Homogeneous Light Source for Surface Plasmon Resonance Imaging

Peter Hausler, Simon Jobst, Johannes Fischer, Carina Roth and Rudolf Bierl

Sensorik-ApplikationsZentrum SappZ, Ostbayerische Technische Hochschule Regensburg,

Franz-Mayer-Str. 1, Regensburg, Germany

Keywords: Surface Plasmon Resonance Spectroscopy, Sensor, SPR-imaging, Miniaturization,

Micro-opto-electro-mechanical Systems.

Abstract: We describe how to build a homogeneous light source for Surface Plasmon Resonance Imaging (SPRi) which

mainly finds its applications in pharmaceutical screening and biotechnology so far. SPR spectroscopy is a

label-free, non-destructive and highly sensitive measurement principle for detecting changes in the refractive

index in close vicinity of a gold surface. A transfer of this technology to a miniaturized sensor will broaden

the range of possible applications. Commercial SPR assays are mainly working with a small number of

sensing spots. In contrast, the SPR imaging system shown here will allow the use of an array of many sensing

spots. In combination with chemical receptors designed as an artificial nose or an electronic tongue, the

simultaneous detection of many analytes is envisioned. So far, lasers or other inhomogeneous light sources

were used to illuminate the sensing surface, which is decreasing the systems sensitivity. We show a compact

(< 60 mm), low cost, LED based light source which is providing a large area (>300mm2) homogeneous top

hat profile. The combination of a high bit-resolution camera with our new light source enables a reflectivity

based surface plasmon resonance imaging system with a high refractive index unit (RIU) resolution.

1 INTRODUCTION

Surface Plasmon Resonance (SPR) technology is

label free, non-destructive and highly sensitive

(Schasfoort, 2017). Due to these properties, SPR is an

attractive measurement principle for chemical

sensors. Nevertheless, there are some drawbacks

limiting its applications so far: most measurement

setups are designed for being used in laboratories and

therefore they are very expensive. The high

temperature sensitivity and the need of trained

personal for its operation impede reliable in-field

sensing. Miniaturized and automated systems could

overcome these problems. Motivated by these

prospects, many miniaturized systems were

developed, (Ribeiro, 2019) includes an overview on

miniaturized SPR systems. Most of these systems

suffer either from missing transportability or from

low sensitivity.

Beside reflectivity based SPR imaging (Fig.1)

many other SPR technologies have been developed to

increase sensitivity, (Wang, 2019) is a review about

the state of the art and an overview of the most

common technologies. However, these technologies

are introducing new components which are

complicating miniaturization.

In order to be able to apply reflectivity based SPR-

imaging to miniaturized sensor systems we developed

a compact, low cost light source, which is providing

a large area, homogeneous illumination.

2 SENSING PRINCIPLE

While a thin gold film is irradiated by light, typically

the entire light will be reflected (Fig. 1). However, if

the light is p-polarised and the angle of incidence is

altered, one can see a narrow dip in the intensity of

the reflected light. This dip is indicating that at this

certain angle of incidence (SPR angle) surface

plasmons are excited. The SPR angle mainly depends

on the refractive index in close proximity to the gold

film which is deposited on a coupler, usually a glass

prism. Therefore, the refractive index on one side of

the gold is constant, which means that any variations

in the chemical composition – and therefore in the

refractive index – next to the other side of the gold

film is determining the position of the SPR-angle.

Selectivity to a special molecule of interest is

Hausler, P., Jobst, S., Fischer, J., Roth, C. and Bierl, R.

Homogeneous Light Source for Surface Plasmon Resonance Imaging.

DOI: 10.5220/0009168701630167

In Proceedings of the 8th International Conference on Photonics, Optics and Laser Technology (PHOTOPTICS 2020), pages 163-167

ISBN: 978-989-758-401-5; ISSN: 2184-4364

163

generated by a chemical functionalization of the gold

film by recognition elements. There are many

different ways on how to utilize surface plasmons

(Wang, 2019). The principle we used is reflectivity

based SPR, since it has the simplest setup. Hence it is

the most promising method for miniaturization.

Reflectivity based SPR has no moving parts, which

makes it more robust. During the design process one

fixed angle at the linear region of the SPR-reflectivity

slope is chosen for measurement. If the refractive

index of the analyte is changing, the position of the

minimum is moving and hence the reflectivity is

changing (R Fig.1).

3 SYSTEM SETUP

Figure 1: System-Setup, made of a LED based light source,

a gold-coated prism with receptors and a detector as well as

a 18 x 18 mm microfluidic chip, which is distributing the

analyte.

Most single or multi-channel SPR setups are using

laser light sources. For SPR-imaging, lasers are not

suitable since they are known for generating speckles

and diffraction patterns (Fig.2). Typical regions of

interest (ROI) do have a size between 50 µm and 5

mm. While having small size ROI, the noise which is

generated by the diffraction patterns is lowering the

resolution of the system dramatically. Most lasers

also have a Gaussian beam profile. If the sensing

surface is illuminated with a Gaussian beam profile a

small movement of the components will cause a

movement of the steep flank thorough the ROI and

hence a high change of the reflected signal. The

replacement of the laser by our LED based light

source overcomes this issue. LED light is very

broadband compared to laser light and it is non-

polarised. Therefore, a wavelength bandpass filter

and a polarizer were integrated into the system.

Furthermore, the beam is tailored and collimated. The

modified beam illuminates a 50 nm gold film which

is on top of a Schott F2 glass prism. The reflected

light is collected by a 2D-camera system, which

records the spatial change of the intensity of the

reflected light beam (Fig.1).

Superluminescence diodes would provide similar

characteristics to lasers but without spatial coherence

but they are excessively expensive. The high price

would impede the use of the SPR technology for

infield sensors.

Figure 2: Comparison of a LED light source left and a laser

light source right; the coherent laser light is generating

diffraction patterns due to dust particle, edges and air

bubbles.

The aim was to build a light source which is

available at a low price and which can be

miniaturized. Unstabilized lasers show mode

hopping, which would introduce noise and lower the

systems resolution. Speckles and other diffraction

patterns would also generate some noise.

Furthermore, the widening of the laser beam would

require a long optical path. Therefore, lasers are not

suitable for miniaturized systems.

Figure 3: Simulation of an SPR signal according to changes

in the excitation wavelength, the thickness of the gold layer

was fitted in every simulation to achieve a minimum

reflection; the SPR curve is getting steeper at higher

wavelength, which raises the sensitivity; simulation was

done with WinSpall software.

The SPR effect is very sensitive to wavelength

shift and its sensitivity is increasing with an

increasing wavelength (Figure 3). On the other hand,

the sensitivity of CMOS cameras is decreasing

PHOTOPTICS 2020 - 8th International Conference on Photonics, Optics and Laser Technology

164

rapidly with an increasing wavelength. Moreover, the

development of a system is much easier if the

wavelength is visible.

In order to have the system components available

for very low price at good performance we have

chosen a 660 nm LED from OSRAM (GH

CSSRM2.24). It is an important wavelength for

horticulture and is therefore readily available. In

order to be able to alter the driving current and to have

low noise at the same time, we used a Keithley

Sourcemeter as a current source for the laboratory

setup.

The beam shaping was done with a single 25 mm

diameter plastic aspheric lenses from Edmund Optics.

Depending on the desired beam characteristics, the

focal length was chosen at a range between 25 mm

and 75 mm (e.g. f40 #66-024). In order to avoid a

wavelength shift, a custom made 2 nm FWHM filter

from Chroma was used. The polarization was

controlled with a 1:9000 polarizer from Edmund

Optics (#85-919).

The prism is made of Schott F2 glass and has P4

polished surfaces to provide best performance.

By altering the position and the focal length of the

collimation lens, the divergence of the light can be

altered. If the SPR system is built with a light source

with low or no divergence a camera can be used even

without objective lenses.

The suitable camera should be chosen depending

on the demands, concerning refractive index unit

(RIU) resolution, system price and volume. The

ximea MU9PM-MH with its APTINA

MT9P031sensor is offering a 12 bit resolution at a

low price and very low volume (15 x 15 x 8 mm)

while the PCO Edge 4.2 is offering an outstanding

resolution of 16 bit. The PCO Edge with its 16 bit is

providing a high RIU resolution and the large chip

size is enabling an acquisition of a large SPR-Image

without the use of a lens. Between these two cameras,

there are many different cameras, which could be

used. Currently, a good compromise between costs,

size and bit resolution are cameras with a Sony

IMX178 sensor like the ISG Allegro.

4 RESULTS

Figure 2 shows the comparison of an LED light

source (left) and a laser light source (right). Both are

providing a “TopHat like” beam profile. In case of the

laser, a very small area at the centre of the Gaussian

beam profile was cut out. This Process is generating

diffraction pattern at the edge of the lenses which

were used to cut out the centre part. Furthermore, the

edges of the prism and microfluidics as well as air

bubbles and dust are generating diffraction patterns.

The result is a very noisy illumination which could be

sufficient for single or multi-channel measurement

with very large regions of interest (ROI) but it is

totally insufficient for high resolution SPR imaging.

Figure 4: False colour SPR image made with a laser based

light source and a Ximea CMV4000 10 bit camera; left:

diffraction patterns originating from edges, dust particles

and air bubbles are visible; right: cross-section along the

black line from the left side, the noise which is originating

from diffraction is clear to see

Figure 4 shows a SPR image which was recorded

with a laser based light source. The cross-section,

which was taken along the direction of the black line

shows that there is tremendous noise which is

lowering the RIU resolution of the SPR imaging

system. To avoid this kind of noise a LED based light

source was developed. LED’s do not have spatial

coherence, therefore they do not generate diffraction

patterns.

Figure 5: False colour SPR image, made with a LED light

source and a ISG allegro 14 bit camera; left: the SPR active

area is illuminated homogeneous, red areas are the sealing

and a air bubble; right: cross section along the the direction

of the black line, the cross section shows the SPR curve

which is originating from the divergent illumination.

Figure 5 shows a SPR image, which was recorded

with a LED based light source. The cross-section,

Homogeneous Light Source for Surface Plasmon Resonance Imaging

165

which was taken along the direction of the black line

shows the SPR curve which is originating from the

light source divergence. A LED cannot be collimated

like a laser, therefore a small divergence will always

remain. Unlike Figure 1, which shows perfectly

collimated light from a laser, a LED based light

source is generating light rays, which are not perfectly

parallel. Therefore, the angle of incidence is shifting

slightly from one side to the other side of the sensing

area. This divergence can be utilized to monitor every

sensing spot at its most sensitive angle. If the SPR

system does not monitor only one receptor but many

different, e.g. 100 spots, it is very likely that every

spot has a different refractive index. This implies that

only one spot is monitored at its most sensitive SPR

angle if a non-divergent light source is used, all the

others are falling behind their possibilities. However,

if a divergent light source is used, the angle of

incidence is splayed and every receptor can be placed

at its most sensitive angle of incidence.

While collimating a laser is very simple,

homogenizing an LED is very complex. One could

use an opal glass but this option would cause

tremendous power losses on the one hand and

focusing on infinity would not lead to a homogeneous

beam shape on the other hand. Therefore, we

developed another technique to create a

homogeneous beam area. We have found an

empirical equation which relates focal length of the

lens, distance of the lens to the desired homogeneous

surface and lens placement:

















⋅exp



















With





: Placement relative to focal plane





: Distance of lens to homogeneous surface

: Focal length of lens

Parameter Value





2.5698







1.9311





37.265







0.0383

Figure 6: Placement of optical components, the 1mm

aperture in front of the opal glass has been omitted in the

schematic.

The SPR sensing area is inclined according to the

beam direction (Figure 1) therefore the beam has to

be homogeneous in a wider area along the beam

direction and not only in one plane. The 



is the

distance between the lens and the centre of the

homogeneous area.

This equation is valid for focus length in the range

of f 20 to f 120 and for a lens – homogeneous region

distance of 40 mm up to 110 mm. The equation

should be understood as an approximation, which is

helping to build a homogeneous light source, which

is based on a non-homogeneous LED chip. Since

most LED’s do have a lens on top, we placed an opal

glass in front of the LED to determine the equation.

However, we did the same procedure without opal

glass and it lead to the same homogeneous results.

The equation without opal glass has different

parameters because of the added lens on top of the

LED.

Placing the SPR prism, the lens and the light

source according to this equation will lead to a

homogeneous illumination of the prism surface. If the

light is s-polarized the whole surface is iluminated

homogeneously. If the light is p-polarized a gradient

like it is shown in Figure 5 appears. The reduction of

the reflectivity is caused by the generation of surface

plasmons and the gradient of the reduction originates

from the remaining divergence of the light beam.

Combining this technology with a low noise current

source creates a light source, which enables high

resolution SPR imaging.

PHOTOPTICS 2020 - 8th International Conference on Photonics, Optics and Laser Technology

166

Figure 7: Plot of the equation, as a function of focal length

and distance between lens and homogeneous surface; red

dots are measurements conducted with the described setup.

Figure 7 shows the plot of the equation. While

utilizing our method one has to deal with four

parameters, which are divergence, width of the

homogeneous area perpendicular to the beam

direction, width of the homogeneous region along the

beam direction and distance between LED and

homogeneous region.

One could use a lens with low f-value to achieve

a short distance between H-region and LED, while

doing so, the divergence will increase and the

spreading of the SPR-angle will be higher. On the

other hand, one could use a high f-value to achieve a

low divergence, this would increase the distance

between H-region and LED. Therefore, one has to

choose the best option for every task.

5 CONCLUSIONS

SPR-imaging was lacking of cheap, large area and

homogeneous light sources. We described how to

build a cheap, LED based light source, which is

providing a homogeneous illumination over the entire

sensing surface. The beam profile is homogeneous

over a wide range along the beam direction and not

only in one single plane. An empirical equation shows

how to place lens and light source in relation to the

sensitive area. We were able to reproduce the

experiment many times and it is in use every day.

However, the equation is not very exact and should

therefore be understood as an approximation. The

empirical equation is lacking a theoretical equation,

thus more investigation has to be done in this field.

However, the concept can already be used and it

enables high-resolution SPR-imaging on large

surfaces. The light source was used for high RIU

resolution experiments. Combined with a PCO Edge

SCMOS camera we achieved resolutions in the 10

-7

RIU range. The results are currently under

publication.

ACKNOWLEDGEMENTS

This work is funded by the Bavarian Ministry of

Economic Affairs, Energy and Technology

(ESB051/006).

REFERENCES

Schasfoort R., 2017. Handbook of surface plasmon

resonance, Royal Society of Chemistry, London.

Ribeiro P., Raposo M., 2019. Optics, Photonics and Laser

Technology 2018, Springer, Heidelberg.

Wang D., 2019. Recent Advances in Surface Plasmon

Resonance Imaging Sensors. Sensors, 19(6), S. 1266.

doi:10.3390/s19061266

APPENDIX

Figure 8: Cross section along the direction of the black line

of figure 5 with s-polarized light.

Homogeneous Light Source for Surface Plasmon Resonance Imaging

167