We designed the first generation device based on our needs, using multiple optical sensors with multiple filters mounted in front of the sensor. We wanted to select fluorescent probes that emit at different wavelengths but at the same excitation wavelength. These fluorescent probes would be excited to fluoresce at this wavelength, and the fluorescence emitted would be detected by multiple probes and the data would be uploaded to a microcontroller, which would then upload the data to a computer. But it soon became clear that there was a major problem with this design: if the sensors were large, the sample size had to be very large, which was a big waste of reagents. Also, multiple optical sensors were used side by side, increasing the cost of the device. As a result, the first-generation device was quickly phased out, and we designed our second-generation device.
This is our device 2.0. Compared with the first-generation device, it only uses a light sensor and has a structure called a filter disc: this is a disc with multiple filters on it, which can do 360 degree rotation, and controlled by servomotor. By rotating the turntable, we can mobilize different filters to make it under a detection light path. But soon we discovered that there was a problem with the design principle of this device. Because there are not a variety of fluorescent probes which can emit different fluorescence at the same excitation wavelength and have very different wavelengths. At the same time, we found that the cost of the filter was very expensive. So we improved our device again.
This is our device 3.0, which eliminates the filter structure very well. This is because the LED emitting light source with a fixed wavelength is much cheaper in contrast, and the LED light source has good directivity and no obvious scattering. On the emission path of the LED, the amount of light emitted by the LED will be rarely received by the detector. Therefore, the light received by the detector theoretically is the light emitted by the fluorescent probe. We can light up the LED lights of different wavelengths to excite the corresponding fluorescent probes in order to accomplish the detection of multiple fluorescent probes one by one. At this moment, we have selected the fluorescent probes, and their excitation wavelengths correspond to dark blue, grass green, yellow-green and red respectively.
In the communication with Bioer Technology Co., Ltd., the director of the device department told us that the amount of fluorescent probes in limited-time detection is very small. Without a filter on the detector for detection, the background light will have a strong effect, which can completely cover up the emitted fluorescence. Moreover, the light emitted by the LED lamp is not pure enough, and a first-level filter is needed. The director of the device department was very enthusiastic and provided us with four pairs of filters corresponding to our fluorescent probes, which greatly promoted the progress of our device and was a milestone event. Our device has been officially produced since this time.
In Device 4.0, we use full-wavelength LED lights as the light source, and at the same time we use the four pairs of filters provided by Bioer Technology to make two filter turntables, which are controlled by servomotors. We use a silicon light tube as a photosensitive element, coupled with a micro-voltage amplifier circuit, to connect with the microcontroller.
We have made such a pair of filter turntables with servomotors based on our needs, and it can work very well. (The reason why we don’t use 360° servo is that the 180° servo can be turned to a fixed angle, and no correction is needed before use)
Our filter turntable has undergone several generations of changes (from left to right). The upper left is our first generation filter turntable. Due to the limitations of 3D modeling capabilities at that time, only a turntable with eight holes was made. We need to add materials to fill other loopholes in the later stage; the upper right is the second-generation turntable, which fills the vacancy through 3D modeling; the lower left is our third-generation turntable, which perfectly drills holes so that the accessories of the steering gear are perfectly embedded in it, which can completely avoid the use of glue guns and greatly increase the flexibility of the turntable; the lower right is the turntable in our final device and it has a small hole at the end to prevent light from passing through the gap between the filter and the turntable, which will affect the accuracy of the experimental results. It can be seen that its light hole is much smaller than the outer diameter of the filter, so the use effect is very good.
In order to make it easier to put the sample into the device, we use special structure for testing: a white box, which allows two slides with a piece of paper in the middle to be inserted and fixed in it. The corresponding light can pass through the sample almost non-destructively through the light hole.
Standard board: In order to make the whole device can be tightly connected by inch screws, we designed standard boards. These boards have light holes in fixed positions and screw holes on four corners to ensure the straight-through of the light path.
With the rules of the standard board, we have designed the next generation (upper left) and the second generation sample plate (lower left). The second generation occupies a smaller volume, saving material costs. (The picture on the right is the real picture of the first generation. Unfortunately, the second generation was built into the installation without taking a picture)
There are many kinds of main components of the photosensitive circuit. We explored silicon light tubes, photodiodes, and phototransistors. The generated photoelectric signal can be recognized by our single-chip microcomputer after first-level amplification. However, after testing, it is found that the light leakage is quite severe, and the intensity of the background light masks the intensity of the fluorescence.
After countless attempts, we made the final product 1.0. But this seemingly "seamless" finished product has many problems found in actual combat tests. The most serious one is that the light leakage is very serious in the entire light path.
Explore the source of the problem:
For some of the above problems, Changqing Luo and I designed the following experiments to verify the feasibility of the final device. For the Rhodamine dye, we selected a pair of filters which meets the characteristics of the Rhodamine dye. We designed a black box with the cheapest waste materials. The black box is composed of a cardboard box with several layers which are light-proof barriers. The pen shell of the marker pen and the rubber cover of the pen are used as the light path. Put the filter, the sample, the secondary filter, and the silicon light tube in it in turn, and connect the circuit.
After placing everything, we added the top cover of the device. According to the reading of the voltmeter, our silicon light tube is in an almost absolute dark environment. After turning on the light, it is found that there is still weak light entering, indicating that the filter can not filter the light completely. We found that the distance between the silicon light tube and the secondary filter has an optimal value, so that the interference will be less. After that, we inserted a glass slide in between, sandwiching a blank filter paper, and we got a stronger signal, which was in line with expectations, because there was fluorescence on the filter paper. And when we put in 0.11 mM Rhodamine, we got very good results. It is 6 times higher than the value of the blank filter paper control! However, when using 0.11um Rhodamine, our device is currently unable to detect it. The reason might be that the self-fluorescence of the filter paper is already stronger than the concentration of the experimental sample.
On September 13, Changqing Luo and I made the first-generation photosensitive chip, which has a reduced size. However, it was found that the use of photosensitive triodes was not effective and needed improvement, so we plan to continue using silicon photovoltaic cell.
Due to the limitation of our ability, the photosensitive element made by ourselves is not effective enough, so we bought the photosensitive element online.
We connected this amplifier correctly, adjusted it to zero and soldered the silicon photovoltaic cell on it, and we fixed it to the partition with screws and glue gun.
There are two potentiometers, which are used to adjust the zero point and the magnification level. We found that when the magnification level is high, its fluctuations will be obvious. Only at the position of the third pair of filters, there is obvious light leakage, and the other filters are normal.
We have overcome various difficulties and finally made our finished detector. Although the appearance is not beautiful, the effect is still very good. (The detailed content will be mentioned on the Engineering page)
To do an iGEM project well, we need to have a lot of knowledge behind it. To build a complete system, we are still far away. But in this summer vacation and our daily spare time, we kept learning and constantly overcame various obstacles. Through the iGEM competition, we did learn a lot of things that others could not learn. We always believe that we will get better!
We are also building a user-friendly software window, hoping that it can give users correct guidance for the first time. However, due to the limited time and the unfavorable factors brought by the epidemic, this ideal cannot be completed before wiki freeze date.