Capillary Electrophoresis

In this instructable, I made a simple capillary electrophoresis setup with capacitively coupled electrode detection method that is simple and cheap (w/o multimeter and power supply, things that I already had, the whole thing costs less than $30). This can be used to analyze biological fluids like urine, blood, food like wine, or to analyze DNA etc. For DNA analysis, which is basically determining the length of particular strand, I will try to reproduce an article "Determination of PCR products by CE with contactless conductivity detection "appearing in J. Sep. Sci. 2012,35, 3009-3513, however my main interest is in making CE system that does not use "external" buffers. I am novice, in fact, never used CE before, so I welcome any suggestions.

Capillary electrophoresis (CE) is an analytical (bio)chemistry technique that can separate or identify ions (or with some modifications neutral species) in a liquid. You can look at wikipedia page for details and general principle of CE.

In addition to capillary itself, capillary electrophoresis setup has 3 major parts:

1. Sample injection

2. Detection system, which can be based on optical detection (typically UV light absorbance of liquid inside the capillary) or electronic detection. Here I used later, capacitively coupled electrode detection because it is very easy and cost-effective.

3. High voltage (be very careful with this) with electrodes.

Sample injection can be hydrodynamic (i.e. using over or under pressure) or electrokinetic (using voltage). Here I used hydrodynamic one. Electrokinetic injection will result in uneven injection because it relies on same principle like CE, but sometimes that can be good. For me, this was also the cheapest and easiest (I had already had tubes and syringes).

Parts:

1. 2 eppendorf tubes

2. Syringe

3. Silicone

I made a hole using syringe needle in the side of eppendorf tube, passed one end of capillary and pushed it towards the bottom of the tube. Next, I put silicone around the hole (both inside and outside) to plug the hole and wait overnght so it is completely cured. This eppendorf tube will be - electrode (anode) reservoir (Fig. 2). The + cathode reservoir eppendorf tube is not modified.

To create over of under pressure I insert the syringe body (without the needle) into anode reservoir and push or pull the plunger, respectively. Amazingly, dimensions of eppendorf tube and my syringe match perfectly to create a nice seal, so basically, eppendorf tube acts as an extension to a syringe body and plunger can serve to control the pressure inside it. This enables me to easily control the liquid flow in, or out of the capillary. You can see the syringe parts (whole syringe and needle separately), and injection system with syringe body fitting in the capillary modified eppendorf tube (Fig. 3).

Detection system is based on capacitively coupled electrodes. Detection system consists of oscillator (basically a wire with AC voltage - emitter), and detector (wire that acts as AC receiver). For both of these, I would connect the output (for oscillator) and input (for detector) to the capillary. I basically cut two stripes of conductive carbon tape and taped them over capillary and each connected wire to make electrical connection (Fig. 5). Separation between oscillator and detector carbon tape should be small (~ 1 mm) and they are on the opposite sides to minimize cross-talk noise (if interested I can point you to papers where they discuss electrode dimensions to capacitive coupling strength).

1. Oscillator source (Fig. 6). You need an AC (sine, square etc) voltage fo 10kHz-1000KHz. I used a ebay bought circuit that I got for $5. It makes a square voltage output with several frequency ranges. The best signal frequency depends on carbon tape electrodes' dimensions. I used frequencies of ~100 kHz.

2. Detection system consists of receiving, carbon tape electrode, and transimpedance amplifier (Fig. 4). Transimpedance amplifier is shown on the scheme and it is slightly modified circuit that is published in Anal. Chem. 1998, 70, 4339-4343 (they did great job). Electronics components which include operational amplifiers, resistors, capacitors, and board, I purchased my from digikey and ebay, and they cost ~$30. Operational amplifiers used in detector require power supply of +/-15V and ground. I had one available so I just used that one. Potentiometers (two resistors that can change resistance) are used to adjust base voltage on last operational amplifier close to zero, so that only useful signal is amplified. Make sure that none of the amplifiers are saturated (they show ~ +/- 15 V on output, pin 6 in that case).

3. Some type of interface to a computer. You can read the output voltage with any of arduino, IOIO or other boards. I used Radioshack multimeter with serial connection and their voltage recording software.

I am aware that the transimpedance amplifier circuit should be soldered, and that I should have used short wires to reduce noise. Also, I should add Faraday cage over the carbon tape electrodes and circuit. I didn't solder circuit output, just used alligator clip to voltmeter.

Warning: high voltage is dangerous. Please make sure you understand your circuit and take precautions. Current limiting helps.

I used simple and cheap, step-up (boost) voltage HV source from ebay that is nominally giving quasi DC of 20 000 V (less than $5). It requires DC voltage input of 3-6 V, and has small current. Due to small diameter of capillary, current required for CE is small (it does depend on buffer concentration.). Initially, I used 2 D cell batteries (total 3V) as input, but when I became comfortable working with this high voltage, I switched input to a simple 3 V AC/DC converter (from 110 V AC). Since you need to know polarity of input to avoid killing the HV step-up circuit, make sure you measure voltage from power supply and you get +3V (not -3V). Also, one needs to determine output polarity since output wires are not labeled. Most voltmeters can't handle voltage this high, but you can use a simple serial connections between resistors, and measure voltage on one serial resistor with small resistance (lookup Ohm's law).

In the metal box that acts as a Faraday cage, I have 2 step-up sources (one for 20, 000V and other for 7, 000 V) that can be used. That is why you can see extra wires.

HV source is lead to cathode (+) and anode (-) electrode. The fluid should flow from cathode reservoir to anode electrode reservoir (-). This is determined by the charge on the inside wall of the capillary.

Electrodes:

Typically people use platinum wire for electrodes in reservoirs. The reason for this is chemical stability (you don't want your electrode to change composition or "leak" into reservoirs. I used needles from syringes that I had around. They are made out of stainless steel and I haven't noticed any degradation during the use. I soldered the connecting wire that leads from + and - wires of HV source to needles. Also I used silicone (one that is used for caulking) to close the hole in the syringe needle and cover the soldered part. The only part that is in touch with the liquid is stainless steed needle outside surface.

So you need silica capillary. Typically these are 25-75 microns inner diameter and they have plastic coating that makes them flexible and less breakable. The great thing is that the capillary coating does not need to be removed. The capillary used has inner diameter of 75 microns and length of about one foot. In the Figure 1, there is also another capillary of the same diameter and shorter length (5 cm) that I use when I need higher electric field.

For this kind of setup, ideal capillary diameter would be 50 microns ID. Also, longer capillary would give higher resolution, people typically use half a meter lengths.

You need to condition the silica capillary. This means to establish a nice, uniform, negatively charged layer on its surface. This is done by passing NaOH solution (stronger, the better). I used 0.1 M NaOH, it works (there is electroosmotic flow) but possibly it could be better if you did 1 M NaOH. Just fill the reservoir with NaOH and press the syringe to flow it trough the capillary to make sure it completely filled inside, fill both reservoirs with base and leave it overnight. Also, sometimes it is necessary to re-condition the capillary wall after the run because some samples adsorb during the runs. For this, people typically use shorter time (not overnight) and weaker concentration (0.1M).

You need to confirm that capillary is open and that detection system works as supposed. I typically inject salt solution and pass it through the capillary hydrodynamically. One should see a peak in recorded voltage vs time plot. In this case, no buffer is in the reservoirs, just distilled water, and sample injected is 150 mM NaCl.

1. Condition the capillary if necessary. Keep the anode outside the reservoirs.

2. Insert the buffer in the anode reservoir. Push the plunger and make sure buffer flows for some time (I count to 50).

3. Fill the cathode reservoir with buffer (if it was not filled previously).

4. While still having the pressure in the anode reservoir (syringe plunger in), put the cathode part of capillary up (I used height of ~5 inches) and into the sample liquid (I used pipette tip to hold sample liquid, very convenient). Next, I release the pressure, i.e. take the syringe body out. This is when the sample starts to be injected. When sample and buffer have similar composition, I just leave the gravity to inject the sample in. If the sample is less dense, I will use underpressure (pull the plunger out for 5s-30s) to force the sample flow into the capillary.

5. Put the electrodes back into eppendorf reservoirs. Turn on the HV.

6. Turn on the power supply (+/- 15 V) for transimpedance amplifier circuit.

7. Turn on the oscillator. This order is to prevent the burn of amplifier and oscillator circuits due to HV spike. When they are not powered, there is a less chance of damage.

8. Start recording.

9. After the run is done, I turn off circuits in reverse order (oscillator, amplifier power supply, and then HV).

10. If you are done measuring for the day, rinse the capillary with distilled water and leave distilled water in both reservoirs. If buffer/sample with high ionic liquids dries inside, it will clog the capillary.

In a picture above you can see what kind of results one can get in these experiments (this one is done with lower HV voltage that in the current configuration but type of plot is the same). The sample was urine in saline sodium citrate -SSC buffer. What we are evaluating is: how many peaks we can see,and their intensity (maximum voltage). CE has serious drawbacks that it doesn't give info what is the kind of molecule/ion gives the peak. For this typically pharma uses mass spectroscopy (give you the molecular weight of the molecule) after CE or other molecular separation techniques. I guess the best use of CE would be monitoring i.e. if you routinely do the measurements of the same thing (let's say urine or saliva) and look out for abnormal peaks. Another one would be evaluation of DNA length when DNA ladder comparison can be made. Unfortunately, one needs a buffer at appropriate pH when you want nice separation, and also large polymers that act as a sieve when you are doing DNA separation. I would love to see protocol that can use only electronics without any consumable chemicals. If you have an idea or paper about this please let me know.

I want to stress that one should not judge the capability of CE based on this setup. This can be done much better (bear in mind that I never even used CE before making this), however I like that this cheap toy setup can be used to play with many different protocols/samples so I think it is perfect for DIY community.