nanoHUB-U Principles of Nanobiosensors/Lecture 1.3: Nanobiosensors Basic Concepts: Types of Biosensors
========================================
 >>

[Slide 1] Welcome. This is the third introductory lecture, discussing the essential aspect of nanobiosensors. If you recall, in the first lecture we talked about why nanobiosensors are so important that it can enable wide variety of applications. For example, in personalized medicine. Or in integration with mobile communication devices. Now, certainly nanobiosensors are important, but we have to first get introduced to the basic definitions so that we can follow along in more detailed discussion. In the last lecture we talked about various types of biomolecules. And we focused on three types. One was small molecule like glucose or nitric oxide that are indicators of variety of diseases, including diabetes and Parkinson disease, respectively. And then we talked about these long polymers, DNA and the amino acid polymer, which is a protein. And the invaders like viruses and bacteria. We also talked about that they diffuse around. Because they are sort of kicked around by water molecules as they try to go from one point to another. And a key element was diffusion distance. Let's continue on this discussion but this time bring in the sensors. 

[Slide 2] So the outline would be that I'll now try to explain to you that, once the biomolecules have landed on the sensor surface, how does the sensor respond? And then, of course, there are a variety that we'll see. And what we'll, one conclusion that I want to sort of come from this is that somehow geometry seems to be a fundamental variable. As things get smaller, you may often hear that the surface to volume ratio becomes larger. You know, when you have a large area, then the surface is essentially a small part of the total volume. Lots of molecules are inside. But, as you make things smaller and smaller, the surface reduces. The volume reduces. But the volume reduces faster. So most of the molecules can stay on the surface. So geometry being an important indicator of biosensors will not be a surprise. But the surprise will be that it is not the surface to volume ratio that's important. I'll explain. And then I'll summarize and conclude. 

[Slide 3] And we'll get started in the next set of lectures. So here is the basic picture; you remember the hypothetical big box which could be a beaker or it could be a droplet. Or it could be a cell. And we have the target biomolecules that we got introduced. And after long random walk, which is the diffusion, the molecule has finally landed on the sensor surface. Now, the sensor, this red sensor on the bottom, must be able to recognize this target molecule. There must be some aspect of it, whether it's the mass of the molecule. Or the charge of the molecule. Or the electron affinity. That it will be able to detect and then correspondingly convert it to something that we can measure. For example, current or voltage or displacement. Something like that. That's the method of transduction. Now, an important point I want to emphasize. That many sensors, an important element of the sensors is as I said, there is this mystery box. And, although I am just drawing it as a mystery box in this particular case, I'll return to this mystery box often. This could be a reference electrode for the experts who know, of various sorts. That makes the sensor work in a noisy environment. But for the time being, 

[Slide 4] let's focus on simple ideas of the sensor itself. So we have this box. Molecules diffusing around. Let's assume there are many molecules. And essentially they can be captured by the sensor surface. You ask yourself this question. That after a certain time T, once you have inserted the biosensors into the analyte concentration, how many molecules would the sensor catch? So, of course, that you can answer by writing a simple differential equation. The differential equation has two parts. The first part, shown here in blue, involves the capture process. So when the molecule comes in, these could be captured by the receptor molecules which recognizes the specific protein. For example, a, or a DNA. And it will be captured by this. Now, this capture is proportional to number of capture probes that are still available. Because if we have captured, if you have all the probes taken up, even if a molecule comes in, it will not be able to sort of land anywhere. So the difference, unoccupied number of receptors is necessary in order for this capture process to work. kF, you can view it as like a stickiness coefficient. That, if a molecule comes in and touches the receptor. If it immediately catches it and never lets go, then the kF will be very high. On the other hand, if it comes in and the molecule conformation is not exactly right and it lets go again without really catching it completely. Then the kF will be lower. So this is the catching term. And, of course, it depends on how many molecules you have close to the sensor surface. If you don't have any molecules, nobody's going to get caught. But this is the number increase in the captured molecule number. But, of course, what may happen that, let's say two molecules have conjugated. After a while, because the thermal vibration, the captured molecule is released. In that case the second term essentially dictates the release process. So, if you want to look at a given time how many molecules you have, this would be a balance between how many has been captured and how many has been released. And, if you solve this differential equation, assuming that there's so many particles so many biomolecules, that diffusion is not very important. In that case, you'd say the total number of particle captured through this capture equation is given by this exponential relationship. In the very beginning T equals zero. Of course, you have captured nothing. You just inserted the sensor. And after a long time, you have reached a steady state because the number getting captured and the number getting released are approximately in dynamic balance. And so, therefore, the total number is not changing. Now, this steady state number that you have is, of course, proportional to how many captured probes you had to begin with. How sticky the sensor surface is relative to this ability to release. So, of course, this would make sense. But one thing I want to emphasize in this expression is that NSS is a finite number. Steady state is a finite number. So, if you have a sensor surface, which is very sticky. And once the particle comes in, it never lets it go. In that case kF is infinity. If kF is infinity, meaning it's very fast, infinitely fast. Then, if this has to be finite, the only way it can be finite is that Rho S is equal to zero. Which means that close to the sensors surface there's no free molecule. Everybody who's coming in is immediately being caught. I will use this simplification often. Especially for DNA capture. So I wanted to show you here in, as a part of the capture equation. 

[Slide 5] So now the biomolecules have arrived on the sensor surface. How are you going to sort of signal that the conjugation has happened? So one way that it's often done is called this labeled approach. So I want to explain to you this first so that I can distinguish it from the approaches of electronic nanobiosensing that we'll be using. So the idea is that, if you have a sensor pixel, one of those small sensors. And you have all these receptors, molecules sitting there. And then these biomolecules are coming and one by one sort of attaching and making, having this conjugation event together. In that case, now specifically, for example, the receptors can be a DNA molecule already ATCG, remember the polymer molecules that we talked about in the previous lecture. And then the target molecule could also be another segment of this DNA molecule. And these two molecules, this molecule is diffusing around and eventually getting captured because A goes with T. C goes with G. And so, therefore, these molecules can get stuck together. We'll discuss those things in detail later. The bottom line is, once these two things have been captured, the question is how are you going to detect them? One way to detect them is to do the following. You see, instead of having these target molecules moving on its own, we attach an extra probe molecule, a label molecule. Like an optical tag. And this optical tag is generally inactive when it's moving around on its own. This is inactive. But, once the binding occurs, then this tag is activated. So then, if you shine light on it, then this tag will light up. Thereby indicating the conjugation process, that something has been captured. But you see what we have just done is used a secondary marker to indicate this conjugation event. We did not really rely on its intrinsic property of the biomolecule itself. Now, this is widely used in DNA microarray and DNA chip. And this is very sophisticated. But in this course we'll not be focusing on this type of label based approach. In electronic biosensing, we'll be focusing on something that relies on intrinsic property of the biomolecule themselves. 

[Slide 6] Let me explain. So there are three types of sensors that we'll be talking about. The first is called a potentiometric sensor. Consider a transistor just like in your iPhone. Let's say you take one of those transistors and before the biomolecule arrives on the gate of the sensor, close to the sensor, there is a certain current flowing from source to drain. If you don't know what a source and drain is, don't worry. We'll talk about that down the road when I really explain the details of this sensor. For the time being, assume that there is a current flow going, electrons going from source to drain in the absence of the biomolecule. Once the biomolecule conjugates on the sensor surface, caught by the target, by the receptor. Because it has charge, it will change the potential of the channel. And hence the name potentiometric sensor. And, therefore, the change in the potential will change the current. So, therefore, you can signal, we can signal the conjugation event through the change in the current. And a more sensitive sensor will essentially respond with a fewer particles or fewer biomolecules captured on the sensor surface. That's it for the potentiometric sensor. By the way, this is a reference gate. This is a mystery box. We'll come back and talk about that down the road when we really discuss the potentiometric sensors in detail. The second type of sensors that we'll be discussing in this course are called amperometric sensors. And the idea is that you have two electrodes, once again, in a fluid. And, generally, the current that flows from this electrode to the other one is relatively small. Because at a given condition, the current flow here is not significant. And the barrier to current flow is very high. Now, once the biomolecule comes in, it behaves like a catalyst. That it changes the ability for this electrode to collect current. And, therefore, this biomolecule, the arrival of the biomolecule, conjugation of the biomolecule is reflected in an enhanced current. And this would be called an amperometric sensor. So these electrodes have names like working electrode. Reference electrode. And counter electrode. But the bottom line is that it converts the chemical information to an information about the current. We'll be using this for glucose sensing, for example. Because glucose don't have any charge. Nor they are not massive enough to change the mass of the sensor beam itself. And the third type of sensor is based on cantilever. Any time, if you have gone to a supermarket, if you put something to weigh on a balance, it has a very similar principle. But here the idea is that, once the molecule lands on the cantilever its, think about the seesaw. If somebody big lands on the sensor surface, of course, the characteristics, the resonance frequency of the cantilever changes. And that can be translated. The change in the frequency can be an indicator of the mass. And the enhanced mass certainly is coming from the capture of the biomolecule. So these are the three types of sensors we'll be discussing in this course. All three, you see, uses intrinsic property of the biomolecule. Charge. Electron affinity. And mass. And doesn't indicate use a secondary level in order to indicate the capture process. 

[Slide 7] Now, an important thing to note here is that, if NS is the number of yellow molecules, biomolecules that have landed on the sensor surface. Now the sensitivity, you can define the sensitivity as the change in the measured quantity. For example, current here. Before and after capture. Divided by the original. So it's like a relative change. So, if you define sensitivity that way, then it turns out this is depending on how you operate the sensor is either proportional to the number of molecules. Or essentially logarithmicly proportional to the number of molecules that you have that has landed on the sensor surface. For amperometric sensors it, again, could be proportional or could be square root dependence. And for the mass similarly, you can have slightly different sensitivity depending on how you decide to operate the sensor. The bottom line I want to say that, although there are pre-factors, depending on the design of the sensor. There'd be factors up front for a given type of sensor going from potentiometric to amperometric to cantilever. They have very similar sensitivity. Once you have chosen the right sensor. For the analyte to be analyzed. Then you have, they have very similar sensitivity. it is as if you go to buy a camera, and all these camera have approximately similar mega pixels. So, therefore, you don't get significantly more from one approach to another. Once you have chosen appropriate analytes for specific, that can respond,that can be detected by the sensor. 

[Slide 8] So that brings us to a puzzle. You see, the puzzle is then in that case, why is it that the sum of the sensors can have this extraordinary sensitivity regardless of the modes of how the sensing occurred? For example, you can have nanomolar sensitivity. Attomolar or a picomolar sensitivity. Extremely sensitive nanoscale sensors. And it cannot be traced to individual modes of operation. Whether it's amperometric or potentiometric, it's not going to give you orders of magnitude change one way or other. As you can see here, the sensitivity is almost agnostic to the trans method of transduction. 

[Slide 9] Now, if you just take a guess and arrange things on a whim like this. So here on this axis, I have plotted the concentration. Millimolar. Micromolar. Nanomolar. All the way to attomolar. Three orders of low M concentration. And towards that end is most sensitive sensors. Towards this end, well, this is the traditional sensors. Now, if you now bring all the experimental results from the literature, you'll see something very strange. And the strange thing is that there looks like there's something to do with geometry. This orders of magnitude enhancement has something to do with geometry. Planar sensors are less sensitive compared to a nanowire sensor. This appears to be very sensitive and far better than everything else. The composite is somewhere in between. So there's something going on with the sensitivity, with the geometry. But what is it exactly? Now, if you ask somebody in the community that what could it be? The answer most of the time will be surface to volume ratio. You know, when you have a big object, only a few molecules are on the surface. Most of the molecules are inside. Volume is big. When you make things nanoscale most of the molecules could be on the surface. And the volume, you also have relatively fewer. So both surface and volume are decreasing, but volume is decreasing faster. And, therefore, you could conclude that that makes nanoscale sensors more sensitive. Not really. 

[Slide 10] Let me explain why. Let's take a cylindrical sensor. Take a cut and let's say the original radius is Rb, before the molecules are captured. And electrons are flowing through this tunnel, going from source to drain, one contact to another. Depending on how much velocity you put, how much voltage you put, that will determine it's velocity. So what will be the current? Very simple. You have the cross-section pi R squared. qND is the density of free carriers. So that gives you the total amount of free carriers. v is the velocity. That's proportional to the drain voltage mobility and other things. Let's not worry about it. That's the total current before a molecule has been captured. Now, once the red molecules have sort of surrounded the sensor, they have surrounded it all around. Let's assume this extreme case. As a result, if these molecules are charged, we are talking about potentiometric sensors here. Then what will happen that the molecules, the free carriers inside, they'll be repelled. And there'll be a region where it will be depleted of 

[inaudible] charges. This is the white region. So right now then, therefore, my current has been reduced because the cross-section has been choked off. What is the new current? The new current, after the capture, is proportional to Ra square, the radius after capture. Now, you can immediately make a connection between the two by realizing that the number of biomolecules that you capture. Remember the molecules we had floating around. And then they got, surrounded, they landed on the sensor surface. And they got conjugated. If you, the number that says Nbio. And 2 pi R is the radius. The R here, you can think about Ra is the, I'm sorry, Rb, which is the radius before the depletion has occurred. The geometric radius. Now, that must be equal, that charge must be equal to the amount of charge that has been depleted. Because the charge must balance. So the red and white should be equal in area. And this is the statement of that fact. That the difference 

[inaudible] is equal in both cases. If you use this information, calculate the sensitivity as a relative change in the current. What you immediately see that sensitivity goes inversely with R, the radius of the nanowire. Which is good because, of course, nanowires are more sensitive. The lower the analyte density, lower the doping density is more sensitive. Why? Because that means the initial current was small. So, therefore, any change in the current would be greatly magnified. Delta is bigger. And Ia, I'm sorry, Ib was actually smaller. And, of course, if you catch more biomolecule, your sensitivity will be more. That's no-brainer. The point I want to make that it does explain the sensitivity gain. 

[Slide 11] But the gain is not significant enough. If you calculate these numbers. Plot it out as a function of the diameter, of the nanowire sensor. And delta I over, delta G over G, which is the sensitivity. Or delta I over I naught. If you plot it out, what you will see that for volt sensors, classical sensors, you have a certain sensitivity. And, as you make the nanowire sensor smaller and smaller and smaller  and smaller. And choking the current off. The sensitivity does go up. But you see it goes up by a factor of 5 or 10. No more than that. Still, in the previous case, we saw there were about 15 orders of magnitude enhancement in the sensitivity. Certainly this is important. But this is not the only explanation. Not even the major explanation about what geometry does to the sensor surface, to the sensitivity of a sensor. And, in fact, this is a key argument that is, 

[Slide 12] I hope you will remember as we go to the next lectures. So the important point I want to make is that the geometry of electrostatics, how the conduction gets choked off is important. But the most important factor, and the surface to volume is important as a factor here. But the most important factor is not this aspect of the geometry. But actually a second aspect of the geometry, which has to do with how a jellyfish catches the food from its surrounding. It is the characteristics of the surface itself, not the surface to volume ratio. Which, as I'll explain later, in a later lecture, that allows this to be far more sensitive than anything else that are around. So the reason I'll show that this is, nanowire sensor is more sensitive than a planar sensor is not because of the surface to volume ratio. But rather you can view this nanowire sensor as if it is a jellyfish in disguise. The physics of particle capture, biomolecule capture and the physics food capture by the jellyfish are identical. And this is how from nature actually nanoscale sensors get the enhanced sensitivity, by mimicking nature. And it has to do with the geometry of diffusion. 

[Slide 13] So let me conclude. I told you about three types of sensors that we'll be talking about. Potentiometric. Amperometric. And cantilever-based sensors. This doesn't use any secondary level. No optical levels to worry about. Less expensive. And could be faster in some ways. And it focuses on the intrinsic property of the biomolecules. Charge. Mass. Electron affinity. And so on and so forth. Now, neither the method of transduction, it doesn't really matter how you detect it. Remember the mega pixel example. They have about similar "mega pixels." Nor the geometry of electrostatics allows to you explain the extraordinary gain in sensitivity. There is something else. And that something else, as I mentioned is biomimetic. In fact, that gives you this extraordinary sensitivity gain. But you'll have to wait till next lecture in order to get started on that story. So, until next time, take care.