Teaching the Truth about pH

Stephen J. Hawkes
Oregon State University, Corvallis, OR 97331

What do you tell a student who measures the pH of 0.1 The following sequence puts the concepts i n a n order in
M HCl a n d finds t h a t ( 1 ) i t i s 1.1when s o u have just which they can be taught logically, with t h e most impor-
taught him a method of calculation t h a t saysit is l.OO?kny t a n t concepts first, and without resorting to untruths. It
explanation reduces to a n admission t h a t s o u deceived can he halted a t whatever level a teacher thinks appropri-
him about the definition of pH. ate to the students.
When the student calculates from t h e published densi-
ties t h a t -log [HI] varies by only 0.01 between 0 and 90°C, pH is a measure of the acidity of a solution.
pH decreases as acidity increases.
why does t h e pH vary (2)from 1.08 to 1.13?
acidity is the tendency of a solution to supply Hi to a
If he should ever find out t h a t the pH of 0.01 M HCI i n reaction.
5 m LiCl is 0.8 (3)instead of t h e 2.00 t h a t you taught him the tendency of a solution to supply H+is represented in cal-
to calculate, what will he think of your professional compe- culations by a quantity called the "activity" of Hi in the solu-
tence? tion.
When we teach methods of calculation t h a t give answers the "activity" is directly proportional to the concentration of
t h a t a r e substautially different from experimental reality, Hi in the solution and is also a function of any other sub-
we accept a grave responsibility. When we teach a defini- stances in the solution whose molecules are close enough to
tion of pH that h a s been obsolete for half a century, we are H+ ions to have any effect on them and is also a function of
guilty of professional malpractice. the solvent itself, which is usually water. It varies slightly
with temperature even when the hydrogen ion concentration
Yet no introductory text t h a t I have seen suggests that does not change.
-log [H+lis only a n approximation to pH. Conversely, when other ions in the solution affect the activity of Ht more than
they discuss the measurement of pH they do not say t h a t neutral molecular species because of the interaction of their
the reading on a pH meter is only a n approximation to t h e electric field with that of the Hf.
-log [Ht' t h a t they have given incorrectly as the definition DHdecreases bv one unit for each tenfold increase in activitv.
of pH. pH may, therefore. be measured hy rhc rxrrnr ro which o so-
Students who have been taught t h a t p H i s -log [H+lwill lution interact* \nth a n electrode th;lt is scnairi\,r t o H'. ur
--
obviouslv exoect t h a t their o H meter will read -log 1Htl.
Even p r ~ f e s ~ owho
-
r s have t c e usual background i n physi-
how it affect3 the tquhbnum h e t w e n n rolorrd arid and its
conjugate base.
cal chemistry have t h e same expectation. Arecent puhlish- the potential at the electrode is the conventional measure of
ed experiment (4) contained this expectation i n the second pH. The electrode is calibrated against standard solutions of
which the H+ activity has been determined as accurately as
decimal! Students who make lab measurements carefullv possible by methods that are outside the scope ofthe course.
and compare their result with the expectation can he ex- Standard solutions have been recommended hy national and
~ e c t e dto cheat. international authorities with recommended pH values for
The consequences are worse when they apply their un- calibration of the electrodes.
--
derstanding of DH to real-world Droblems after thevrrmdu- .because the ~otentialat an electrode varies emonentiallv
ate. Our teaching has lead them into error.
Such teaching i s unnecessary.
poses as 10-P".
How to Tell the Truth in solutions so dilute and so pure that Hi ions are not influ-
enced bv anvthinn- exceDt
. the solvent.. DH
. = -log 1Htl.
Any approach to p H must be simple enough for introduc- u~ ~

in mos&al;tians, pH differs from -log IHt1 in the first deei-

tory students to understand, must not be misleading, and mal, but this may sometimes be near enough for very ap-
must lead correctly into whatever calculations we think i t proximate calculations.
necessary to teach. But let u s understand what i s impor- in most calculations the activity must be used.
tant. MacInnes, one of t h e leading workers on p H theory in the activity often is represented as lgammaiIH+l where
his time, made this clear (5)when h e said: (gamma1 represents the effects of the solvent and the other
substances in the solution: in aqueous solutions it is between
In possibly all but one case in a thousand, it is not necessary 0.05 and 13 in every case that I have found but is usually
to consider the meaning of pH in terns of solution theory at all, between 0.5 and 1.0. It is 0.7 in seawater
but only to accept the numbers as a practical scale of acidity
and alkalinity. The Definition of pH
So how much should we teach about pH? Are our calcn- The International Union of Pure and Applied Chemistry
lations necessary? Whatever our answer to this, i t is clear (6)h a s defmed pH essentially a s the reading on a pH me-
t h a t we need to emphasize the practicality of t h e pH scale t e r t h a t has been standardized against standard buffers.
a s far more important than any calculations for which i t That i s
may be used. Moreover, our decision to teach calculations
should not compromise the essential simplicity of the con-
cept of pH.

Volume 71 Number 9 September 1994 747

 This is the same definition as that used in the standards needed is to reverse the calculation. The most useful
of all national bodies that have ruled on the question. The source of critically evaluated constants is the six-volume
actual standard buffers varv from one authoritv to an- work by Smith and Martell (111, but they give concentra-
other. Most use NRS ~NationhBureau of ~tandarhs) buff- tion ratios. They explain that they do this so as to make the
ers as standards. while IUPAC and some others use 0.05 m definition fall into line with the definitions of the stability
potassium hydrdgen phthalate, but all give the same val- constants of metal complexes that are the main part of
ues of pH within 0.005 unit. They are chosen so that their their compilation and these are most usefully defined as
recommended pH equals -log aH+as closely as theory per- concentration ratios. They give the method of reversing
mits the activity to be calculated. their calculation to obtain the above "practical" definition,
The symbol pH has its origin in a paper by Sorensen (7) by adding a constant, which is tabulated in their introduc-
in 1909 about the effect ofH+on the action of enzymes. The tion, to the log of the concentration ratio. Of course, both
concept of activity was not available to Sorensen in 1909, the concentration ratio that is usually given in introduc-
so he believed that the output from his electrodes bore a tory texts and the more useful practical constant have the
Nemstian relation to [HI] and when inventing the useful complication that they vary with the ionic strength of the
term pH for this paper he naturally defined it as -log [H'I. solution especially if the acids are ions as in NH4+or
It is ironic that the natural and inevitable misconception of HP04". Usually this is ignored in introductory texts, al-
this pioneer haunts our introductory texts nearly a century though the concentration ratios used in them to define K.
later, even though their authors have the benefit of a mod- are even more dependent on ionic strength than the prac-
e m education. Sorensen himself came to realize his mis- tical constants. This deceives students into believing that
conception within a decade and invented (8)a new term they can be used in situations where they are useless, and
paH, where aH is the activity of Ht, but this became short- this deception stays with them even after they have be-
ened (9) in common usaee to DH. In time. it was realized come full professors. This problem will be elaborated in
that even the activity was too'dificult to measure and in- later papers.
volved uncertain methods of interoretine exoerimental I have calculated pK.(practical) for the acids most com-
data. So one national body after anbther,&d'eventually monly quoted by teachers of introductory chemistry, using
IWAC, agreed on the unambiguous and simple definition the data and the correction in (11)and they are compiled
that depends only on an experimental datum. Theoretical in the table.
electrochemists need alternative definitions that are usu-
ally given special subscripts and refer to various methods
of defining activity. There is a useful discussion of such ap- Rationalization
proaches in the recent text (10) by Galster, but it is not The error is ingrained so deeply from our own training
necessary to include such refinements in introductory in high school and college that we seek a rationalization
courses. The simple definition as the output from a pH me- that releases us from it. The one I have heard most often is
ter is the most useful for the practical purposes that will be that there are two definitions of pH, one involving concen-
valuable to graduates of an introductory chemistry course. tration and the other involving activity. Atenn can be de-
Eauilibrlum Calculations p& (Practical) for Acids Commonly Quoted in Texts
Equilibrium calculations involving H' must necessarily
involve the activity rather than the concentration, both be-
cause it is the activity that affects the equilibrium and be-
cause in any real equilibrium it is the activity, not the con- Ionic strengthC
centration, that is experimentally measurable through the
pH. The activities of the other substances in the solution 0.0
usually are not known, but their concentrations are known -
H3P04 2.15
from the stoichiometry. The useful definition of the equilib-
rium constant, therefore, involves the activity of H+ and HzP04i 7.20
the concentrations of everything else. The acid dissociation ~ ~ 0 4 ' - 12.38
constant K, for example would be HzC03 6.35
aH+[A-l HCO3- 10.33
&=- CsH5(COzH)z 2.95
[HA]
CsH5(COzH)COz- 5.41
This is variously known as the "mixed" or "Brmsted" or
"practical" constant. In dilute solutions with negligible (COZH)~ 1.25
ionic strength it is identical with the concentration ratio HOzCC02- 4.27
and also with the activity ratio which is the thermody- NH4' 9.24
namic definition. But few real-world problems involve CH3NH3* 10.64
such solutions. Indeed, most of the problems that appear
in textbooks have ionic strengths greater than 0.1, so that HF 3.17
the thermodynamic constant is useless, and the mixed con- CBHSCOZH 4.20
stant is essential and may be only 213 of the concentration CHjCOzH 4.76
ratio, as is the case in seawater. HCN 9.21
The name "practical constant" should convince readers
that it is unethical to use any other variation of K, in an HCOzH 3.75
introductory course. Introductory students will have no HN02 3.15
use for the esoterica of the electmchemist, but will need a H20a 14.00 13.89 13.89 13.93
practical method of getting correct answers to practical 'for H20. Ib(Mactica1I is the Dractical Kw, that is awJOH1. Where I O H l
problems if they need such calculations at all. occurs in a n y &i'culation;it may, therefore, be replacsd bj. ~,ipractical)iam.~
Where are such values tabulated? They are the original b ~ ~ decimaln d not available or varies too greatly with the nature of the
data from which other forms of K,, such as concentration interfering ions to be generally significant.
ratios or the ratio of activities are calculated, so all that is Tonic strength = 0.5oi[(molaritiesofion$+(chargeof ions)?]

748 Journal of Chemical Education

fined in any way we choose, but this dual definition has no authors of textbooks someday will indicate that the expression
foundation in either convention or lopic. pH =-log [HI1is not a definition and that it may be anly a very
I t is logical to define pH in a way that is useful in prac- rough approximation.
tice. Because the concentration of Hi cannot be obtained His "forlorn hope" has been realized in most handbooks,
from the experimental pH, a definition in terms of concen- but is not yet fulfilled in any introductory text that I have
tration makes subsequent calculations impossible, while seen. Perhaps he was too nice. Let us lay it on the line.
definition in terms of activity lends itself to calculation of Authors that define pH a s -log [H+I are deceiving their
other quantities. readers and leading them into serious error. Those of us
I can find no evidence of a convention allowing both defi- who have taught this error for decades, a s I have, are
nitions. I n the nine reference works (6,12-19) that I have guilty of professional malpractice and need to correct our
consulted, not one quotes both definitions a s valid. Three teaching in the future.
of them quote -log [H*] as a n approximation, and two give The error can be avoided easily, a s has been shown
it as the only definition. The remainder give the definition above, by a trivial change in pedagogy that will not cause
correctly. difficulty to teachers or students.
How Close is Activity to Literature Cited
The solution being tested and the standard solution usu-
ally differ from each other in ionic strength and in other
ways. This makes even the determination of activity a n Verlag, Berlin.
approximation because of junction potentials in the cell, 3. Peterson.0. Chomk-lng. nchn. 1368.40.76.
althoueh
" it is much closer than DH is to -loe IHil.
good enough approximation? T L ~IUPAC document (6)
--
- Is it a
4. Cawley J. J.J. Cham. Edue. 1993.70. 596598.
5. MacInnes, D.A Seieneo 1948,108,693.
6. Mills, I.: Cvitaa, T:Kallay, N.; Homann, K; Knchitsu, K. Quonliiies, Units, and
puts the difference between pH and -log a H + a s no more Symbols in Physiml Chemistry: Blackwell Scientific: Lpandon, 1988: pp 5P55.
than 0.02 for aqueous solutions more dilute than 0.1 M. I n 7. Sorenren, S. P L. Bioehrm 2.1909.21, 131 and 201. Pvblifhed simultaneovsly in
Compt Rend. l h u . Lab. Carisberg 1909,8,1.
more concentrated solutions the difference may be greater 8. Sorensen, S. P. L.; LindersVom-Lsng, K. Compl. Rend. l h o . Lab. Corlsbwg 1924.
but may be offset by compensating errors. In standard 15, No.6.
9. Hamed. H. S.:Owen. B. B. Th4 Physical Chrmistry ofElectrolytie Solutions; Rein-
seawater. for examole. the difference (20) is about 0.05. hold: New York, 1943; p 317.
oceanographers either specify standard buffers that are 10. Galater H.pHMeosvremen(;VCH:New York, 1991, pp 4 7 5 6 .
specifically designed to eliminate this error when measur- 11. Mamll, R. M.; Smith, A. E. Critical Sfobilih Comlonfs.Vols. I 4 ; Plenum: New
York. 1975-1989.
ing the pH of seawater, or they use sophisticated methods 12. m i a x L.; K ~T J.~compmhensiue
~ , nictionory of physicoi cheminry; bentice
of correction that are outside the scope of a n introductory Hall, 1992; p 306.
13. Lewis, R.
course, or determine the activity d ~ c ~ t r o ~ h o t o m e t r i r a ~ ~ y . ,003."" SO" J. Condensed Chpm&l Dictionary: Van Nostrand Reinhold: New York,
A""",

The official procedure (211 for determining the pII of blood 14. Bevan. S. C.;Gregg, S. J.; Rosseinsky, A C~onciseElymologiedDictionary ofChem-
involves
~-~ ~ ~secondam standards referred to the NBS stand-
~
iny;Applied Science Publishem: London, 1976; p 2 8 4
15. K q t . D.Dcctionary of Chem~callkchnology; Elsevier. 1980. p 284.
ard buffers, with solutions that simulate those properties 16. Pa~khh,S.P.M&row-Hill Enwclopedia dChcmisiry. 2nd ad.: MeGrsw-Hill, 1993.
of blood that affect pH measurement. 17. Lide. D.R., Ed. CRCHondbook orchemistry ondPAysles, 74thed.; CRCPuhlishing,
1993. pp G34.35.
Conclusion 18. Dean, J. A. Langds Hondbwk ofChrm&try. 14th ed.; MeGraw-Hill, 1992: pp 8-103
to 8-108.
I t has been nearly half a century since MacInnes (5) 19. Parhr, S. P McGmw-Hill Dictionary ofScipnLifi and lkcknicol h s , 4th ed.;
MGraw-Hill: New York, 1989: p 1404.
wrote 20. Culbcrson, C. H. In M m i m Electmch~mlsfry:Whiffield. M.; Jagner,D.,Eds; Wiley
New York. 1981: p 200.
The relation pH = -lag [Htl is an approximation anly, and 21. Maaa,A. H.J.:Wei6berg.H. F.:Bumett,R. W.:Mullc~Plathe,0.;Wimber1ey.P. 0.:
using it as a definition may lead to serious errors. It is the Zijlstrs,W G.; Durst, R.A,: Sigaard-Anderson, 0 . J. Clin. Chem. Ciin. Biochem.
author's rather forlorn hope that compilers of handbwks and 1987,281-289.

Volume 71 Number 9 September 1994 749