Towards optical biosensors based onwhispering-gallery modes in microsphereresonators
2019
Whispering gallery mode (WGM) sensors have attracted a significant level of interest
recently due to their high level of sensitivity in the field of life sciences as biosensors.
However, integrated WGM sensor devices are still in their infancy. In this work, we
introduced a sensor device structure which is a step closer to commercial exploitation
and mass production of monolithic WGM sensor capable of multiplex detection.
The project started with the aim to design and then fabricate a sensor device
which can utilize the advantages of high Q factor of theWGM to detect biomolecules.
The sensing device would be able to detect small concentrations of biomolecules and
would support multiplex detection.
In this project we have designed and fabricated the sensor device successfully.
The design and fabrication of the sensor device consisted of fabricating an array of
planar single mode SU-8 WGs on glass substrate, developing a method to spin coat
and cure MY-133, which is a low RI material, matching the RI of water, and then
immobilising the microspheres on top of the WGs.
The first task in designing and fabricating the array of WGs was to select the
materials and estimate the dimension of the WGs which would serve the purpose
of the project. Therefore, we needed to find the appropriate materials, i.e., for the
substrate, the core, and the cladding layer of the WG and the microspheres. The
array of planar WGs were fabricated on a glass cover slip serving as cost-effective
substrate, and allowing to observe and analyse the light coupling into the WGs and
microspheres from the substrate side when the sample would be submerged inside the
fluid chamber filled with water in our measurement setup. Since the sensor device
will be submerged into the water filled fluid chamber, we researched materials of RI
close to that of water at the operating wavelength of the DFB laser (784 nm) to be
used as the cladding material of the planar WGs.
We have researched materials which have a RI close to water at 784 nm. However
many of these materials could only be cured by heating at high temperature. Such a
temperature curing procedure was used previously in our group to cure the cladding
layer on top the of SU-8WGs and then an attachment layer to glue the microspheres.
This temperature curing procedure involved temperatures above 100�C, which is not
suitable for microspheres functionalised with biomolecules such as antibodies. A few
alternative materials were investigated to address this issue. After going through
extensive safety protocols for processing in the clean room MY-133, a low RI material
was chosen for processing.
Apart from the safety issue, we choose MY-133 of RI 1.33 at 784 nm because it
could be spin coated and then cured by UV light. The UV curing method provides is fast, and a low temperature method. Curing by UV light allows to work with
biomolecules (antibodies) already functionalised on the microspheres. The principle
of light curable materials was discussed in Sec. 2.6. We have developed a standard
operating protocol to deposit MY-133 on SU-8 WGs given in Appendix B.2.
We choose SU-8 as the core material because it is highly transparent for the laser
wavelength, and it has a high RI of about 1.6, which is higher than both the glass
substrate and the cladding layer, allowing for light to propagate by total internal
reflection.
After choosing the appropriate WG materials we estimated the WG dimension,
discussed in Chap. 3. The thickness of the SU-8 layer for single mode propagation
was estimated between 450 to 1000 nm for a 3 μm wide WG. We estimated that
the thickness of the MY-133 on top of the WGs should be 550±50 nm for optimal
coupling between the WGs and the microspheres.
The next challenging part in this project that we faced after choosing the material
and estimating the dimension of the WG was to choose the correct type of SU-8 to
deposit the estimated target thickness of about 900 nm that can be reproducible. We
tried first with SU-8 2000.5 as this type of SU-8 is usually used to deposit a typical
layer of thickness less than 1 μm. However it was hard to reproduce the target
thickness. We therefore used SU-8 2002, which is usually used to deposit thicker
layer of SU-8. We experimented with SU-8 2002 and found out that it needed to be
diluted. We diluted SU-8 2002 with SU-8 2000.5 and found that a ratio of SU-8 to
2000.5 (2:1) is suitable for a reproducible SU-8 WGs of 900 nm thickness.
The second task in fabricating the planar WGs was to spin coat the diluted
MY-133 and then cure the MY-133 layer. MY-133 can be purchased in the form
of a gel and needs to be diluted by a fluorinated solvent in which MY-133 is completely
soluble. We chose HFE-7500 for its high boiling point (128�C) compared
to alternatives. The diluted MY-133 was spin coated to deposit an approximately
750 nm thick layer. We first tried to cure the deposited MY-133 layer by keeping the
sample submerged under water while exposing to UV light. However, this process
produced a tacky surface, so we had to develop a curing unit allowing to keep the
sample under inert gas during exposure. The second generation of this curing unit
can cure a MY-133 layer within one minute in inert gas such as nitrogen.
Once the WGs were fabricated we characterised them. Usually DekTak can be
used to measure the thickness of the WGs but in this project we have developed an
alternative method to measure the height of the WGs optically by using DIC microscopy.
It is a non-destructive method of characterising transparent material. We
have verified the DIC technique using DekTak measurements on the same samples.
We have also used DIC to determine the RI of the cured SU-8 layer.
The third task in this project was to glue the PS microspheres on the WGs. For
this we have spin-coated a layer of about 150 nm thickness of MY-133. Microspheres
were then drop casted in water onto this layer, and they adhere to the layer, making
contact with the cladding layer underneath. We characterised the PS microspheres
by first estimating the RI of the PS microspheres using the DIC technique. Then we
wanted to determine the footprint of the microspheres in the MY-133 gluing layer.
In this project we have used PS microspheres of 30 μm diameter, which gives the
submerged height of the microspheres in the MY-133 layer about 400 nm. However,
the DIC microscopy technique resulted in a significantly lower height - a point which
is open for future investigation.
The next task was to cleave the sample so that it can be fitted within the fluid
chamber. We have faced a set of issues, for example, when we were cleaving the
samples to define facets of the WGs, the SU-8 layer was peeling off from the facet.
In order to address this issue we have designed a cutting tool. The use of the cutting
tool improved the peeling issue. Fig. B.1 shows sample where the SU-8 layer were
peeled off when using a scriber to cleave the sample for opening the facet, and the
improved result obtained using the cutting tool.
We have developed an optical setup for biosensor experiments. The setup includes
a distributed feedback (DFB) laser, a fluidic chamber, and a linescan camera
to detect the WGMs of the MS among other optical components. The biosensor
device was positioned inside the fluid chamber by resizing it using the cutting tool,
retaining the majority of WGs intact to allow light travelling from WG input to
output. We can align the laser light from the DFB laser to a single WG from the
array of WG on the sensor device and observe the output light from the WG in a
2D camera. We can also determine the quality of the WGs by observing the light
intensity at the end of the WG in that camera.
We have further developed the experimental data acquisition software to generate
and control trigger signals for the laser and camera in optical measurement setup.
The software was also used for optical alignment and to save experimental data for
future analysis. We have shown that the laser light couples to the WGs and analysed
the blinking microspheres excited by the laser light. We have also analysed the scan
recordings from the linescan camera. Finally, we presented the sensitivity result
from our experiments in Chap. 5.
This work is an important first step towards an integrated biosensor based on
WGMs of microsphere resonators. We have successfully fabricated an on chip sensor
device and coupled the device with the laser light. We have also successfully excited
the microspheres. However, the Q-factor that we obtained from our experiments
is not as high as we expected. We have made suggestions to improve the Q factor
discussed in Chap. 5. In order to be able to couple an array of WGs to excite many
microspheres for multiple detection one should use the originally planned cylindrical
lens in the setup as shown in Fig. 4.1, combined with a micro-lens array. Such an
array could be fabricated by 3D laser writing with a Nanoscribe tool now available
in the department.
We have fabricated a sensor device with non-functionalised microspheres. Thus
this device can only be used for non specific sensing. Time did not allow us to fabricate
a sensor device with functionalised microspheres but for specific sensing one
needs functionalised microspheres glued on the sensor device. We have bought functionalised
PS microspheres and discussed briefly about specific sensing in Chap. 5 as
well. Importantly, the design and fabrication of the device was undertaken accommodating
the use of batch-functionalised microspheres, so that once sensing with
high-Q modes is achieved, introducing functionalised beads should not pose additional
technical issues. Thus the next step would be to fabricate a sensor device
with functionalised microspheres.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
0
References
0
Citations
NaN
KQI