Essential Thermodynamic Concepts for Biochemists Neal Woodbury Arizona State University Thermodynamics is a very powerful formalism. It is powerful because it applies to everything. The thing to remember about thermodynamic formalism is that it has nothing to do with chemistry. As you will see, all the real chemistry is contained in the empirical constants and fudge factors (if you want chemistry, you need to learn about quantum mechanics and electrostatics). Thermodynamic formalism is about populations. It really does not matter whether you are talking about populations of molecules or chickens, the principles apply equally well (assuming you have something like a mole of chickens so that the statistical fluctuations average out). Thermodynamic formalism is derived for ideal gases (collections of particles that do not interact in any way except that they run into each other elastically – that’s not chemistry). The application of thermodynamics to everything else simply involves adding the aforementioned empirical constants and fudge factors. So lets start at the beginning. In thermodynamics we make two fundamental assumptions: 1) The total amount of energy in the Universe does not change (OK, mass/energy if you want to get technical, but biochemistry does not involve any thermonuclear reactions). 2) The amount of disorder in the Universe is constantly increasing and can never decrease. Another way of saying this is that things spread themselves around in a random fashion if you shake them up a bit, and if you have a large number of things, they are never going to become any more organized. The more you shake, the more disorganized (random in space and time) the things will become. Sometimes people refer to the definition of temperature as another “Law”. Basically temperature is random motion of molecules and is proportional to kinetic energy. The proportionality constant is kB, the Boltzman constant. At some very low temperature, motion entirely stops (this is absolute zero in degrees Kelvin). Now we need to convert these fairly obvious ideas into some kind of mathematical formalism. This is all thermodynamics is, actually, a mathematical formalism for these two concepts that can be applied to the world to make quantitative predictions. For the first Law we have to decide what the physical manifestations of energy are and how we will measure them. Actually, we never measure absolute energy (except maybe temperature). We only measure changes in energy. The three kinds of energy changes that are important to biology are heat (called q), mechanical work (called w) and electrical work (another kind of work, w). When a system gives off heat, its total energy decreases (but the total energy of the surroundings increases so the total energy of everything is constant). When it takes up heat, its energy increases. When a system does mechanical work on the surroundings (pushes something against a force), its energy decreases. When the surroundings do work on the system (compression, of the system for example), the system’s energy increases. Electrical work is like mechanical work except that we are moving a charge either against or along an electrical potential (for example, it takes work to move two positive charges closer together). We normally measure heat through changes in temperature (though the system can give off or pick up heat without changing temperature if other things happen at the same time). We normally measure mechanical or electrical work by measuring the force that we are pushing against and the distance we push against it. We can now start writing these ideas down in mathematical form: dq C p dT dw fdx PdV dw zFdE These are the differential forms of the equations defining the infinitesimally small changes. An incremental amount of heat (dq) is given in terms of the heat capacity (Cp, the heat energy something can absorb for every degree in temperature that it increases) and an infinitesimally small change in temperature (dT). An increment of mechanical work (dw) is just the force (f) times the distance moved (dx) which if put in terms of three dimensions is the pressure (P) times the change in volume (dV). You can think of this in terms of the work required to blow up a balloon. Finally electrical work is just the energy required to move a particle of charge z through a tiny voltage dE (F is the Faraday constant and just makes the units come out right). So, we are now ready to state the first law in terms of the energy of our system: dUsys = dq + dw Usys is the energy of the system, often called the internal energy. Any change in U is either because heat went in or out or because work was done on or by the system. The change in Utot for the universe is zero (that is actually the first law). The first law is a statement setting some limitations about what can happen – you cannot do things which change the total energy of the universe. However, it is far from the whole story. I can drop an egg on the floor. It breaks into a thousand pieces and makes a big mess. However, the energy of the Universe did not change. The energy of the falling egg was converted into random molecular motion when it hit the floor (heat). Furthermore, the energy of the Universe would not change if exactly the reverse reaction occurred. If the energy lost as heat when the egg hit the floor jumped back into the egg, causing it to come back together and fly up into my hand. However, we all know that the first process (dropping the egg and smashing it) happens, but the second process (having it spontaneously reassemble and fly back up into my hand) does not. Neither process violates the first law of thermodynamics. What are we missing here? Remember that the second law says that if something is going to happen, it must result in the Universe moving towards an overall state of greater disorder. Well that pretty much solves the egg problem. The egg got pretty disordered when it hit the ground and it is not going to reorder itself on its own. Further, if the universe gets more ordered in one place, it must get even more disordered in another. I could, in principle, put the egg back together, but I would have to expend considerable energy doing it. That means I would have to burn up food which will make the atoms in the food much more disordered. In the end, I will have made the Universe less ordered by putting the egg back together (you can try this argument out on your Mother when she tells you to clean up your room). OK, so how are we going to state all this in quantitative terms. We need some measure of order or disorder. From the egg example, we have some idea that the problem with the egg putting itself back together is that the energy that resulted in it becoming disordered (the energy of its motion when it hit the flow) was turned into the random motion of heat. So heat lost has something to do with it. But it also has something to do with temperature. Basically, there would not be as much of a change in the amount of random energy in the system if there was already a lot of random energy to begin with. It other words, the higher the temperature, the smaller the change in disorder because at high temperatures there is already a lot of random motion. We therefore measure the change in disorder as something called entropy (S), which is a ratio of the heat lost (actually, the heat lost reversibly, but let’s not worry about that just now) to the temperature: dS = dqrev/T The second law states that for any process that will actually happen: S tot 0 In other words, the disorder of the Universe must increase or the process will not take place. Now in chemistry (or biochemistry), it is going to be a lot easier to think of things in terms of energy than in terms of entropy. Notice that if you multiply entropy by temperature, it has units of energy (heat). So it might be easier to think about the second law of thermodynamics in terms of the quantity TdS, or if we consider a macroscopic change, TS. T is always a positive number if we deal with degrees Kelvin, so we can say that TS tot 0 or TS tot 0 We give the energy -TStot a special name. We call this energy the free energy change of the process and usually designate it G. Remember that a free energy change associated with a process is always a free energy change for the entire universe. It is not the free energy change of just the molecules involved in the process. It includes the effect of heat or work done on the surroundings as well. This means that another expression of the second law of thermodynamics is: G 0 OK so based on the above discussion, we can see that G = -TStot We can now split this up in terms of the entropy change for the system and that for the surroundings: G = -TSsur -TSsys What constitutes the system and the surroundings? Basically, the system is the molecules involved in our reaction or process and the surroundings is the rest of the universe. Now we have to come back to the definition of entropy change in terms of heat and temperature. As long as we agree to keep the temperature of the surroundings constant, we can arrange things so that heat is transferred to it or from it in a reversible fashion. This concept of reversible is kind of tricky, but for heat transfer it basically means that the temperature of the thing under consideration does not change (we have not changed, in some sense, the ability of the surroundings to transfer heat one way or the other). For the surroundings, this is certainly true (the surrounding are infinitely large compared to the system). Thus: G = -Tqsur/T -TSsys = -qsur -TS The heat transferred TO the surroundings is just the negative of the heat transferred FROM the system. Thus we can write: G = qsys -TS Biochemistry, thank goodness, occurs at constant temperature and pressure (this just saved us many pages of grief). If the pressure is held constant, we have another name for the heat generated or consumed by the system. We call it the enthalpy change and give it the symbol “H”. Thus we can write an equation which should start to look pretty familiar to you: G = H -TS OK, this is conceptually useful, but most of the practical applications of thermodynamics in biochemistry rely on understanding its relationship to concentrations of chemicals in reactions. How are we going to generate a mathematical formalism for this? This is where what I said in the very first paragraph becomes important. Basically, the free energy associated with a chemical reaction under some particular set of conditions is an empirical constant. You cannot calculate it using thermodynamics. You can maybe calculate it using quantum mechanics and electrostatics, but only for very small molecules. In biochemistry, essentially everything of importance in this regard is empirical. Great. We just went through four pages of this, and NOW I tell you that you just have to look the value of the free energy up in a book anyway. What’s the point? The point is that you only have to look the value up in a book for one set of conditions. In principle (with a few occasional empirical fudge factors that we will discuss later thrown in), you can use thermodynamics to calculate the free energy under all other sets of conditions. This brings us to the concept of the standard state. Basically, solution chemists decided a long time ago to tabulate reactions for the state where everything was at about room temperature and all the concentrations of the reactants and products were at 1 M, except for reactions were the solvent is one of the reactants, in which case the standard concentration of the solvent molecule in the reaction is the concentration of the pure solvent. Thus, in essentially all biochemical reactions, water is always at its standard state (which will mean we can ignore it in our thermodynamic calculations for the most part). You should also note that in biochemistry, unlike chemistry, the standard state of H+ is 10-7 M (pH 7.0) rather than 1 M. In addition, free energies are reported on a per mole basis. That is, the free energies we will deal with will actually have units of energy/mole (the amount of free energy change when one mole of reactant goes to one mole of product). OK, now you are saying, hold on a second Dr. W., you are not going to tell me that you are going to run a reaction on a protein in a solution with 1M protein!! After all that could be hundreds of kilograms of protein per liter!! Not possible. The answer is that you do not have to actually ever run the reaction at standard state. The free energies are measured at some reasonable concentration and then adjusted as described below until the free energy is what would be expected if you could run the reaction at standard state. That value is what is reported in the table. So here is the way to think about the free energy change at standard state. This is important if you want to understand what is going on. This is the energy of one mole of reactants being consumed to form products under conditions where NOTHING changes including the concentrations of the reactants and products. This is possible, after all, if there is a very large amount of reactant and product to begin with. For example, a thousand liters of 1M everything will not change concentration appreciable if you convert one mole of reactant to product (there was a thousand moles of reactant in the one thousand liters at 1M after all). We call this standard value of the free energy change G0. You look this up for the reaction of choice in a table somewhere. I have not really shown you this, but free energy is a state function. This means that the energy change between the beginning of a process and the end of a process is independent of how you get there. So, if we want to know the free energy of a reaction that is not at the standard state, you can do this in three steps: 1) Determine the free energy change involved in changing the concentration of the one mole of reactant that is going to react from the real concentration to the standard state concentration. 2) Run the reaction at standard state giving a molar free energy change of G0 (molar means per mole). 3) Determine the free energy change involved in changing the concentration of the one mole of product from the standard state concentration back to the real concentration. So what is the free energy associated with changing concentrations of a compound? This is entirely an entropy term. The more spread out something is (lower concentration) the greater the entropy. This has nothing to do with the reaction itself, which is all tied up in the G0 term. Yes, OK, we will have to ultimately consider the attraction between molecules in solutions which is not entirely entropy, but hold off on that – we will invent another fudge factor to take care of it in a minute. Actually, in biology, most things are present at low concentrations so that this is not as large an effect as it might otherwise be. First lets consider particles that do not interact with one another. This is an ideal gas. Forget for the moment that we do not run enzymatic reactions in the gas phase. From the point of view of thermodynamics, it really does not matter. We can concentrate molecules closer together and increase their concentration or let them be farther apart and decrease their concentration. In biology, molecules are often at pretty low concentrations and surrounded by a homogenous medium (water) making them act in a sense as though they were in the gas phase (separated and not interacting with each other). What does it take to push ideal gas molecules closer together? Now we need that first law stuff we wrote out above. The only kind of energy that an ideal gas has is kinetic energy. Movement. Temperature. That’s it, there is nothing else these inert particles are allowed to do but move. This means that if we hold the temperature constant while we change the concentrations of the molecules in the reaction, there will be no change in internal energy. That means that w = -q. Also, because we are holding the temperature constant during the process of changing concentration, the heat is reversible. So remember we said that this was entirely entropy thus: Gcon. change = -TSconc. change = -Tqrev/T = -qrev = w What is the work associated with compressing a gas at constant temperature? Remember that: dw = -PdV To get w we just integrate between the uncompressed and the compressed state (this is the only real calculus in this whole document, promise!). But we have a small problem. As we change V, P is going to change as well. Fortunately, we know that for an ideal gas, pressure and volume are inversely proportional at constant temperature (ideal gas law, remember?): PV=nRT Or P = nRT/V Substituting we see that: f Vf nRT dV G w PdV dV nRT nRT ln V V Vi Vi Vi Vi Vf Vf V Here, Vf and Vi are the final and initial volumes, respectively. Since we are doing all this on a molar basis (the free energy change is for one mole), n = 1. Also, we are not losing or gaining material, we are just making it more or less concentrated. Thus the volume that the material is held in is just inversely proportional to the concentration. In other words: Vf/Vi = Ci/Cf So we can see that the free energy for changing the concentration of a molecule is given by: C G RT ln i C f C RT ln f C i Remember ln(a/b) = -ln(b/a). We are almost there. Now we have to actually do the three steps we talked about above. This is most easily done using an example. Let’s consider the reaction: A + B C The initial concentrations of A, B and C are just CA, CB and CC. Now lets do the three steps: 1) Change the concentration of one mole of A and one mole of B from their current concentrations to 1M. So the free energy to change A from CA to 1M is just: 1M C RT ln A RT ln A G A RT ln 1M CA So what’s with [A] all of a sudden? Well we use that nomenclature all the time in chemistry as though it was the concentration of a chemical. Indeed it has the same value as the concentration in moles per liter does, but it is actually unitless. It is a ratio of the concentration in moles per liter divided by 1 mole per liter. It has to be unitless or all those equations where we take logs of concentration would never work. You can’t take the log of a unit. So [A] means the concentration of A relative to a solution of one molar A. It is now pretty easy to calculate the free energy change for changing the concentration of B: GB RT ln B 2) Next we need the free energy of the chemical reaction itself under standard conditions. This of course is just G0. 3) Finally we need to take the products formed (in this case C) and change them back from 1M to the true concentration, CC. This is the same as above but the opposite sign: C GC RT ln C RT ln CC 1M The signs are reversed because in this case the initial concentration was 1M and the final concentration was Cc (the ratio was inverted). Now, finally, we just add up the free energies for the three steps: Grxn = GA + GB + G0 + GC or C Grxn RT ln [ A] RT ln B G 0 RT ln C G 0 RT ln AB Now this should look familiar. If you are confused about the manipulation of logs, remember that ln(a) + ln(b) = ln(ab) and ln(1/a) = -ln(a). Anyway, we can generalize this equation as: Grxn G 0 RT ln Q Where Q is just the quotent of all the relative concentrations of products multiplied together divided by all the relative concentrations of the reactants multiplied together. This is the master equation we use all the time in biochemistry. Note that if we consider equilibrium, which the situation where the free energy change in the reaction is zero (it does not spontaneously go in either direction), we have: 0 = G0 + RTln(Keq) or G0 = -RTln(Keq) Where Keq is just the ratio Q when the system has come to equilibrium. Note that this means that there is a fixed ratio of concentrations for a reaction at equilibrium that is independent of how much of any reactants or products you started with. This ratio is called the equilibrium constant. Note that it does depend on temperature. Fine, fine, but hemoglobin is not an ideal gas! Well that’s true, but first, as long as it does not see too many other hemoglobin molecules and just floats around by itself in water, it is not that different from an ideal gas from the point of view of changing concentrations. Second, to the extent that attractions between molecules do occur, we realize that this makes it easier to push molecules together than you would otherwise predict. In other words, the molecules act like they are at a lower concentration than they really are. No problem, we will just make up a fudge factor (normally less than one, but not always), called an activity coefficient, and multiply our hemoglobin concentration by that. Now we have an effective concentration that does, more or less, obey the rules. We can come up (empirically) with fudge factors like this for all of the molecules we work with. We refer to this effective concentration as an activity. OK what can we do with this? The above equations are themselves useful for general biochemical reactions. Let’s consider a few useful special cases. We deal a lot with pH equations, membrane potentials and redox equations in biochemistry. They are all just this equation in disguise. Lets work through them. First redox equations: Redox reactions are just moving electrons around. This is electrical work and can be thought of in terms of the energy for moving a charge with or against a voltage. We stated earlier that the energy for doing this was just equal to -zFE, where z is the charge (the number of electrons transfer during the reaction), F is the faraday constant for getting the units right and E is the voltage you are moving the charge through. Thus to convert from a free energy to a voltage you just divide the free energy by -zF: G/zF = -G0/zF + -(RT/zF)ln(Q) or E = E0 - (RT/zF)lnQ For a reaction between a reduced material going to an oxidized material, Q is just ([Red]/[Ox]). This gives the familiar redox equation: E = E0 - (RT/zF)ln([Red]/[Ox]) Now this is actually the voltage for a half reaction. If something is going to be oxidized, after all, something else must get reduced in the process. But you can just add the potentials of half reactions together to get complete reactions. How about membrane transport? Here we have one type of molecule moving across a membrane. The reaction is Ain Aout We can write as usual: G = G0 + RTln([Aout]/[Ain]) If there is no voltage across the membrane (usually there is) or if A is not a charged molecule (and thus does not feel the membrane voltage), then at standard state when there is the same amount of A on the inside and the outside, the system is at equilibrium. Equilibrium means that there is not tendency for the reaction to go one way or the other. In other words the free energy is zero. If G is zero and if Ain = Aout (which results in the log being zero) then G0 for this reaction must be zero. This makes sense; we would expect that at the standard state when the inside and outside concentrations were the same that the free energy of the reaction should be zero. Plugging this in gives the following for transport of molecules across membranes if the molecules are uncharged of the membrane has not potential: G = RTln([Aout]/[Ain]) This “reaction” is ALL entropy. There is no real chemical reaction taking place, you are just moving molecules from one place to another. If you make the inside concentration high, the free energy change is negative and flow outward is spontaneous. If you make the outside concentration high, the free energy change is positive and flow inward is spontaneous. But what if we are dealing with a charged molecule and there is a voltage across the membrane? Then the free energy for transfer of a mole of charge across the membrane is not zero any more (G0 is not zero). It will depend on the membrane voltage as we discussed before. People usually refer to membrane voltages as So the energy of transferring a charge z across a membrane when the concentrations on the two sides are the same is just zF. The complete equation including both the effect of different concentrations on the two sides of the membrane (chemical gradients) and a membrane potential across the membrane is: G = zF + RTln([Aout]/[Ain]) Note that now at equilibrium, if the membrane potential is not zero, the concentration gradient of a charged molecule will not be zero. At equilibrium G = 0 so: RTln([Aout]/[Ain]) = zF This equation is important for calculating the ion gradient at equilibrium from a membrane potential or the other way around. For a more complete description of membrane potential calculations and conventions concerning the assignment of the sign of the membrane potential and the direction of the ion transfer see the Membrane Potential lecture notes. Another equilibrium equation of great importance comes from considering the acid base reaction: A + H+ AH+ G = G0 + RTln([AH+]/[A][H+]) If we let this system come to equilibrium (G = 0) we have = G0 + RTln([AH+]/[A][H+]) Now lets divide by RT = G0/RT + ln([AH+]/[A][H+]) Convert the ln into log10 0 = G0/RT + 2.303log([AH+]/[A][H+]) divide by 2.303 0 = G0/2.303RT + log([AH+]/[A][H+]) split the log up 0 = G0/2.303RT - log([H+]) + log([AH+]/[A]) We call –log([H+]) pH. Rearranging we have (careful about minus signs) pH = -G0/2.303RT + log([A]/[AH+]) Remember that G0 = -RTln(Keq). The equilibrium constant for this reaction is usually called KA and the negative log of this equilibrium constant is called pKA. I think you can see that pKA = -G0/2.303RT if you rearrange things. This gives the familiar equation: pH = pKA + log([A]/[AH+]) These are the fundamental thermodynamic equations that we use in Biochemistry and a description of their origins. None of this involves high level math. The concepts associated with entropy and free energy take some thinking about, but if you consider them in the terms presented here, I hope it will all start to come together.