Hello Adafruit forums!
I am trying to measure the concentration of polystyrene microspheres in a microfluidic solution running through a cuvette, and am here to solicit advise on how you think I could best go about this and what I might need to purchase.
Thank you in advance!
Trying to measure concentration of particles in solution
Moderators: adafruit_support_bill, adafruit
Please be positive and constructive with your questions and comments.
- belharethsami
- Posts: 7
- Joined: Wed Nov 14, 2018 4:57 pm
- zener
- Posts: 4567
- Joined: Sat Feb 21, 2009 2:38 am
Re: Trying to measure concentration of particles in solution
Generally you do it optically, either with transmittance or reflectance, or by scatter level. You might experiment to see if the conductivity is affected but that probably won't be a good approach. What light source you use might depend on the cuvette material and what wavelength can pass through it. Ideally your solution has no bubbles since these could be interpreted as spheres.
- belharethsami
- Posts: 7
- Joined: Wed Nov 14, 2018 4:57 pm
Re: Trying to measure concentration of particles in solution
Thus far I have tried doing it optically using the ideal absorbance wavelength tested using a spectrophotometer, but the numbers I am getting are all over the place. My current setup uses an LED connected to an Arduino and a photodiode on the other end, both made for the ideal wavelength (851nm).
- adafruit_support_bill
- Posts: 88037
- Joined: Sat Feb 07, 2009 10:11 am
Re: Trying to measure concentration of particles in solution
I've done it with the cuvette mounted in a microscope stage and an imaging system to identify and count the beads. But that required quite a lot of processing horsepower in addition to some expensive optical and mechanical systems.
If you post some more detail about your setup, we can take a look for details that might cause inconsistencies in your results.
If you post some more detail about your setup, we can take a look for details that might cause inconsistencies in your results.
- Attachments
-
- Capture.PNG (830.84 KiB) Viewed 334 times
- zener
- Posts: 4567
- Joined: Sat Feb 21, 2009 2:38 am
Re: Trying to measure concentration of particles in solution
First you need to test your transmitter and receiver setup with no liquid. Just make sure you can get stable readings when varying the strength of the signal. Once you get that working then you can start working with clean fluid to get some baseline readings and see if that is at all consistent. Then move on to a solution. I forgot about the vision approach. Scatter might be a better way to go than absorption. I don't know.
- belharethsami
- Posts: 7
- Joined: Wed Nov 14, 2018 4:57 pm
Re: Trying to measure concentration of particles in solution
@adafruit support bill, I am currently using a photodiode whose detection range is centered at 850nm and an LED at 851nm. I have potentiometers connected to the LED and photodiode to control the strength of the signal and the gain, respectively. @zener I have and have gotten linear results. The same is true when there is deionized water in the cuvette. For some reason I get different results from run to run when I use the microsphere solution and sometimes I have more light passing through (apparently). I would also like to mention that the solutions I am using have close to 3 million particles per milliliter, and so am not sure about the feasibility of using the microscope setup you mentioned.
- adafruit_support_bill
- Posts: 88037
- Joined: Sat Feb 07, 2009 10:11 am
Re: Trying to measure concentration of particles in solution
I am currently using a photodiode whose detection range is centered at 850nm and an LED at 851nm. I have potentiometers connected to the LED and photodiode to control the strength of the signal and the gain, respectively.
That much we gathered from your previous posts. But there are many details in the implementation of the optics, electronics, fluidics and/or mechanics of the system that can affect the consistency of the results.
The particle density has implications for the fluid dynamics as well as for the optics. One of the problems is achieving consistent distribution of the spheres within the cuvette (or at least the field of view of your sensing system). You may find it necessary to perform a precision sample dilution to measure higher population densities. Results can then be simply multiplied by the inverse of the dilution ratio.I would also like to mention that the solutions I am using have close to 3 million particles per milliliter, and so am not sure about the feasibility of using the microscope setup you mentioned.
- belharethsami
- Posts: 7
- Joined: Wed Nov 14, 2018 4:57 pm
Re: Trying to measure concentration of particles in solution
As far as implementation, the LED and photodiode are in a 3D printed part to hold them both snugly and firmly against the side of the cuvette. I typically wrap it in aluminum foil afterwards to prevent outside light from affecting it, although I typically do not see any significant difference in the results. The cuvette is 10mm wide. As far as the second point, it unfortunately is not possible to dilute due to the nature of the experiments being run. Even if the concentration is uneven, I would not mind integrating to find the total number of particles passing through per second or something of the like, but the numbers I am getting now vary from experiment to experiment, typically not intra-experiment. I would also like to thank you guys again for your help thus far.
- zener
- Posts: 4567
- Joined: Sat Feb 21, 2009 2:38 am
Re: Trying to measure concentration of particles in solution
I was going to suggest what Bill said, make a dilution. If you can't do that then I think you need to find another way. Is this a school project or a commercial application? Some other ideas off the top of my head:
- Conductivity
Make a much smaller aperture to shoot the light through (distance between transmitter and receiver)
Scatter: https://en.wikipedia.org/wiki/Static_light_scattering
Viscosity?
- belharethsami
- Posts: 7
- Joined: Wed Nov 14, 2018 4:57 pm
Re: Trying to measure concentration of particles in solution
Okay. Thank you for your help! @zener
- belharethsami
- Posts: 7
- Joined: Wed Nov 14, 2018 4:57 pm
Re: Trying to measure concentration of particles in solution
Update:
The setup now has an LED connected to the 3D printed part (mentioned earlier) with a fixed resistor and is connected as before. The photodiode should collect transmitted light to scale with concentration. Photodiode output is sent to a TIA (Texas Instruments OPA380AIDGKT) with fixed gain and setup according to recommended setup on the datasheet and then sent to be read by an Arduino Uno DAC at 50Hz using analog read. We see ranges for DI alone of 750 - 850. This is the same as before and is seen with varied concentrations, too. Why might we be seeing this, and what could be done to minimize this? If this setup does not work, how can a calibration curve be built without expensive optics? Ultra-high precision is not needed. We only need to be able to determine the difference between 2.5 million and 2 million particles per milliliter, and not the difference between 2.51 million and 2.49 million particles per milliliter.
The setup now has an LED connected to the 3D printed part (mentioned earlier) with a fixed resistor and is connected as before. The photodiode should collect transmitted light to scale with concentration. Photodiode output is sent to a TIA (Texas Instruments OPA380AIDGKT) with fixed gain and setup according to recommended setup on the datasheet and then sent to be read by an Arduino Uno DAC at 50Hz using analog read. We see ranges for DI alone of 750 - 850. This is the same as before and is seen with varied concentrations, too. Why might we be seeing this, and what could be done to minimize this? If this setup does not work, how can a calibration curve be built without expensive optics? Ultra-high precision is not needed. We only need to be able to determine the difference between 2.5 million and 2 million particles per milliliter, and not the difference between 2.51 million and 2.49 million particles per milliliter.
- adafruit_support_bill
- Posts: 88037
- Joined: Sat Feb 07, 2009 10:11 am
Re: Trying to measure concentration of particles in solution
Hard to guess with the details provided. If the results are consistent with no sample present, then the inconsistent results are likely due to an uneven distribution of the particles in the cuvette. The emission pattern of your LED is probably not uniform across the field. And the same is likely true of the sensitivity pattern of your diode. If the particle concentration is not uniform, the areas of brightest illumination and/or highest sensitivity will dominate the results.
- zener
- Posts: 4567
- Joined: Sat Feb 21, 2009 2:38 am
Re: Trying to measure concentration of particles in solution
If I understand what you said, you have a photo diode connected to an op-amp so that with varying amounts of received light it puts out a varying voltage. And then the Arduino reads that voltage through an analog input. Is that correct?
- belharethsami
- Posts: 7
- Joined: Wed Nov 14, 2018 4:57 pm
- zener
- Posts: 4567
- Joined: Sat Feb 21, 2009 2:38 am
Re: Trying to measure concentration of particles in solution
What is the voltage the op amp puts out, measured with a multimeter, with the different solutions (DI water, different concentrations)?
Please be positive and constructive with your questions and comments.