authors have addressed most of my initial concerns and the paper is now publishable in Nature Communications.
Before publication, I highly recommend the following changes to be made to the manuscript, though: 1) Fig. S5 should be moved to the main text -and possibly be integrated with another, already existing figure in the main text (my suggestion would be to combine it with Fig 4 in the main text). The main reason for recommending this change is that the manuscript will greatly benefit from a true proof-of-concept, which is a comparison of the reconstruction of a sub-wavelength structure with an electron micrograph of the same structure -and this is what Fig. S5 provides. It will get lost in the supplement, though....
2) The references in the supplemental information file need to be changed. Reference 2 is not an appropriate reference for SOFI. Here, the original SOFI paper by Enderlein should be cited:

Color codes used in this response letter:
Blue Italic: original review comments; Black: our responses; Red: revisions made in the manuscript.

Response for Reviewer #1
The authors have significantly revised this manuscript for resubmission to Nature communications compared to the previous version. They have added significant new content, especially regarding details about the depth penetration and the applicability of the technique to different experimental aims. They have also made significantly more effort to properly char acterise the resolution they do achieve, including with different reconstruction approaches.
Overall the paper has been significantly improved and they have clearly answered all of the points that I originally raised. I feel that the paper is now of a sufficient standard to be published in Nature Communications.
[Reply] We appreciate the reviewer's positive evaluation on our study and valuable comments.

Response for Reviewer #2 [Question 2-1] I have reviewed this manuscript twice now (Reviewer 2), and although the revised paper has additional details, on the conceptual side, there is limited novelty to grant publication in Nature Communication. The main idea of the manuscript of generating superresolved speckle patterns and to exploit them for super-resolution microscopy is well established in the nanoscopy community.
The author in their response letter wrote "the previously established speckle illumination super-resolution methods are all limited by diffraction-limited speckles" which is incorrect.

[Answer 2-1] Discussion about novelty and super-resolved speckle patterns
We thank for the reviewer's time and efforts that have gone into the careful examination of our study. In order to make a better flow to describe our work and novelty, we check the reference that reviewer #2 mentioned. By doing so, we clarify that the previously reported work is nothing related to the super-resolved speckle patterns. The established super-resolution methods are based on either fluctuation-based microscopy, ESI, or localization-based microscopy, dSTORM. The Supplementary Figure S11 of the paper shows not the presence of super-resolved speckle but a dSTORM image of multi-mode interference pattern of a strip waveguide which visually looks like speckle due to the sparse nature of STORM reconstruction. The interference fringe size of ~140 nm corresponds to the diffraction-limit of λ/2neff within the waveguide. This interference pattern was not used as "a speckle illumination pattern" in their imaging methodology. Moreover, we would like to emphasize that, a most recently published paper from the same group ("Structured illumination microscopy using a photonic chip" (Helle, Ø. I. et al. Nat. Photon. (2020)) reports 2.3 times enhancement (~209 nm) in imaging spatial resolution by using a photonic waveguide structure to deliver the illumination patterns. The photonic waveguide can generate higher resolution illumination patterns compared to patterns generated in free space purely because of the high refractive index of the "dielectric materials" composing the photonic waveguide (refractive index of Ta2O5 or Si3N4). Therefore, it is necessary to introduce plasmonic materials/metamaterials for better resolution improvement. To the best of our knowledge, our study is the first experimental demonstration of super-resolution imaging with highk (~7k0, where k0 is the wavevector of incident beam) speckle illumination generated by metamaterials.
[Question 2-2] Moreover, authors commented that previous work requires specific fluorophore dyes which is also in-correct. For intensity fluctuation method to work, either intrinsic photokinetics of the fluorophores are needed or the fluctuations (modulation) of the illumination light itself is sufficient. Both the cases are well known and are being exploited by the community. In this manuscript, the authors have exploited the fluctuations caused by the illumination, itself, similar to several previous literatures, some of them copied below and many others exits.  260-274 (2018).

[Answer 2-2] Discussion about novelty
Based on the discussion at [Question 2-1], the three references that reviewer #2 mentioned were all discussed in the previous response letter for the reviewers' comments. We would like to clarify again that the previously reported illumination-based methods are using diffraction-limited speckles.

Ref.
Algorithm On using the speckle pattern for illumination, there is no prior information and thus loss of temporal resolution becomes evident. Author stated about pre-determined illuminations using speckle idea but no scientific information was provided, so it is difficult to foresee how this would work. The study carried out in this manuscript is systematic and is more suitable for publication in Scientific Reports.

[Answer 2-3] Discussion about imaging speed and pre-determined illuminations
We would like to remind the reviewer that the number of images needed for the SIM reconstruction is highly related with the resolution improvement factor compared to the diffraction limit. We use more images in our speckle-MAIN mainly because our resolution is much higher than conventional SIM. The periodicity in traditional SIM is an important prior knowledge but pre-determined speckles represent much more prior information enabling improved image reconstruction with less measurements via compressive sensing. This eventually will lead to higher imaging speed.
We agree that our current demonstration of speckle-MAIN is only focused on high lateral resolution achievement using randomly varying high-k speckle illumination. We believe our following study will provide a complete demonstration for high-speed super-resolution imaging. Since the detailed experiment is by itself a mammoth task, it will be undertaken by us in future studies. However, the idea about how to use pre-determined speckle illumination for high-speed imaging has been illustrated in Nanoscale, 9, 18268-18274 (2017). In fact, this is the major distinction between our speckle-MAIN and MUSICAL. The pre-determined speckle illumination makes the compressive sensing scheme possible which will easily improve the imaging speed by 10 folds and the improvement factor can be much greater if the object is sparse. Intrinsic intensity fluctuation cannot be pre-determined, so that no prior information can be utilized in MUSICAL and there is no way to apply the compressive sensing scheme, limiting their potential speed. Please see more detailed discussions in the above mention reference reported by our group.

Response for Reviewer #3 [Initial statement] From my point of view the authors have addressed most of my initial concerns and the paper is now publishable in Nature
Communications. Before publication, I highly recommend the following changes to be made to the manuscript, though: [Reply] We appreciate the reviewer's positive evaluation on our study and valuable comments. The detailed responses are below for each of the comments.
[Question 3-1] Fig. S5 should be moved to the main text -and possibly be integrated with another, already existing figure in the main text (my suggestion would be to combine it with Fig  4 in the main text). The main reason for recommending this change is that the manuscript will greatly benefit from a true proof-of-concept, which is a comparison of the reconstruction of a sub-wavelength structure with an electron micrograph of the same structure -and this is what Fig. S5 provides. It will get lost in the supplement, though....

[Answer 3-1]
We appreciate the reviewer's valuable suggestions. Following the reviewer #3's excellent suggestion, we made modifications in Fig. 4 in the main text. [Revision in main text] After reduction to only 80 sub-frames, we still can resolve two particles with ~60 nm center-to-center distance by using an objective with NA = 0.8 (Fig. 4 ai), indicating the robustness of the speckle-MAIN technology. Figure 4 j-m represents a super-resolution image of a more complex object at a wide field of view. The object is made by a dense drop-casting of quantum dots emitting at 605 nm.

[Question 3-2]
The references in the supplemental information file need to be changed. [Answer 3-2] We thank the reviewer for the correction. As the reviewer correctly pointed out, we have made revisions to the references in the revised Manuscript.
We thank again for all the reviewers' time and efforts that have gone into the careful examination of our manuscript and for giving valuable and insightful comments. These comments have certainly helped us improve our manuscript in the revised form. We hope the revisions we made in the manuscript and our responses have addressed the review comments, and the manuscript in its revised form is considered suitable for publication in Nature Communications.