Received date: August 07, 2018; Accepted date: August 28, 2018; Published date: August 31, 2018
Citation: Pechprasarn S, Melnikov A, Chow TWK, Somekh MG (2018) Widefield Confocal Microscope for Surface Wave K-Vector Measurement. Nano Res Appl. Vol.4 No.2:4. doi:10.21767/2471-9838.100029
Copyright: © 2018 Pechprasarn S, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Optical surface waves are guided light on surface of optical structures. There are several optical structures that support optical surface waves, such as, nanostructures, gratings, optical waveguides and metamaterials. Optical surface waves have proven themselves very promising candidates for several applications including, biosensing, optical computing and optical circuitry. The characteristics of surface wave can be characterised by the wave vector (k-vector) of the surface wave. In this paper, we will discuss how a modified confocal microscope integrated with a phase spatial light modulator allows us to measure both the real part and the imaginary part (attenuation coefficient) of the surface wave k-vector. Surface plasmon resonance (SPR) excited on a uniform gold surface through Kretschmann configuration is employed as an example in this talk. Note that the system is not limited to the SPR. It is also applicable to other types of optical surface waves. We have demonstrated in our recent publication that the modified confocal not only provides the k-vector measurement both real and imaginary, it also allows us to separate different loss mechanisms in SPR. One limitation of the system was the single point detection. Here we will discuss the current stage of our development in widefield confocal surface plasmon microscope, which allows us to measure multiple points simultaneously. This has been achieved by integrating another amplitude spatial light modulator in the image plane of the objective lens. Use of orthogonal illumination patterns allows the image plane to be sequentially amplitude modulated and post-processed to recover the confocal image.
Optical structures; Biosensing; Metamaterials; Light intensity
Recently scientists and engineers have been interested in applying optical surface waves  in several fields including biosensing [2-6], refractive index measurement [7-8], voltage sensing [9,10], ultrasonic detection [11-13], microscopy [5,14,15], optical computing  and optical circuitry  because of their unique optical properties. The optical surface waves are guided electromagnetic waves on surface of optical structures and devices, such as, metamaterials [18,19], gratings , nanostructures  and optical waveguides . Optical surface waves can be characterised by their wave vector (k-vector) of the surface wave. The k-vector of optical surface wave consists of the real part k’ and the imaginary part k’’. The real part k’ indicates the k-vector value, which the optical surface wave can be excited. The imaginary part k’’ is the attenuation coefficient.
In this paper, surface plasmon resonance (SPR) excited on a uniform gold surface through Kretschmann configuration  is employed as an example of optical surface wave. Note that the methodology developed in this paper is not limited to only the SPR, but it is also applicable to other types of optical surface waves.
Since 2012, our group has been working on developing confocal surface plasmons microscopy [2,8,24-28] for quantitative interferometric imaging of biological samples. Simplified schematic for the confocal surface plasmon microscope for the SP wave vector measurement (ksp) is shown in Figure 1. It depicts a defocused plasmonic gold sample 50 nm thick where the SPs are excited at its plasmonic angle θsp. The SPs excited at point ‘a’ depicted in Figure 1 propagates to point ‘b’ and beyond it. Confocal point pinhole allows reradiated SPs that appear to come from the focal point to be detected (path P2). The SPs reradiated from the other planes will be blocked by the confocal pinhole . This allows us to have a well-defined detection path . On the other hand, the light around the normal incident beam does not excite the SPs. This beam is directly reflected by the gold surface (path P1) and detected by the detector. These two beams can then form and interferometer. This modified confocal microscope is operated by defocusing the sample axis z, so called, V(z) measurement . The other k-vector components that do not contribute the interferometric signal is then attenuated and blocked by amplitude pupil function as shown in Figure 2. The amplitude modulation is achieved by a phase spatial light modulator (phase-SLM) aligned so that it is in one of the conjugate planes of the back focal plane of the objective lens. The amplitude pupil function is modulated by an antiphase pair of consecutive pixels on the phase-SLM to cancel out the light intensity. The phase-SLM does not only provide the amplitude modulation, it also allows us to do electronic defocus by imposing the defocus phase profile on the phase-SLM as shown in Figure 3(a). This enables us to the measure the V(z) curve without physical defocusing the sample  as shown in Figure 3(b).
There are ripples in the V(z) curve shown in Figure 3(b). These ripples can be related to the plasmonic angle θsp by the following equation:
Where Δz is the ripple period in the V(z) curve.
n is the free space wavelength
n is the refractive index of the coupling media
θsp is the plasmonic angle
From the Equation (1), the recovered plasmonic angle is ~46.78° degrees, which agrees with the plasmonic angle of air backing sample. This also indicates the real part of the ksp or k’sp. perform phase stepping measurement . the phase SLM can be employed to modulate the phase of either path P1 or P2 by 90 degrees step. This allows us to perform phase stepping measurement and recover the relative phase φ(z) of the SPs as shown in Figure 4. The φ(z) can be calculated by:
Where φ(z) is the relative phase between the relative phase between the SP and the reference beam.
I1, I2, I3 and I4 are the V(z) curves in intensity with 0 degree, 90 degrees, 180 degrees and 270 degrees phase difference between path P1 and P2 respectively.
For the imaginary part of ksp or k’’sp, this can be measured by blocking the P1 and allowing the path P2 to be detected by the confocal pinhole. The exponential decay curve can be detected by measuring the V(z) measurement. The attenuation coefficient can be determined by fitting a linear function to the natural logarithmic scale of the V(z) curve .
This system with a single phase SLM can only be used to scan a single object’s point at a time. In this paper, we will demonstrate that by complementing the phase-SLM with another amplitude spatial light modulator (amp-SLM) this allows us to perform multipoint scanning simultaneously.
One of the main aims of this paper is to develop a widefield confocal surface plasmon microscope by designing a confocal microscope that allows us to measure multiple points simultaneously. Here it is achieved with sequential orthogonal coding in the image plane so different illumination points do not suffer from crosstalk. The uniqueness of the proposed system lies in the fact that a spatial light modulator (SLM) in the back focal plane can be used to use to perform additional processing to all the beams in parallel.
We have designed and constructed the microscope system based on two SLMs as shown in Figure 5. As indicated above the first SLM was aligned in one of the conjugate plane of the objective lens image plane serving as an amplitude SLM modulating the intensity profile of the objective lens image plane allowing the image plane to be spatially time-coded. The second SLM acts as a phase SLM providing phase modulation in the back focal plane (BFP). The experimental setup for the multiple point scanning confocal phase SPR system is shown in Figure 5. The light source used in the system is 10 mW He-Ne laser at 633 nm free space wavelength. The laser beam magnified at the magnification stage is projected on the amplitude-modulating SLM’s surface (LETO Holoeye SLM), where it undergoes amplitude modulation controlled by computer-generated modified Hadamard sequences. The amplitude modulated optical beams are then projected on the phase SLM’s surface (PLUTO–NIR-011 Holoeye SLM) where corresponding phase profile (defocus and pupil mask) is introduced. After passing through another magnification stage, which matches the size of the phase-SLM screen and the size of the objective lens back focal plane (BFP), both amplitude and phase modulated illumination patterns are projected onto the sample by 1.25 NA objective lens (1.25 NA oil immersion objective by Zeiss). The same objective lens captures the light reflected by the sample. Two CCD cameras, the same magnifying optics and two beam splitters are used to capture the images of the image and back focal planes. The confocal pinhole is formed by a digitally at the image plane by processing the images obtained by the CCD camera. This allows us to change the size of the pinhole in the post processing stage.
Amplitude modulation is implemented using an SLM, which modulates intensity profile of the incident light by switching individual pixels ON and OFF, as it is illustrated by Figure 6. Independently controllable SLM’s pixels open the way to create illumination patterns, which bear no correlation with one another. This allows us to maintain confocal effect for every individual amplitude SLM pixel, since every illumination/ detection point is interference-free from the rest. A Hadamard code provides a convenient illumination sequence, since it will generate efficient orthogonal sequence. This is used to produce amplitude modulated signal in this project.
Original Hadamard code, HO, is a linear block code of N=2n length, comprising 1’s and -1’s and every line of code is correlated with itself only. However, it is not possible to implement negative intensity and therefore -1’s are going to be replaced with 0’s in illumination (modified Hadamard code HM). In 1D intensity distribution I(x) at the image plane of the microscope, using amplitude-modulated illumination, is a 2D matrix:
Where S(χ) is the sample function and O(χ) is the intensity point-spread function of the microscope, subscript “i” denotes individual lines of modified Hadamard code, and ⊗ is a convolution operator.
In order to recover individual confocal images of the sample I (x, N) , generated due to the different illumination patterns, I (x, N) needs to be correlated with original Hadamard illumination patterns, but since modified Hadamard code must have a zero mean to maintain orthogonality the mean value must be subtracted from I (x, N) first:
where “k” is the index of the pattern used to test the image against and k = i is a dot-product operator. Only when k = i , the value of will be nonzero, thus preserving confocality of individual images in multipoint scanning arrangement. The process of confocal image recovery is illustrated in Figure 7.
Since, both amplitude and phase modulation are required in order to achieve multipoint V(z) scan, the total number of 2D images captured by the image CCD camera is a product of Hadamard sequence length and number of defocus scan points. For every given defocus value, the whole Hadamard sequence must be obtained.
In order to extract the SPR related information, every Hadamardcoded stack of images, for the given defocus value, must be processed, according to the Equation (4). Then, confocal pinholes must be defined and placed on the recovered individual pointspread functions, and finally, the plasmon resonance angle can be extracted for every confocal pixel using the Equation (2) or from the gradient of φ(z).
Figure 8 shows the V(z) curve results recovered from 128-bit long Hadamard sequence and (b) recovered φ(z). It can be seen form the figure that the Hadamard sequence allows us to have a better quality signal and signal to noise ratio is much better than the results shown in Figure 3b. This is for a single point measurement.
The next step, we took multiple points measurement as shown in Figure 9. It can be seen from the figure that the method allows us to take a confocal image for each z defocus slice.
Confocal imaging is very broadly employed not only in the field of material science but also in cell physiology , histology , and neuroscience . Especially in the fast-growing field of living tissue/organ imaging. We believe that the high speed parallel focal spots imaging is very crucial in such biological imaging. In the previous section, we have demonstrated how the widefield confocal surface plasmon microscope works and how the system speeds up the confocal imaging. In this section, we will discuss some issues of the current system.
Signal to noise ratio and input light intensity
The current system was designed and aligned to ensure that the image plane was illuminated with a uniform beam profile. This was achieved by magnifying the Gaussian HeNe laser beam profile by 35 times. Less than 3% of the 10mW power arrives the CCD camera because of the magnification and necessary optical beam splitters. This has forced us to use quite a long Hadamard sequence, i.e., 128 bits to recover a decent signal.
We found that the experimental results obtained during the night time were more reliable than the experimental results obtained during the day time. This might be because of the bigger temperature fluctuation during the day time. We will enclose the system, control the temperature and humidity in the laboratory. If necessary, we will use an objective PID temperature control kit to ensure that the objective lens temperature is stable.
Spatial light modulator speed
The SLMs that we are using in the system is very slow. The crystal response time is around 0.1s. It therefore takes very long acquisition time. This makes the temperature fluctuation and drifting effect more severe.
Non-zero background amplitude SLM and low scattering light level
The amplitude SLM that we use in the system cannot produce a true zero intensity background level and the light scatter from a pixel is very low. This makes the point spread functions appear as a very faint signal on top of a non-zero background. This degrades the signal to noise ration and forces us to use a longer Hadamard sequence to recover the signal.
In this paper, we have explained and demonstrated how our proposed widefield confocal surface plasmon microscope measure the complex k-vector of surface plasmons. The system requires two spatial light modulators. Firstly, a phase-SLM aligned in a conjugate plane of objective back focal plane provides electronic defocus and amplitude pupil function modulation in the back focal plane. Secondly, an amplitude SLM aligned in a conjugate plane of objective lens image plane provides Hadamard sequential coding in the image plane. The coded images are then stored and the confocal image can be recovered by the method described in details in the Materials and Methods section. However, there are a number of issues that need to be addressed, such as, the non-zero background intensity of the amplitude SLM, the signal to noise and the speed of the SLMs. These issues will be addressed and reported in a later publication.
All Published work is licensed under a Creative Commons Attribution 4.0 International License
Copyright © 2019 All rights reserved. iMedPub LTD Last revised : March 22, 2019