In fact, this approach (with somewhat more strict mathematics) was applied by Van 't Hoff at the end of the XIX century to create the equation describing the temperature dependence of the reaction rate. He obtained equation (d(lnk)/dT)V=E/RT2+B (note that this B is different from yours) and a few years later Arrhenius developed his famous equation that is essentially the same but gives some physical meaning for the parameters E and B. You can find details in any good textbook on the chemical kinetics.

So your idea is not wrong, but all useful information that can be extracted from it was already found long ago.

And what about the transition state theory? Can be the B chosen as B = - Gt + R*T*ln(kB*T/h), where Gt is Gibb's energy of transition state of the reaction, kB is Boltzman constant, T temperature and h is the Planck constant? 

It seems to me that the bottom line, thermodynamically, is that the Gibb's free energy differences of the substrates can only give you the net driving force, not the kinetics, i.e., not the rates. For instance, introduce an enzyme, and the rate will increase dramatically without changing the net driving force for the reaction. 

Yes, thank you. It starts to be clear for me.

The B is not a constant, it is dependent on the transition state of the reaction, which is together with substrates "the driving force of the forward kinetics" and together with products "the driving force of the backward kinetics". As a result the reaction rate coefficient is as in the transition state theory: kf = (kB*T)/h *exp((Gs-Gt)/(R*T)), where Gs is Gibb's energy of substrates, Gt is Gibb's energy of transition state, kB is Boltzman constant, T temperature and h is the Planck constant, R gas constant. 

The enzyme creates different chemical path, which is faster than the original reaction, because of lower Gibb's energy changes from substrates to transition states. And the equilibrium of both paths (the orginal reaction and the enzymatic reaction) will be the same because of the principle of detailed balance.

Yes. This now has the ring of truth.

For your proposal to work, the reaction must follow microscopic reversibility. Only elemental reaction steps do so. Therefore, your proposal only applies to elemental reaction steps. It absolutely does not apply to multi-step reactions by principle. When it seems to apply due to experiments, it is only by good fortune or serendipity.

Adding an enzyme to a reaction system changes the overall mechanism. Even when you would perhaps be considering elemental steps, you cannot compare apples (a reaction without an enzyme) and oranges (a reaction with an enzyme) as you seem to want to do.

Of course, as "enzymatic reaction" I mean the elementary reactions A+EC and CB+E, where E is an enzyme, C is enzyme-substrate complex, and AB is the original slow reaction without enzyme. As in Michaelis-Menten kinetics but with both reaction reversible. I hope, that the mentioned calculation of forward rate coefficients is applicable on each of these elementary reactions in both directions.

What I think is that it is not possible at all to find the reaction rate from thermodynamic quantities.

An example:

Two reactions with the same DG do not have the same velocity.

Consider the N2 dissociation. It happens with the following reaction

N2+X=2N+X.

Now the rate coefficient depends on X. You can find many references in literature about these rates.

Another example, Methane combustion. At room temperature the rate is 0. DG is high.

Thermodynamic quantities tell you how far you are from the equilibrium, not if the equilibrium state can be reached and, if yes, at what speed.

The Arrhenius law includes some dynamical parameters such as activation energy, that takes into account the height  of the barrier (not a thermodynamic function) and a power constant to consider the slope of the potential energy surface.

Activation energy of the reaction is Gt-Gs, where Gs is the Gibbs energy of the substrates, Gt is the Gibbs energy of transition state of the reaction.

@M Matejak: "Of course, as "enzymatic reaction" I mean the elementary reactions A+EC and CB+E ..."

Assume both reaction are elemental. This does not mean that A B is elemental. Also, the first two reactions use an enzyme, while the third does not. Apples and oranges.

In conclusion, you absolutely cannot analyze the thermodynamics of the first two reactions to pull out the kinetics of the third reaction.

Hmm, the goal is to make a support for integration of known user defined chemical systems. Our software is called Physiolibrary. If you want to build more complex chemical systems using Physiolibrary then install the last release of OpenModelica, where is the Physiolibrary accessible as system library.

The last Physiolibrary 2.3 still has dissociation constant and forward reaction rate as parameters of each chemical reaction. But may be the next release will calculate dissociation constants from Gibbs energies of formation and activity coefficients of substrates and products. And it brings also an idea to calculate the reaction rates coefficients..

http://www.physiolibrary.org

http://www.openmodelica.org

Thermodynamics only deals with differences of state functions such as internal energy, enthalpy, Helmholtz and Gibbs free energies with respect to a reference state.  Therefore, your dG Terms have to be replaced by "DELTA"G Terms.  It is physical nonsense to expect Separation in Terms of kinetics for Forward and backward reactions from thermodynamics alone.  You Need kinetic Information, at least in one direction, in order to "understand" the chemical System.  Kinetics and thermodynamics are UNRELATED, always have been, and only the Ratio of Forward and backward reaction yields the Equilibrium constant that is controlled by thermodynamics.  This Separation is a pipe dream of many scientists, and in the seventies a few scientists in your Country (Czech Academy of Sciences) went astray with wanting to do just this.  Nothing but raw nonsense resulted from this "initiative"!  If you have two measured rate constants corresponding to Forward and its microscopic reversible reaction thermodynamics is a powerful constraint for testing the accuracy of the measured chemical kinetics.

Michel, please excuse me the names of the symbols, which I mean always in the sense of "DELTA"G.

Do you think, that the equation kf = (kB*T)/h *exp((Gs-Gt)/(R*T)) from the transition state theory can not be used this way? Or you want just to say that for user is better to set kf than Gt as parameters of reaction?

(The meaning of symbols: kf - reaction forward rate coefficient, kB - Boltzman constant, T - temperature, h - Planck constant, Gs = "DELTA""f"G"0""s" - sum of Gibbs free energies of formation of substrates, Gt = "DELTA""f""G""t" - Gibbs free energy of formation of transition state of the reaction, R - gas constant)

Transition state theory uses the formalism of Equilibrium thermodynamics, but has nothing to do with thermodynamics as I understood from your original question.  The term Gs-Gt corresponds to the enthalpy of activation (constant pressure experiments) or internal energy of activation (constant volume experiments).  For the relationship between activation Parameters and measured activation energies see: D.M. Golden, J. Chem. Ed. 48, 235 (1971). 

If in the original question is selected the reaction-dependent parameter B = - Gt + R*T*ln(kB*T/h), then it fits also the kf = (kB*T)/h *exp((Gs-Gt)/(R*T)).

I do not have an access to the article (D. M. Golden, "Standard states for thermochemical and activation parameters," Journal of Chemical Education, vol. 48, p. 235, 1971/04/01 1971.) just now. If you do not mind, please send it to me (marekmatfyz.cz). Thank you.

Let me first say that I completely object to the statement by Michel Rossi: thermodynamics can say a lot about processes; that is why it was invented in the first place!

One example of its use is the relation that Marek proposes: we use the fact that in equilibrium two opposite reactions balance and from that fact we can extract a relation which only involves the Gibbs free energies of initial and final states. Hence, the suggestion that Marek makes is totally correct. There may be one problem, as B contains the backward reaction rate it will be strongly temperature dependent. This makes it relatively useless unless one has an Ansatz for this temperature dependence. This is where Transition State theory comes in.

But if one remains at one temperature and wishes to study the effect of the substrate only where it is assumed that different substrates do not affect B this is the equation to go for.

 It is futile to attempt to obtain any information with regard to rates of reactions from thermodynamic data, if one believes in the principle of thermodynamics. Time has no role to play in thermodynamics while time plays a crucial role in kinetics. 

Having said that, I would like to make a few observations that might bring into focus some problems that arise in discussing the issue on hand.

1. Chemical kinetics involves the concept of dynamic equilibrium - that a given chemical reaction occurs in opposite directions all the time. The rate of the given reaction is the net rate ( the difference in the rates of forward and the backward reactions). When the net rate is zero, the given reaction is said to be in equilibrium. This concept however, leads to problems - for example, in the equation: ΔG0 = -RT lnK = -RT (kf/kb), the ratio of rate constants is not dimensionless in general, as in cases where the orders of the forward and backward reactions are not equal. In such cases, the ln of a quantity with dimensions creates mathematical problems.

2. Every reaction can be considered, in principle, to be a unique combination of two component reactions - as for example, the two half-cell reactions in redox reactions. The two half cell reactions occur at the  two electrodes of an electrochemical cell. The two half cell reactions occur in opposite directions - Oxidation and reduction. We note that the rates of these two reactions occurring in opposite directions are always equal and equal to the rate of overall reaction. These rates give no information about the energetics of the reactions involved. 

It is not proper for me to elaborate more on this issue here. I hope the above gives some of the problems involved in the discussion of the issue raised by Marek Matejak. The problem is not as simple as it appears to be! 

1. Yes, a dissociation coefficient must be defined unitless. As ratio beween activities of products and activities of substrates. An activity defined as the activity coefficient multiplied by mole fraction of the substance is unitless and makes dissociation coefficient unitless. The physical units of kf and kb are the same (e.g. "mol.mol-1.s-1"), so their ratio is also unitless. However, there is a problem with logarithm of nonunitles coefficient as ln(kf). Any idea how to solve this?

2. Yes, the backward rate must be possible to express in the same way as forward rate.

If activity is defined the way you suggest as a unitless quantity, we will be inviting more trouble. For one thing, the concept of order of a reaction will disappear from chemical kinetics. In chemical kinetics, the units of rate constants differ when order of reaction differs. That arises because activities have units. 

What happens if the total number of moles in the system keeps changing as the reaction progresses? Such will be the case when ever the number of moles of products produced differs from the number of moles of reactants consumed for every unit of reaction. The denominator in the ratio of mole fraction will not be constant, bringing in many problems.

Therefore, by defining activity to be a unitless quantity, we will be introducing more problems than we solve.

Of course, the problem of ln(kf) remains as youpointed out. 

I would like to reemphasize my comment made some time ago despite Ger Koper's vote. ( Science is NOT a matter of democracy, the majority may easily be wrong!).  There is just no way a rate of reaction (forward or reverse) may be guessed a priori from the standard free energy (Helmholtz) or free enthalpy (Gibbs)!  The only way you can use thermodynamics is to calculate/predict the microscopically inverse reaction.  Thermodynamics is a constraint to the ratio (i.e. the equilibrium constant) but is unable to separate both rate constants, forward and reverse.  Therefore, thermodynamics has nothing to do with kinetics, however, it is a constraint in cases one of the rate constants is known.  This is always useful when the reverse rate constant of an equilibrium cannot be measured, for instance because it is too small.  Linear free energy relationships are empirically useful, but have only qualitative justification, and may only be used for the same "family" of reactions, never in a general sense.

Hmm, the Second Law gives the direction of the reaction, doesn't it?

PS I do not see where democracy or majority vote came from, I agree it is a matter of providing arguments based on fundamental principles.

Incorrect! DELTA S (!!) (your "entropy" is in fact an entropy DIFFERENCE, is a state function and determines the position of an equilibrium!)  If you start (as usual) with a non-equilibrium mixture Keq will lead to an equilibrium mixture, whose concentrations will be different from the concentrations you started with. However, no time information (rates!) can be given:  thermodynamics does NOT tell you how fast the establishment of the equilibrium will take.  For that you will have to measure the relaxation time towards equilibrium which is a function of both forward and backward rate processes.

I update the description of this question above by adding the unknown value C, which should express the problem with scaling of both rate coefficients of the chemical reaction.