We had another device idea which functions as follows: As the sample just need to be excited at specific wavelength to emit at another wavelength, we just have to excite it, and detect the emitted light through a photoresistance. So the device could look like this:


While discussing with a friend which do his master in optic, he told me that the usually way to make these kinds of measurements was to work with at least two filters. One of these filters serves as showed in the previous device example, but the second is able to reflect the wavelength which aren’t interesting for the measurements. One easy way to structure the device components could be like this:

Device with dichroic filter

The way we found to quantify the GFP intensity was to connect a simple photoresistance as the first resistance of a voltage divider so that we can just measure through the arduino uno microcontroller’s analogic pin:

Voltage divider

The Vout can be measured by the simple formula:

The photoresistance occupy the R1 place

The photoresistance is at the R1 place. A photoresistance react to light intensity by decreasing it’s resistance when the light it receives is increased, and inversely decreases its resistance when it is put in a darker environment. So when the light increase, R1 decreases so that Vout increases. But when the light decreases, R1 increases so that Vout decrease.

For quantifying the Vout this simple code was used in the microcontroller

const int sensorPin = 0; //The number of the pin Which is used to measure Vout

int lightLevel; //Which represent the Vout measure

void setup() //This function is called only at the moment the program is lauched
Serial.begin(9600); //A function to initiate the Serial function which
                    //serves to use the serial monitor to display the
                    //Vout values measured with the Voltage divider on
                    //the computer.

Serial.println("Start of the measurements"); //A first sentence to note
                                             //where the measurements begin

void loop() //This function will be called as long as the program works
lightLevel = analogRead(sensorPin); //AnalogRead is a function which
                                    //simply calculate the Vout value
                                    //which can be between 0[V] and 5
                                    //[V] and convert it to a number
                                    //between 0 (for 0[V]) and 1023
                                    //(for 5[V])

Serial.println(lightLevel); //Display the Vout value on the
                            //serial monitor of the computer

delay(1000); //Wait 1000 [ms] (= 1[s])

The idea was then to make a first graphic with known arsenic concentrations, of the intensity in function of the Arsenic concentration in order to calibrate the device. With this graphic it would be easy to determine the arsenic concentration in quantifying the GFP emission with the device, and looking in the graphic at which concentration it corresponds.

So with a few tools it could be easy to build a simple compact device which is able to quantify, depending on the different LED’s wavelengths, photoresistance and filters available, all kinds of fluorescence type.

So we needed to get optic filters. I began research on the web for all the different filter’s offers. I wanted to find two kinds of filters and two exemplar of both:

One filter and on dichroic filter letting wavelengths between 500 nm and 510 nm (the eGFP emission wavelength) pass but impermeable around 488 nm (the eGFP’s excitation wavelength) and one filter and one dichroic filter letting wavelengths around 360 nm (the Tryptophan emission wavelength) pass but impermeable around 280 nm (the tryptophan’s excitation wavelength).

I mostly searched on the Andover corporation website and the chroma website because it seemed that all the filters societies had around the same prices and because there was a wholesaler in Morges near the EPFL place which sells filters from both companies. But I also searched on the Rosco page and on a lot of eBay pages.

One problem I encountered during my researches was that we learned that the GFP we should work with, because it was used in the Van der Meer’s bacteria, was in fact eGFP which has a different excitation wavelength (around 488[nm]) which is much closer to the GFP and eGFP emission peaks (509[m]) than the normal GFP excitation peak which is around 395 [nm]. Thus the filter research became much more difficult, because the wavelength fork where the filter could change between no transmission (around the excitation wavelength) and maximum transmission reduced from 114[nm] to only 21[nm].

The better product Sachiko finally found was this filter for arsenic detection.

One thing I thought about during this research was the capacity of our photoresistance to perceive different wavelength, and I was surprised to discover on it’s datasheet that it was insensitive to wavelength lower than approximately 470 [nm] (on the datasheet the Cds one). So for the bacteria detection which should detect wavelength about 360 [nm], it would be necessary to find others photoresistance which detect it.

Finally we brought no filters at all because we found the filters we needed in the Biop facility at EPFL. This place makes different microscope available for all the Life science faculty needs. I meet José Artacho which nicely lend me the filters with a filter holder.

Finally came the day we tested the prototype. All the tools needed are visible on this picture:

Device components

So the tools are a computer to write the script for the microcontroller and supply the two hirshman plates which connect the excitation LED and the photoresistance with the measurement components, the filters in the filter holder, some structural elements and the Feldshlossen (red and blue) box to isolate the photoresistance from the LED’s lightening. After assembling it looked like this:

Device built

zoom into the device

The measurement device was able to discern between ambient light, which generally gives values between 70 and 110, very intense light which gives values about 400-600 (maximum was 1023), and absence of light was quantified at 0 or 1. So we tried to do some tests with bacteria which express constitutively eGFP (thanks to a plasmid), and with Dextran, a polysaccharide which reacts like GFP, but no light was detected by the photoresistance. So we tried to do the measurement without the filters to avoid the filter transmission factor which decrease the light intensity and we discover that the light from the LED was not detected when the LED was just put behind the filter holder. Another thing we tought about was the fact that the sample holder bottom was not really made for letting light pass through because of it’s curvature which refracts the light. So we tried to measure from the sample holder’s sides and we concentrate the light emission with a lens:

Device without filters
Here the filter holder is used only for compartmentalization as the

But unfortunately we only arrived to quantify an intensity of about 4 which isn’t really significant. Maybe the use of a mirror all around the sample holder could be a solution and the use of more than 1 LED could intensify the GFP excitation, which could induce a signal increase.

More tests have to be done and new conclusions will follow.

One thought on “Device 2 – Optic Filter Acquisition

  1. Pingback: Trip to Sion | biodesign for the real world

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s