Talk:Acid dissociation constant

pKa and the logarithmic expression

There is a problem re pKa as logarithm of a quantity K_a which is non-dimensionless. How does the dimension and its units affect the numerical values of the logarithmic pKa? This aspect must be clarified in the attached note!--109.166.129.52 (talk) 12:53, 18 September 2019 (UTC)

When $K_{\mathrm {a} }={\frac {K^{\ominus }}{\Gamma }}$ and both $K^{\ominus }$ and $\Gamma$ have the same dimension, $K_{\mathrm {a} }$ has no dimension. See, for example, Rossotti and Rossotti (1961). Petergans (talk) 09:40, 19 September 2019 (UTC)

What exactly is the dimension of Γ? Isn't it usually considered non-dimensional?--109.166.139.27 (talk) 14:38, 10 November 2019 (UTC)

The answer is given above. The dimension of $K^{\ominus }$ and $\Gamma$ will depend on the unit used in the measurement of concentration, commonly mol dm^-3. LogK is possible when the value of $\Gamma$ is (implicitly) equal to 1, whatever the unit of measurement. Petergans (talk) 20:35, 10 November 2019 (UTC)

More precisely: the answer is given above in the Talk page section Discussion of dimension, as well as in the article section Dimensionality. Dirac66 (talk) 01:32, 19 November 2019 (UTC)

Strong acids and bases

(moved from larger_than_about_12_is_more_than_99%_dissociated_in_solution , above) I have now verified some of the textbook references in the section Strong acids and bases, and will make some changes. The book by Dasent does not mention solvent leveling, so I will replace it as a reference by Porterfield and also Shriver and Atkins, with some revision of content. I have been unable to find a source saying that the limiting pK_a values are –2 for strong acids and 12 for strong bases. The article now attributes the latter value to Shriver and Atkins, who actually say that pK_a is approximately 0 for strong acids and pK_b is approximately 0 (or pK_a approximately 14) for strong bases. Note that these values are symmetric about pH 7 as they should be. The buffer capacity curve above is also symmetric about pH 7, so it cannot lead to an asymmetric prediction such as –2 and 12, which is in fact unsourced if the value attributed to Shriver and Atkins is corrected. Dirac66 (talk) 02:53, 16 November 2019 (UTC)

I am surprised that no source was found to verify the statement concerning strong acids. It is trivial to prove. For a strong acid, HA, [H⁺] + A^- ⇌ HA, K=[H]²/[HA], so when [H⁺]/[HA] = 0.01 (1% dissociation) K=0.01 [H]. It follows that for a 1M solution of acid K=0.01, pK = -2. Many references to a value of -1.76 can be found in the literature. These values are indicative as they depend on an arbitrary criterion such as [A^-]/[HA] < 0.01. I prefer the value of 2 because of the arbitrary nature of the defining criterion.

The plot will not be symmetric because with strongly alkaline solutions the self-dissociation of water has to be considered simultaneous with the alkali protonation. A value of 14 is an unrealistic upper limit. 12 represents an approximate upper limit for what can be determined experimentally and consequently for what can be found in the literature. Of course these limits apply only to aqueous solutions. Petergans (talk) 11:27, 16 November 2019 (UTC)

Yes, the statement about strong (aqueous) acids is easy to prove IF we only consider the one equilibrium HA ⇌ H⁺ + A^-, or more correctly HA + H₂O ⇌ H₃O⁺ + A^-. Similar calculations for strong bases are equally easy if we only consider the one equilibrium B + H₂O ⇌ BH⁺ + OH^-.

However you have pointed out above that at very basic pH the problem is complicated by the second equilibrium H₂O ⇌ H⁺ + OH^-, or more correctly 2 H₂O ⇌ H₃O⁺ + OH^-, leading to large changes in buffer capacity near pH 14. If this is true (and we still require a source), then the same equilibrium should also complicate the calculations at very acidic pH due to large changes in buffer capacity near pH 0. These changes are in fact shown in the plot of buffer capacity that you have inserted above, which is symmetric about pH 7 with large increases in buffer capacity in both regions, near pH 14 and near pH 0. Dirac66 (talk) 22:18, 16 November 2019 (UTC)

There are good reasons why the reaction H⁺ + H₂O → H₃O⁺ is ignored in discussions concerning acid strength in aqueous solutions: it is effectively a quantitative reaction and so has no effect on equilibrium constant calculations, except when the solution is extremely concentrated. Also, the hydronium ion forms weak complexes, such as H₄O₉⁺, with other water molecules. The reactions H⁺ + nH₂O → H_2n+1O_n⁺ cannot be treated individually, Put another way, in the context of aqueous solutions, all protonation constants are macro-constants and this fact is almost never stated explicitly as the extent of proton aquation in any given solution cannot be quantified. Petergans (talk) 11:09, 17 November 2019 (UTC).

Microconstants

Two comments on this section:

First, since pK_a = 14 - pK_b for a base is more complicated than pK_a for an acid, it would be better to start with an example having two nonequivalent sets of acid groups, such as citric acid. Then spermine with two nonequivalent sets of basic groups could be mentioned as a second example starting with the word "Similarly".

Also the most recently added point about the additivity of microreaction constants is of interest, but the equation cannot be correct. Surely it should be K_a (or K_b) which should be additive and not the logarithmic pK_a. And as always it would be best to have a source for the statement, which would serve here to confirm the correct additivity relation. Dirac66 (talk) 15:24, 17 February 2020 (UTC)

Regarding the 2nd point, thank you for pointing out the error. The source is ref.64; I worked on the computer program HypNMR, but not on the experimental part of the paper or on the determination of the micro-constants. Petergans (talk) 21:37, 17 February 2020 (UTC)

More on micro-constants: https://pubs.rsc.org/en/content/articlelanding/1994/P2/p29940000265, Michal Borkovec and J. M. Koper, Analytical Chemistry 2000, 72, 14, 3272-3279 Petergans (talk) 22:30, 17 February 2020 (UTC)

Thanks for changing the five pK_a to K_a. For the case of a base such as spermine with two isomeric conjugate acids, I think it is actually K_b which is additive:

$K_{b}=\mathrm {\frac {[HB^{+}][OH^{-}]}{[B]}} =\mathrm {{\frac {[HB_{terminal}^{+}][OH^{-}]}{[B]}}+{\frac {[HB_{internal}^{+}][OH^{-}]}{[B]}}} =K_{terminal}+K_{internal}$

On the other hand for an acid such as citric acid with two isomeric conjugate bases, a similar reasoning is true for K_a. Perhaps we could include both cases, one with the algebra above and the other with the word "Similarly". Dirac66 (talk) 17:47, 19 February 2020 (UTC)

It is clear that a macro-constant value is the sum of the values of the micro-constants with both association and dissociation definitions. We (ref 64) followed the literature cited above in choosing which convention to use. Petergans (talk) 22:20, 19 February 2020 (UTC).

Are you saying that for a specific case such as spermine with one base and two isomeric (or tautomeric) conjugate acids, that K_b is the sum of the K for two microreactions with a common reactant, while K_a of the same macroreaction is the sum of the K for two microreactions with a common product?? I had been considering only the case of a common reactant. If the additivity is true for a common product also, then for an example with two microreactions of equal K, then both K_b and K_a will be double the K of a corresponding microreaction, so that K_bK_a = 4 x 10^-14 !??

Also I have now looked at the paper by Frassineti et al. I do see that the Results section first mentions protonation constants and then switches to pK_a for deprotonation before mentioning macro- and micro-constants. Does this mean that additive macro-constants are defined for both common-reactant and both common-product cases, without worrying about the fact that K_aK_b is no longer 10^-14? I don't see this mentioned in the paper? Dirac66 (talk) 21:51, 20 February 2020 (UTC)

A macro-constant value is always equal to the sum of the micro-constant values. With spermine,

Association: $K_{a}^{macro}=K_{a}^{terminal}+K_{a}^{internal}$ (1)

Dissociation: $K_{d}^{macro}=K_{d}^{terminal}+K_{d}^{internal}$ (2)

Similar relationships do not apply to log K or pK values. Obviously, the total concentration is the sum of the concentrations of the micro-species.

Also, any dissociation constant value is the reciprocal of the corresponding association constant value.

$K_{a}^{macro}=1/K_{d}^{macro}$ (3)

$K_{a}^{micro}=1/K_{d}^{micro}$ (4) for each micro-species Petergans (talk) 23:51, 20 February 2020 (UTC)

Thank you. These are four simple and clear equations. But the problem is that the third is incompatible with the others, as can be seen by multiplying the first two and then using the fourth to evaluate two of the four terms.

$K_{a}^{macro}K_{d}^{macro}=(K_{a}^{terminal}+K_{a}^{internal})(K_{d}^{terminal}+K_{d}^{internal})=1+K_{a}^{terminal}K_{d}^{internal}+K_{a}^{internal}K_{d}^{terminal}+1$

Since all four terms are positive, this sum must be greater than 2, whereas your third equation says it should be 1. Can you identify the error in this argument please? Dirac66 (talk) 02:04, 21 February 2020 (UTC)

I think it's to do with the definitions. I define all association constants in terms of [H⁺], whereas your definition above for dissociation constants is in terms of [OH^- ]. Therefore the expression for any product K_aK_d with these definitions must either include [OH^- ] or use K_w / [H⁺] in place of [OH^- ]. Petergans (talk) 11:00, 21 February 2020 (UTC)

But I did switch to your definitions in my last edit on this page. It is true that my third and second last edits do include [OH^-] and 10^-14. However my last edit at 02:04, 21 February 2020 (UTC) is based on your most recent edit and has $K_{a}^{terminal}K_{d}^{terminal}=1$, rather than 10^-14. So we are now using the same definitions, but I still find an inconsistency between your 4 equations. Dirac66 (talk) 18:53, 21 February 2020 (UTC)

I can't figure out what's wrong. Using the defining expressions for the macro-constants, (electrical charges not shown)

$K_{a}K_{d}={\frac {[HA^{i}]+[HA^{t}]}{[H][A]}}{\frac {[H][A]}{[HA^{i}]+[HA^{t}]}}=1$ Petergans (talk) 14:41, 22 February 2020 (UTC)

Aha! Here you have an expression for $K_{d}$ (macro understood) with two terms in the denominator, which I suspect is correct. However your four numbered equations above include Dissociation: $K_{d}^{macro}=K_{d}^{terminal}+K_{d}^{internal}$ (2), with two terms in the numerator. It would seem that to obtain the simple sums (3) and (4), we must define $1/K_{d}^{macro}$ to be additive rather than $K_{d}^{macro}$ itself. Like combining electrical resistances in parallel. Since you are more familiar than me with the details of equilibrium constants, I will ask you if the literature agrees. Dirac66 (talk) 19:01, 22 February 2020 (UTC)

I don't know where to look. We have always used association constants for acids and bases for the specific reason that stability constants for complexes of metal ions have always been defined as association constants. As far as I know, all general-purpose computer programs involving stability constants use association constants for all species for the same reason. Nevertheless, it is still commonplace to see dissociation constant values being cited (as pK_a) in the literature for organic acids and association (protonation) constant values for organic bases. In fact the [Sirius ] auto-titrator software is designed for the determination of dissociation constants of acids. Petergans (talk) 20:14, 23 February 2020 (UTC)

Resolution

I have found the origin of the problem. Using association constants, adding one proton to spermine, L

L + H ⇌ LH(i)+ LH(t)

[LH(i)] + [LH(t)] = K(i)[L][H] + K(t)[L][H] = (K(i) + K(t)) [L] [H]

So, the macro- association constant value is equal to the sum of the micro- association-constant values.

But, for removing one proton from LH and using dissociation constants

LH(i) + LH(t) ⇌ L + H

[L][H] = K'(i)[LH(i)] + K'(t)[LH(t)]

The macro- dissociation-constant value is not equal to the sum of the micro- dissociation-constant values. Petergans (talk) 13:23, 24 February 2020 (UTC)

I agree for the case of spermine, which can be characterized as per Frassineti et al. (p.1048) as a base containing 2 (nonequivalent) sites capable of accepting a proton (association). (I would refer to spermine though as B, not L.)

On the other hand citric acid which I mentioned above is an example of an acid containing 2 sites capable of donating a proton (dissociation). For citric acid the macro-dissociation constant is a simple sum and the macro-association constant is not. In general K_assoc and K_dissoc cannot both be additive for the same system, since that leads to incompatible equations as I showed above.

For the article, I think the Microconstants section should mention both cases, since many important examples of each can be found. Also I think it would be better to mention the citric acid type first, since the article is called Acid dissociation constant, and the convention of describing both acids and bases by pK_a (rather than pK_b) makes the theory for acids slightly simpler so that it should be presented to the reader first. Dirac66 (talk) 20:36, 26 February 2020 (UTC)

You make a valid point. However, a citation is needed. A quick search for "microconstant" gave, e.g.

https://www.ncbi.nlm.nih.gov/pubmed/7849133 and https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4074347/.

I'm sure the paper by Alex Avdeef is reliable, but I can't access the text to get a reference for the method used. Also, Helmut Sigel has used a chemical method - blocking one site by esterification to get at the dissociation constant for the other site. Please feel free to follow this up. Petergans (talk) 09:35, 27 February 2020 (UTC)

I had forgotten that I created the section Equilibrium constant#Micro-constants which gives more details on micro-constants, including citations for computational methods. Petergans (talk) 14:51, 27 February 2020 (UTC)

The two papers you have cited most recently above are for more complicated systems. The spermine system has one base and two conjugate acid isomers, while the citric acid system has one acid and two conjugate base isomers, again for a total of three species and two equilibria. But the niflumic acid studied by Avdeef is an amino acid with one conjugate acid, one conjugate base and one zwitterion isomeric to the original molecule, for a total of four species with four equilibria.

For citric acid, I found one 1961 reference by Martin here [1] but I do not have access to the complete article. So since the spermine paper is the best reference we have for a 3-species system, it is probably best to continue to use spermine as the first and major example for this section. Then we can briefly mention citric acid as an analogous acid system with Martin as source, and then perhaps the 4-species example(s) with zwitterion.

And one more minor correction. For the spermine system, $K_{b}(notK_{a})=K_{terminal}+K_{internal}$ as discussed above. Dirac66 (talk) 23:25, 27 February 2020 (UTC)

citric acid

This structural representation of citric acid, H₃L is, perhaps, misleading. The molecule is tetrahedral at the central carbon atom. It has 3-fold symmetry (ignoring rotation about C-C bonds and assuming equivalence of the protons in each -CO₂H group). The anions H₂L^- and HL^2- both have local 2-fold symmetry at the central carbon atom so they have no distinguishable micro-states to which micro-constants might apply.

The article on zwitterion is relevant but its quality was very poor - now re-written. Isoelectric point is a macro property and is therefore irrelevant to the micro-states involved in forming a zwitterion.Petergans (talk) 19:15, 28 February 2020 (UTC)

Petergans (talk) 19:15, 28 February 2020 (UTC)

New source

I have found an interesting new source for this section: Splittgerber, A. G.; Chinander, L.L. (1 February 1988). "The spectrum of a dissociation intermediate of cysteine: a biophysical chemistry experiment". Journal of Chemical Education. 65 (2): 167. It presents an undergraduate experiment for the determination of microconstants for the deprotonation of cysteine at the nitrogen and at the sulphur, using simple UV spectroscopy. As a Journal of Chemical Education article it is somewhat more pedagogical than the spermine paper, and starts by explicitly presenting equations relating micro and macro constants, for subsets of reactions corresponding to three cases of interest here.

For parallel deprotonation of the neutral cysteine zwitterion at two sites (N and S), so that one reactant leads to two products, the JCE paper states that the macroconstant K₂ = K_A + K_B (using the authors' notation). This case is equivalent to the deprotonation of citric acid which I have mentioned above. And incidentally, citric acid does not have C₃ symmetry, since the central carbon is bound to 2 -CH₂COOH groups and 1 -COOH group, not 3 equivalent groups.

For convergent deprotonation of the cysteine (-1) ions produced in the above paragraph to a common (-2) ion, 1/K₃ = 1/K_C + 1/K_D, so 1/K is additive here. Of course if we consider the protonations (associations) in the reverse direction, then K for association is additive. So we have a source for saying that if K is additive for reactions in one direction (here protonation), then 1/K is additive in the other direction (here deprotonation). Another example of this case is the spermine example in this Wikipedia article, for which parallel protonation of one neutral base leads to two (-1) anions, and again the macroconstant K for association (protonation) should be additive. At the moment however pK is incorrectly marked as additive, implying that K is multiplicative.

The third macroconstant considered is for the double deprotonation of the zwitterion at both N and S, which can proceed by two parallel paths through either (-1) ion to the final unique (-2) ion product. For this case the dissociation constants K are really multiplicative so that the pK are really additive.

I believe this article can be used as a source for a more complete Microconstants section presenting the various cases. Dirac66 (talk) 02:25, 5 March 2020 (UTC)

Can you send me a reprint - peter.gans@hyperquad.co.uk? Another "simple" example is spermidine which has 3 non-equivalent amino groups, 3 micro-states each for LH⁺ and LH₂²⁺. Petergans (talk) 10:53, 5 March 2020 (UTC)

Examples of values when the quotient of activity coefficients is different from 1

I think the article should present examples of values for this constant when the quotient of activity coefficients Γ is different from 1 for acids like sulfuric acid, nitric acid, phosphoric acid, carbonic acid, etc.--109.166.136.166 (talk) 14:39, 26 April 2020 (UTC)

What sources could are known to include such data for these acids?--109.166.136.166 (talk) 14:48, 26 April 2020 (UTC)

Putting the quotient of activity coefficients Γ to be equal to 1 is an arbitrary assumption. It can be applied to any equilibrium constant since K^T=KΓ. The assumption is needed because the thermodynamic constant, K^T, must be dimensionless; otherwise Log K does not exist. It is equivalent to the assumption that the properties of solution are the same as properties of an ideal solution, so K^T has the same numerical value as K. Values for activity coefficients can be calculated using the Debye-Hueckel equation, Pitzer equations and others.Petergans (talk) 08:35, 28 April 2020 (UTC)

Variation of pK_a of acetic acid with ionic strength.

I think that values in tabular form for the dependence on ionic strength of Γ when a fixed reference state (as mentioned in a section above) is used should be included. The experimental values of activity coefficients for these acids, obtained by various methods, not those based on calculations (with the mentioned equations) which rely on various assumptions like Debye-Huckel, are more valuable for the article. The activity coefficients values for concentrated strong acids like those mentioned also remove the problem of negative pK_a discussed above which arises when activity coefficients are ignored.--109.166.139.1 (talk) 01:48, 10 May 2020 (UTC)

Also the connection to the law of dilution including activity coefficients for concentrated strong acids should be emphasized. Usually in absence of lower than 1 or greater than 1 values of activity coefficients the law of dilution in its classic form has a non-constant constant K_a.--109.166.139.1 (talk) 02:03, 10 May 2020 (UTC)

For a salt A_pB_q, the quantities $\gamma _{{\ce {A^+}}}$ and $\gamma _{{\ce {B^-}}}$ cannot be determined individually from experimental data. The quantity that can be obtained is the mean activity coefficient, $\gamma _{\pm }={\sqrt[{p+q}]{\gamma _{\mathrm {A} }^{p}\gamma _{\mathrm {B} }^{q}}}$. The value of this quantity varies with solute concentration. The effect of this variation is illustrated by the graphic taken from acid dissociation constant and repeated here. Petergans (talk) 10:15, 10 May 2020 (UTC)

Association and dissociation constants: section notation disagrees with rest of article + textbooks

I have just noticed that the section "Association and dissociation constants" uses an unsourced and nonstandard notation which conflicts with rest of the article. The usual textbook notation as in the rest of the article uses K_a for an acid DISsociation constant, where the subscript a is for acid, and K_b is the association constant for a base. However this section now claims that K_a is an "ASsociation constant" for an acid, a definition which is not the usual K_a, and that K_d (a symbol not usually used in this topic) is the acid dissociation constant normally denoted as K_a. This is extremely confusing and conflicts with the rest of the article and also with every textbook I know.

The two recent incorrect edits by 94.63.152.174 seem to have been an attempt to clarify this confused notation. I think it would be better to either delete this section entirely, or else to find sourced equations which presumably would use K_a for the acid dissociation constant as in the rest of the article, and another notation (K_a^-1?) for the acid association constant. Dirac66 (talk) 02:55, 9 August 2020 (UTC)

There are historical factors behind the apparent discrepancy. With organic acids it makes sense to use a dissociation constant as a property of an acid. It also makes sense to use an association constant for protonation of an organic base. On the other hand, association constants are the obvious choice for metal-ligand and supramolecular complexes. Both conventions are in current use in their respective field, so both conventions need to be respected in WP. I have tried to do this by including a section in the article in which the relationship is described.Acid dissociation constant#Association and dissociation constants I am not aware of a source that treats both association and dissociation in one place. That is why I "invented" the notation. The individual definitions are "common knowledge".

In WP I have chosen to use the association convention, where possible, because that is the convention that must used in all the general-purpose computer programs that, in turn, are used to obtain stability constant values from experimental data. This is necessary because the chemical system may involve both ligand dissociation and metal-ligand association reactions.

HL = L^- + H⁺ (dissociation) and M^m+ + L^- = ML^(m-1)+ (association).

In that context, hydrolysis of a metal-ligand complex must be treated as protonation of the hydrolysed moiety. There are commercial auto-titrators, sold mainly to the pharmaceutical industry, that are supplied with software for the determination of pK_a values of acids, included as part of the instrument software. Consequently, both conventions need to be respected in WP. Petergans (talk) 14:28, 10 August 2020 (UTC)

May I add that this is ancient stuff. See Rossotti & Rossotti (1961) The determination of stability constants pp 5-10 Petergans (talk) 15:36, 10 August 2020 (UTC)

OK, I will accept that the association/dissociation nomenclature and symbols are used instead of acidity/basicity in some sources, such as Rossotti & Rossotti. But it is very confusing for readers to switch notation and definitions in one section of the article without warning, especially since the symbol K_a is changed from acidity to association = basicity! I would suggest instead stating explicitly that this is an alternate notation for this section only. We could say something like: "The acid dissociation or acidity constant K_a is sometimes written K_d for dissociation (cite), and similarly the basicity constant K_b can be written K_a for association (cite). In this case K_a does not mean acidity constant but rather association constant. This alternative notation will not be used in other sections of this article." I think this would help to reduce the readers' confusion. Dirac66 (talk) 01:37, 11 August 2020 (UTC)

To remove ambiguity my solution would be to give specific examples like "association:A+B=AB K=[AB]/[A][B] and dissociation: AB=A+B K=[A][B]/[AB]". Notation, including alternatives, can be described after each example. How can we resolve the possible ambiguity in notation between pKa (dissociation) and Ka (association)? The "a" can stand for either "acid" or "association". As I mentioned before, the natural choice in organic chemistry is to use dissociation for acid deprotonation and association for base protonation but this cannot be followed in general-purpose computer programs, where a ligand may be a base like ethylene diamine, or a conjugate base like the acetate ion. Petergans (talk) 10:33, 12 August 2020 (UTC)

Such extra explanations may help, but I think it is also essential to specify that the section uses a different notation than the others. The natural tendency of most readers is to assume that what it said in one section applies in another section of the same article, unless the contrary is explicitly specified. Of course it is helpful to explain also the reason for the variation, e.g. by saying that one notation agrees with chemistry textbooks and the other is used in computer programs to calculate equilibrium constant values (from what data?) Dirac66 (talk) 21:59, 12 August 2020 (UTC)

Possible solution

We could adopt some slightly modified notation for the association constant, instead of K_a which is already used for acid dissociation constant in the rest of the article. Perhaps K_A or K^a or K_assoc. Any of these would be slightly different from K_a and so preserve the convention that one symbol should have a unique meaning in a given article. But at the same time any of these is close enough to K_a so that it would be reasonable to add that it is called K_a in computer programs for equilibrium constants. Could I ask Petergans to suggest which of these variations seems most appropriate? Dirac66 (talk) 20:52, 13 August 2020 (UTC)

That makes sense. My preference is for K_assoc. Perhaps a footnote is also needed to explain that there is a conflict of notation in the literature: the symbol pK_a is generally used with dissociation constants, but K_a is used with both association and dissociation constants, depending on context (specify?). As we are inventing the notation for the purpose of clarity in WP, why not use either K_a or K_association? Petergans (talk) 08:42, 14 August 2020 (UTC)

OK, I have now rewritten the section using K_assoc (and pK_assoc) as I understand it. Please check that I have things right, as I realize that you are more expert than I in equilibrium constants. Also some sources would be helpful, for example a description of a computer program which uses K_a(ssoc) and K_d(issoc) instead of acidity and basicity constants. Dirac66 (talk) 01:38, 16 August 2020 (UTC)

Many thanks. I will check the text later.

An old paper of ours (Gans, Sabatini and Vacca, Talanta, 43 (1996)1739-1753) has literature references for some 60 computer programs concerned with software for stability constants' computation. If I remember correctly, all the programs use the association convention, exclusively. It all goes back to LETAGROP (ca. 1960) which was the first general-purpose program to be published. On the other hand, textbooks on organic chemistry traditionally use the dissociation convention for acids and association for bases.

The association convention has always been used for complexes of metal ions, as they may be viewed as association reactions (in solution) of a metal ion and a ligand even when the ligand replaces water molecules in the coordination sphere of the aqueous metal ion, [M(H₂O)_m]^z+(aq) + nL(aq) ⇌ [ML_n]^z+(aq) + mH₂O(aq), the concentration of water is effectively constant.

A similar situation obtains in relation to hydrolysis of metal ions which is represented as removal of H⁺ in computer programs and addition of OH^- in descriptive texts. This also applies to solubility products of metal hydroxides; the value may given as K_sp, equal to the product [M][OH]ⁿ or K^*_sp, equal to [M][H]^-n. The difference between the numerical values of log K^*_sp and log K_sp is 14n. The issue arises because the rôle of solvent is ignored when dealing with pK values, even though we may write the equilibrium as AH + H₂O ⇌ A^- + H₃O⁺. It is an acceptable approximation that the concentration of water in an aqueous solution of an acid or base is effectively constant and independent of the pH of the solution at all but the highest acid or alkali concentrations (ca. 1M and above). Petergans (talk) 13:21, 16 August 2020 (UTC)

Values for common substances

Many articles list the basicity constant of a substance, so the table should be extended in such a way to simplify comparing basicity and to common bases.

Inline text blanks

As seen in the Wikipedia app running on Android 10, there are multiline blanks for <chem>…</chem> formatted expressions. I don't know the syntax, so am not trying to correct these. Nikevich 18:54, 21 December 2021 (UTC)

Definition of pKa

The term is used a lot (more than 150 times in the article), but I am not finding an explanation for what it actually means, unless it is hidden somewhere deep in the text. Please define technical terms and symbols at the first appropriate opportunity, preferably at first use. Cheers,· · · Peter Southwood ^(talk): 05:16, 10 January 2023 (UTC)

It is defined in the Introduction by the equation pK_a = -log₁₀K_a, but admittedly there is a problem for readers who don't know what log (logarithm) means. Perhaps we should add a simple example: For a weak acid with K_a = 10^-5, log K_a is the exponent which is -5, so that pK_a = 5. And for acetic acid with K_a = 1.8 x 10^-5, pK_a is close to 5. Dirac66 (talk) 14:29, 10 January 2023 (UTC)

We also could explain the trend, because the whole idea of "equilibrium constant" is already technical. Something like "A higher Ka means the compound is more acidic because at equilibrium it is more dissociated. For convenience and due to the range of values encountered in nature, it is often converted to a logarithmic scale, pKa, where a lower value means a compound is more acidic." Compare how our article here seems only advanced level, with the lead-section of pH, which includes lay-reader-level details: "Acidic solutions (solutions with higher concentrations of H+ ions) are measured to have lower pH values". DMacks (talk) 15:28, 10 January 2023 (UTC)

I have now added the above explanations to the article, at the end of the introduction. Dirac66 (talk) 02:41, 29 March 2023 (UTC)