Achieving High Spatial Resolution Surface Plasmon Resonance

Subscriber access provided by Fudan University

Technical Note

Achieving high spatial resolution surface plasmon resonance microscopy with image reconstruction Hui Yu, Xiaonan Shan, Shaopeng Wang, and Nongjian Tao Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b05049 • Publication Date (Web): 08 Feb 2017 Downloaded from http://pubs.acs.org on February 8, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Achieving high spatial resolution surface plasmon resonance microscopy with image reconstruction Hui Yu1,2, Xiaonan Shan2, Shaopeng Wang2, and Nongjian Tao1,2* 1 State Key Laboratory of Analytical Chemistry for Life Science, School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210093, China 2 Center for Bioelectronics and Biosensors, Biodesign Institute, Arizona State University, Tempe, AZ 85287, USA *Corresponding authors: NJ Tao: [email protected]

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 15

ABSTRACT

Surface plasmon resonance microscopy (SPRM) is a powerful platform for biomedical imaging and molecular binding kinetics analysis. However, the spatial resolution of SPRM along the plasmon propagation direction (longitudinal) is determined by the decaying length of the plasmonic wave, which can be as large as tens of microns. Different methods have been proposed to improve the spatial resolution but each at the expense of decreased sensitivity or temporal resolution. Here we present a method to achieve high spatial resolution SPRM based on deconvolution of complex field. The method does not require additional optical setup, and improves the spatial resolution in the longitudinal direction. We applied the method to image nanoparticles, and achieved close-to-diffraction limit resolution in both longitudinal and transverse directions.

KEYWORD: Surface plasmon resonance, biosensors, spatial resolution, deconvolution


Page 3 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Surface plasmon resonance detection and imaging are powerful methods for studying and quantifying molecular interactions1,2. Using a high numerical aperture objective, high spatial resolution surface plasmon microscopy (SPRM) has been demonstrated3, and applied for studying biological samples, including cells4, bacteria5, virus6, DNA molecules7, liposomes8 and proteins9. Recently, SPRM-based electrochemical current and impedance imaging techniques have been developed, allowing study of local electrochemical reactions of heterogeneous

surfaces10,11,

catalytic

reactions

of

nanomaterials12,13,

and

cellular

processes4,14. The temporal resolution of SPRM can be as fast as µs, which is superior to fluorescence microscopy and scanning probe microscopy. However, its lateral resolution is limited by the finite decaying length of the plasmonic wave along the propagation direction (longitudinal)15. For example, the SPRM image of a nanoparticle has a parabolic tail of many microns long in the longitudinal direction6,10,16,17. One approach to improve the SPRM spatial resolution is to decrease the decay length by either using short wavelength light to excite the surface plasmons,3 or introducing metal nanostructures on the metal film.18 This approaches results in less well-defined surface plasmon resonance, and thus loss in the SPRM sensitivity for imaging biological materials and detecting molecular binding processes.3 Another approach is to obtain a SPRM image by scanning a focused laser bean line by line across the surface. Mechanical scanning of the laser beam in the approach inevitably lowers the temporal resolution that is required for studying fast binding kinetics and surface processes.19,20 Finally, a high spatial resolution SPRM image may be reconstructed from images taken at different incident light angles.21 The need of acquiring multiple images at different angles complicates the optical setup and slows down the imaging speed. In this note, we describe a digital method to improve the spatial resolution of SPRM. It uses imaging processing in Fourier space and deconvolution of complex field

22

, so that it



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 15

does not require change in the optics nor acquisition of multiple images. More importantly, it improves SPRM spatial resolution without compromising its high temporal resolution and sensitivity.

EXPERIMENTAL SECTION Materials and experimental setup: The SPRM experiment was performed on an inverted microscope (Olympus IX-81) with a 60X high numerical aperture (NA 1.49) oil immersion objective. Light with wavelength of 680 nm from a super-luminescence diode was used to excite surface plasmons on a gold thin film (47 nm) on a glass slide, and the plasmonic images were recorded by a CCD camera (AVT Pike F-032B) at a frame rate of 106 frames per second. The images were processed and analyzed by a Matlab program. 100 nm polystyrene nanoparticles (Microspheres-Nanospheres, Cold Spring, NY) bound to the gold film in 150 mM phosphate buffer, were recorded with SPRM. Image processing: Raw SPRM images recorded by the camera were converted to 16-bit tiff format files with a Matlab program. The background noise was removed by subtracting the first image. 2D Fast Fourier Transform (FFT) of the image was performed with the Matlab program to convert the SPRM image of nanoparticles from real space to k-space, which revealed a ring-like feature. The radius of the ring was obtained to determine the wavevector of propagating surface plasmon wave.

RESULTS AND DISCUSSION We first present the theoretical basis of the proposed image reconstruction algorithm, and then describe the algorithm step by step. Next, we demonstrate the effectiveness of the algorithm by applying it to experimental SPRM images of nanoparticles. We then conclude the paper by discussing the advantages and remaining limitations of the present method.


Page 5 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


SPRM image mechanism and reconstruction. Rigorous SPRM theories can be found in literature23,24, but here we present a simplified approach. The principle of SPRM image of an object (e.g., nanoparticle or virus) near the metal film is illustrated in Fig.1a, showing that a SPRM image is formed by two basic processes (Fig. 1b)17: scattering of the object by surface plasmon wave, which results in a scattered wave denoted by Esc (phase image shown in Fig.1b), and interference of the scattered field (Esc) with the planar surface plasmon wave (Ep) (phase image shown in Fig.1b).17 Accordingly, SPRM image is expressed as, (, ) ∝ +

,

(1)

where I(x,y) is the SPRM image intensity at location (x,y). In the single scattering regime25, we can describe the scattered field as the convolution of the subject, O (spatial distribution of refractive index), and point spread function (PSF), h, given by Esc=O*h, where * is convolution, h is the field scattered by a point scatterer. h can be represented by a cylindrical wave with amplitude decaying over propagation distance, ℎ =

(2)

where is the polarizability of the object (nanoparticles), ! is the decaying constant of the surface plasmon wave, k0 is the wavevector of the surface plasmon wave, and r is the distance to the scatterer. In Fourier space, the decaying cylindrical wave described by Eq. 2 is a ring with radius of k0, and thickness of !. The interference of the scattered field for a point-like object and propagating plasmonic wave results in the characteristic parabolic tail in the SPRM image, which renders the lateral resolution poor, as discussed above. To improve the lateral resolution of SPRM, we wish to reconstruct the object image (O) from the measured SPRM image (I(x,y)) with the procedure outlined in Fig. 1b. The basic



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 15

idea is to obtain the scattered field, Esc, from I(x,y), and then reconstruct the object image, O, via deconvolution of Esc with h. We describe each step of the procedures in details below using SPRM of a 100 nm polystyrene nanoparticle as an example. Fig. 2(a) shows the measured and calculated SPRM images of a nanoparticle. The intensity profiles across the nanoparticle images along the plasmonic wave propagation direction reveals the slow decay in the intensities of both measured and calculated images. To obtain Esc from I(x,y), we used a filtering method in Fourier space similar to the reconstruction of 3D images in optical holography (Fig.1c).26 First, we multiply the SPRM image I with the propagating plasmon field, Ep, which leads to

∙ # = $ + |&' | ( ∙ + ∙ &' + ∙ &' ∗ , (3) where the first term on the right-hand side is a planar wave, the second term is the scattered field, and the third term is the conjugate of the scattered field. The three terms partially overlap in Fourier space (yellow dashed square in Fig. 2b), and the second term (Esc) corresponds to a well-defined ring (blue dashed circle at the center of Fig. 2b), which determines the image contrast of the object. To isolate the second term (Esc) from the other two terms in Fourier space, we apply two filters (Fig. 2c), M1 and M2 to Eq.3, which leads to *+&' ,~*. ∙ # / ∙ 01 ∙ 0 .

(4)

The purpose of M1 is to isolate Esc (the second term) in Eq.3 from other terms, which passes only the ring in the Fourier space. M1 takes the form of, 01 234 , 35 6 =

:

789 : ;< : = @>? :

, (5)

which equals 1 on the ring, and decays rapidly with a constant of 31 (the thickness of the ring) from the ring. Because the part of the ring within the yellow dashed square in Fig.2b is noisy, we applied

another filter M2 as defined by 6 ACS Paragon Plus Environment

Page 7 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3 > 31 1 0 234 , 35 6 = A (?):/>: : , 3 < 31

(6)

to remove the noise within the square, where 3 = 34 cos H + 35 sin H, and k1 and ks2 are

chosen according to the signal to noise ratio around k0, and H describes the propagation direction of the surface plasmon wave (Fig. 2b). Although the two filters defined above remove part of the scattered field, the overall signal to noise ratio does not decrease. We will discuss this point later. With &' obtained with the procedures described above, and h defined by Eq.2, we now

turn to the reconstruction of the objective image (O), which is related to &' and h in Fourier space according to, *+K, = *+&' ,⁄*+ℎ,.

(7)

The object image (O) is given by, K = |* 1 +*+K,01 ,|. (8)

We note that because *+ℎ, is in the denominator of Eq. 7, which may amplify noise in &' , we applied filter M1 in Eq. 8 to reduce this noise. The SRPM image in k-space after each step is illustrated in Fig.1c. Application of the method to experimental images. To examine the performance of the image reconstruction method, we imaged 100 nm polystyrene nanoparticles. Fig. 3(a) shows a typical SPRM image of multiple 100 nm polystyrene nanoparticles, each with a long parabolic tail. The parabolic tails of the nanoparticles often make it difficult to identify signals from different nanoparticles. Following the procedures described above, we first determined the scattered field from the SPRM image with Eqs. 3-6, and then the propagating plasmon wave (Ep) from the radius (k0) of the ring in Fourier space. Next, we determined the point spread function h, by computing the field scattered by a single nanoparticle with Eqs. 36, using an experimental image of a single nanoparticle as input (e.g., Fig. 2a). Knowing Ep and h, we reconstructed the object image with Eq. 8. 7 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 15

Fig. 3b shows the reconstructed image of Fig. 3a, where each nanoparticle appears as a bright spot with the parabolic tail removed. Note that all the nanoparticles remain in the image after reconstruction, even those with low contrast ‘tails’ (i.e. those indicated by the blue arrows in Fig.3a). To evaluate the improvement in the spatial resolution along the plasmonic wave propagation direction, we plot the SPRM intensity profiles across the same nanoparticle from the original and reconstructed images in Fig. 3c and d. We achieve an improvement in the spatial resolution in the longitudinal direction by ~3 folds (Fig.3c) without affecting spatial resolution in the transverse direction (Fig.3d). The achieved spatial resolutions are ~310 nm (full-width-half-maximum) in both longitudinal and transverse directions, which is close to the diffraction limit of the optical system, which is about 230 nm. The scattered wave in the present method was calculated by assuming single scattering of the plasmonic waves by the object, which should be applicable to the study of nanoparticles and viruses as long as they are not too densely packed on the surface. Multiple scattering may not be negligible when the size of the object is large, and/or the average distance between different objects is small, compared to the wavelength.27,28 In those cases, a more sophisticated algorithm based angular spectrum reconstruction techniques29 may be applied.

CONCLUSIONS We have demonstrated a simple method to achieve high spatial resolution of SPRM without additional optics or sacrifice of its temporal resolution and sensitivity. Using the method, we have obtained SPRM images of nanoparticles with close-to-diffraction limit spatial resolution in both longitudinal and transverse directions. We anticipate that the method can be applied to the study of nanomaterials, including nanoparticles and nanowires, and biological species, including bacteria, viruses and macromolecules.


Page 9 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ACKNOWLEDGEMENTS We thank National Natural Science Foundation of China (Grant No. 21327008), and Gordon and Betty Moore Foundation for financial support. COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests. REFERENCE (1) Homola, J.; Yee, S. S.; Gauglitz, G. Sensors and Actuators B: Chemical 1999, 54, 3-15. (2) Nguyen, H. H.; Park, J.; Kang, S.; Kim, M. Sensors 2015, 15, 10481-10510. (3) Huang, B.; Yu, F.; Zare, R. N. Analytical chemistry 2007, 79, 2979-2983. (4) Wang, W.; Yang, Y.; Wang, S.; Nagaraj, V. J.; Liu, Q.; Wu, J.; Tao, N. Nat Chem 2012, 4, 846-853. (5) Syal, K.; Iriya, R.; Yang, Y.; Yu, H.; Wang, S.; Haydel, S. E.; Chen, H.-Y.; Tao, N. ACS nano 2015, 10, 845-852. (6) Wang, S.; Shan, X.; Patel, U.; Huang, X.; Lu, J.; Li, J.; Tao, N. Proceedings of the National Academy of Sciences 2010, 107, 16028-16032. (7) Yu, H.; Shan, X.; Wang, S.; Chen, H.; Tao, N. ACS Nano 2014, 8, 3427-3433. (8) Viitala, L.; Maley, A. M.; Fung, H. W. M.; Corn, R. M.; Viitala, T.; Murtomäki, L. The Journal of Physical Chemistry C 2016. (9) Maley, A. M.; Terada, Y.; Onogi, S.; Shea, K. J.; Miura, Y.; Corn, R. M. The Journal of Physical Chemistry C 2016, 120, 16843-16849. (10) Shan, X.; Patel, U.; Wang, S.; Iglesias, R.; Tao, N. Science 2010, 327, 1363-1366. (11) Shan, X.; Huang, X.; Foley, K. J.; Zhang, P.; Chen, K.; Wang, S.; Tao, N. Analytical chemistry 2009, 82, 234-240. (12) Shan, X. N.; Diez-Perez, I.; Wang, L. J.; Wiktor, P.; Gu, Y.; Zhang, L. H.; Wang, W.; Lu, J.; Wang, S. P.; Gong, Q. H.; Li, J. H.; Tao, N. J. Nat. Nanotechnol. 2012, 7, 668-672. (13) Fang, Y.; Wang, H.; Yu, H.; Liu, X.; Wang, W.; Chen, H.-Y.; Tao, N. Accounts of Chemical Research 2016. (14) Lu, J.; Li, J. Angewandte Chemie 2015, 127, 13780-13784. (15) Berger, C. E. H.; Kooyman, R. P. H.; Greve, J. Review of Scientific Instruments 1994, 65, 2829-2836. (16) Shan, X.; Diez-Perez, I.; Wang, L.; Wiktor, P.; Gu, Y.; Zhang, L.; Wang, W.; Lu, J.; Wang, S.; Gong, Q.; Li, J.; Tao, N. Nat Nano 2012, 7, 668-672. (17) Yu, H.; Shan, X.; Wang, S.; Chen, H.; Tao, N. Analytical chemistry 2014, 86, 89928997. (18) Kim, D. J.; Kim, D. J. Opt. Soc. Am. B 2010, 27, 1252-1259. (19) Somekh, M. G.; Liu, S.; Velinov, T. S.; See, C. W. Appl. Opt. 2000, 39, 6279-6287. (20) Watanabe, K.; Matsuura, K.; Kawata, F.; Nagata, K.; Ning, J.; Kano, H. Biomedical Optics Express 2012, 3, 354-359. 9 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 15

(21) Banville, F. A.; Söllradl, T.; Zermatten, P.-J.; Grandbois, M.; Charette, P. G. Opt. Lett. 2015, 40, 1165-1168. (22) Cotte, Y.; Toy, M. F.; Pavillon, N.; Depeursinge, C. Opt. Express 2010, 18, 1946219478. (23) Demetriadou, A.; Kornyshev, A. A. New Journal of Physics 2015, 17, 013041. (24) Demetriadou, A. Scientific reports 2015, 5. (25) Bozhevolnyi, S. I.; Coello, V. Physical Review B 1998, 58, 10899-10910. (26) Kim, M. K. PHOTOE 2010, 018005-018005-018050. (27) Bozhevolnyi, S. I.; Vohnsen, B. Physical Review Letters 1996, 77, 3351-3354. (28) Bozhevolnyi, S. I. Physical Review B 1996, 54, 8177-8185. (29) Barton, J. J. Physical Review Letters 1991, 67, 3106-3109.


Page 11 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


FIGURES

Figure 1. SPRM image formation and reconstruction processes. (a) Schematics of SPRM imaging. A camera records the reflected light from the gold/glass surface, including a planar plasmonic wave Ep and scattered plasmonic wave Esc; (b) The SPR image formation process is represented by the interference between the scattered plasmonic wave; The amplitude image of Ep, the phase image of Esc and the simulated SPRM image of a nanoparticle are also shown. (c) SPRM image reconstruction steps. See main text for details.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 15

Fig. 2 SPR image of nanoparticles and reconstruction. (a) Experimental (top) and theoretical (bottom) images of single polystyrene nanoparticle, and the decaying intensity profile at the direction of propagation. (b) Fourier space image of the SPRM image in (a). The blue dashed line shows the typical ring corresponding to the scattered wave, and the yellow dashed square


Page 13 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


indicate the overlapping of twin images and (c) The masks used for image reconstruction. Relevant parameters were chosen according to Fourier space image in (b).



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 15

Figure 3. SPR image reconstruction of 100 nm polystyrene nanoparticles. (a) Raw SPR image of nanoparticles. (b) Reconstructed image with improved resolution. Intensity profile across the nanoparticle marked in (a) and (b) by the blue dashed lines in longitudinal (c) and transverse (d) direction showed the improvement of spatial resolution. Blue and red curves and markers in (c) and (d) correspond to raw image and reconstructed image respectively.


Page 15 of 15

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC


Achieving High Spatial Resolution Surface Plasmon Resonance

Recommend Documents