Escolar Documentos
Profissional Documentos
Cultura Documentos
PETERA. STEWART"
Department of Physiology and Biophysics, Brown U n S v e ~ s i f Provtsl'ence,
~),
RI, U.S.A. 02912
Introduction
Quantitative analysis of ionic solutions by direct
application of fundamental physical and chemical
principles has always been possible in principle, but
has been effectively prohibited until recently by the
overwhelming amounts of calculation required. Now,
however, computers can do at1 those calculations easily
and quickly, so that direct quantitative analysis is
simple and extremely useful. The approx irnations , simplifications, and "tricks" which the precomputer situation required are no longer needed, and the confusion
which they have generated is no longer necessary. The
result is a new quantitative approach to acid-base
which has revolutionized our ability to understand, predict, and control what happens to hydrogen ions in
living systems (Stewart 1981).
'Present address: P.O. Box 24, Beer Harbor, WA, U.S .A.
98243.
The physicochernical analysis required is at a remarkably elementary level, is easy to understand, and
can be readily applied to biological sca8utions. The
mathematics required is only high school algebra.
While this is a simple and practical approach to ionic
solutioms, it requires significant changes in many gemerally accepted concepts of hydrogen ion behavior. The
purpose s f this review is to outline the analysis and to
suggest a few of its most useful implications. More
complete details are spelled out Ira the above reference.
The analysis will be developed by progressive application to four isolated solutions: ( I ) pure water, (2)
strong ion solutions, (3) weak acid "buffer'9 solutions,
and finally, (4) solutions also containing C 0 2 . At
each stage, we shall ask, and answer quantitatively, the
fundamental question: "what factors determine the hydrogen ion concentration ([H' 1) in this solution?"
so soluBiological solutions always contain @02,
tions with C 0 2 are most interesting. They are also the
The analysis to follow will establish what does determine changes in [ H t ] , and the identification of these
two kinds of variables will be an essential part of that
analysis.
Theoretical development
Pure lvutPr
Liquid water is so much a part trf all living systems
that it is easy to overlook its very peculiar properties.
Three of them are iinportant for this analysis: high dielectric constant, very slight dissociation into H i and
OH ions, and extraordfnarily high molar concentration
(about 55.5 M).
The high dielectric constant of water causes substances whose atoms are held together primarily by
electrostatic bonds to dissociate into their component
ions when they dissolve in it. Water is therefore a
strongly ionizing solvent. That it has this effect on itself
partly explains its own slight dissociation. While the
extent of water dissociation is minute, enormous biological importance is attributed to it. Although often
overlooked, the large molar concentration of liquid water is also iinportant for the analysis of hydrogen ion
bchavior.
Since the water dissociation reaction equilibrates rapidly, we must always have
KL is called the ion product,for H7&rtc.r.It is very sensitive to temperature, because Kw is. In solutions other
than pure water, its value also depends on osmolarity,
because the [H20]term does, and on ionic strength, and
on the presence of specific ions.
The numerical value of K k , as well as of all the other
constants we shall introduce later. also depends on
whether the [] symbols denote concentrations or activities. Strictly speaking. activities or effective concentrations should always be used when dealing with reaction ratcs or equilibrium constants. Fortunately, the difference between activities and concentrations is usually
not large, and concentrations are so much simpler to
think about that it is preferable to use them at first.
When the resulting theory is applied in practice, it may
then become important to use activities to ensure numerical agreement with real system behavior. It should
also be recognized that activity corrections apply to
chemical concentrations, but not to charges per ~ e .
1446
k'z.
As cmly strong ioras have been added, no ncw equilibria are involved. Water dis\c~ciationequilibrium is
still constrained by equation 121. (Balce again we have
two independent, simultaneouq equations, [2j and [ 51,
expressing the rule\ of the game for [ I I ' J aend [Oil I.
We can combine and rearrange tl~eanby the rules of
algebra just as before, but this tianc the results are the
follow ing qttacfratic exprcsxions:
[6]
(H-1
t'K\l
+ ( [ N a ']
[Cl ] ~ ' / 4
( [ N a b] - [CI 1112
These expressionc answer the question '\+hat c1etermines the [I1' 1 of pure water?" The aalswca- i, ""K:".
'H'hcy also tell
that K: dctcrn-aints the [OII ] of pure
water.
S t r o ~ gi o ~ ls o k ~ f f o ? ~ ~
Electrolytes are substances whose rnolectales itr
solution dissociate to sorrle degree into ions. Strong
electrolytes dissociate completely for all practical purposes. We can not. therefore, write a diysociation uquiilibriium for theral. Strong ion, are simply always present
in the solution at whatever cc~ilcentration they were
originally added. Ac the] remain completely iiissociated, it fi~llowsthat they do not participate in any
reactions \albthin thc~soh.ifiotz.'They t-aaay react with other
entities such as ceI1 anennbranes, but these are outside
the solution. and we are concerned fc~rhe tnornellt only
with strictly isolated solution,.
In ma~nrnalianextracellular solutia~nc,the most concen!rated inorganic \tn,ng ion\ arc Na and Cl . K ' and
SO, are also strong ionc. but they are a l w a y ~present
in n~a~ch
lower concentrations than Na or CI . Sormaetirncs. Ca" and ~ g may
"
function as strong ions. but
they also occur at low concea~traaionsin coanparison
with N a ' and CI , so that we usually neglect them.
Lactic acid ha$ a dissociation constant tsn the order of
I 0 ' eyuiv ./I,, so that lactate behaercs as an organic
strong ion when present in plahnasa or cxtracelltalar fleaids. It Inlay be quantitatively signlficatat under some
circumstances. Intracellularly, the nlo\t iaamportant
+
171
loll-]= \/K:
+ ([Na']
-k
[CI ]1'/4
( [ N a ' ] - (Cl j)/2
These equations dell us that ( H ' ] ant1 (OH -1 arc functions of, or determined by, the values of both K,: and
([N,atl - [CI 1). If we consider K: to be a known
constant parameter, then ([Nat1 - [Cl-1) B3 clear11 the
quantity u hich determines the values of both [ H ) and
(OH ] in thin system.
Bcforc exploring in detail what these cquatio~lssay,
consider the quantity ( [ S a t ] - [Cl 1) in [ 5 ] , [(>I,and
[71. It is the rzcf strong icva positive charge, the concentration of strong cation positive charge in the solutioa~
which is Paot electrically balanced by strong anion negative charge. We know the vtllues of [ N a ' ] and [CI I
because we know how match of each wc atfded to make
up the litre of solution, so we know the vale~eof this ( )
term. If trther strong ions were present, such as K I or
lactate . their known concenta-ationswould also appear
within the ( ). and only there. both in thc electrical
neutrality equation, [ 5 ] .and in [6] and 171. It is therefore pernnissikle, and very convenient, to represent this
( ) quantity by a sl~ortsymbol instead of having to write
~ , ~
it out in detail each time. We call it the S ~ T O B PCIM
d[flirefzc-~.
ranti represent it by the symbol (SIIIj.
[SIfIJ is formally defiraetf as (the suan of all the strong
cation concentrations in the scplutiol~)nainxi (the surn of
all the strong araicpn co~lcentrationsin the bc~lutic~n).
Node that if strong anions exceed strong cations, then
[SID] will be ncgativc. [SID] slmcbuld be thoe~gtntof as a
ION SOLUTIONS
STRONG
CH+S
AND
C8H-1
"
"
0
VERSUS
CSIBI
0.2
9.8
0.6
0.4
STRONG I 8 N DIFFERENCE,
CSIDS
FIG. 1. Hydrogen ion ([H+]f and hydroxyl ion ([OH I) conccntratisns, for any strong ion solution, plotted against strong
ion difference ([SIDJ), over the [SID] range from - 10 "to + 1 0 ' equiv./l. When [SIII] = 0, [H'] = [OH 1 = VIE,
as
indicated by the arrow. Note the smooth gentle curves of transition from linear to nonlinear behavior on going through [SID]
= 0. K k = 4.4 x HO '"(equiv./l,)'. the value for blood plasma at 37"C, for this and all the following figures.
-0. &l@5
0.OG15
fl
181
IH']
191
[OH ]
\ / K ; + ( [ S I D J ~ Z )-: [SIDJ/Z
=
\/K:
+ ((SIDjjZ)' + [SID]/2
STRONG
CH+3
AND COH-3
LOG-LOG
7
ION SOLUTIONS
VERSUS
PLOTS
la-3
CSIDI
r CH+I , CDH-1
+18-5
STRONG I O N DIFFERENCE,
+18-3
CSIDI
RG. 3. Log-log plots of [H+3 and [OH-] versus [SIB] for any strong ion solution, over the [SIB] range from - B to
+ 1 equiv./l. The left side of this graph has the horizontal scale direction reversed, and plots Hog([H+j)and log ([OH-]) versus
log([SIB]). The slopes of the linear portions of the curves must be interpreted accordingly.
equations graphically, and provide a number of surprises. The algebra behind some of those surprises is
illuminating, and reinforces their significance.
First, KL is of the order of 18-l4 (equiv./~)', while
[SID] is usually on the order of lo-' equiv./L. In blood
plasma, for example, [SID] is normally close to 40
rnequiv./L. K:, is therefore negligible compared with
( [ S I D ] / ~ )in~ [8] and [9], so that the value of the whole
square root term will be +[SID]/2, regardless of the
sign of [SIDJ. In the rest of the equation, however, the
sign of [SID] makes a profound difference to the way in
which the values of [H'] and [OH-I depend on it.
Whenever strong acid anions exceed strong base cations, then (SID] is negative, [H'] is larger than [OH-],
and the solution is acidic. Equations 8 and 9 can then be
rewritten to a very good approximation (at least 1 : 10')
as:
for example) results in [ H ' ] increasing by x . It is important to understand that this happens because the
added Cl-- ions increase the magnitude of - [SID] by x,
and [H'] must be kept equal to -[SID], not simply
because of the added hydrogen ions. Similarly, adding
x moles per litre of Na' (as NaOH) will make -[SIID]
more positive by x, and result in [H' j decreasing by x,
even though no H t per .ye have been taken out of the
solution. This simple linear behavior of [PIil with
- [SID] can be seen clearly on the left side of the graphs
in each of Figs. 11, 2, and 3. i t only occurs whe~z[SiD]
is negative and the solution is acidic.
[SID] is positive, on the other hand, when strong
base cations exceed strong acid anions. In this case,
[OH-] is greater than [H'] , the solution is alkaline, and
to the same excellent approximation as before, 1131 and
(9) become
~ 1 1 3 [OH-]
-K:/[SIDI
In words, [OH-] is very small in a strong acid solution, and varies inversely with the value of (SID], while
[H"] is large and equal to the magnitude of [SID]. As
a result, [ H t ] seems to behave like a strong ion in such
acid solutions, even though it is rzot one. Adding x
moles of H' per litre to such an acid solution (as HCI
[13] [PI']
KL/[SID]
1450
[s~rq
Solution
Initial
Final
Initial
Final
No. I
No. 2
No. 3
0.041
0.001
-0.039
0.039
-0.001
-0.041
1.07X10"
4 . 4 lo-"
~
3.9X 10 '
B.13xlO''
1 . 0 lo-'
~
4.1 X 10
[A-j
K A x [HA]
[HI] x [OH-]
Change
Ii~itial
+ 5 . 5 ~ 1 0 - ' ~ 11.99
1 .OX 10 '
10.36
t 2 . 0 BBB
~ '
1.41
Final
Change
11.95
3.00
1.39
-0.02
-7.30
-0.02
1141 [H']
pH
[M'J
A:,'
[H+]'
+ ( K , + [SID]) x
[IJ
+ I "
+ (K,$ x dlSIDl
X
[H']
[AloTl)- KGB
Kh x K: = ( I
WEAK
ACID SOLUTIONS
H VERSUS
STRONG
ION DIFFERENCE,
CSID3
&SID3
Fmc;.4. Combined plots of [H+] and [OW-] versus [SIDI (Icft side) and pH versus ISID] (right side) for a weak acid solution
with [ATC)T]= 0.019 cquiv./L and K A = 3 x 10 ' equiv./k. Thc vertical dashed lines are at [SID] = [ATOT]= 0.019 equiv./L,
and [SID] = [ATOT]/2= 0.0095 equiv./l. as indicated on the right hand graph.
[OH-] F O ~ I( [SID1 [ATOI 1, Kd K A )
[ H t l = Flr([SID], [ A L o ~KL,
~ , KA)
ACID SOLUTIONS
WEAK
[HAS
AND
-0.01
STRONG
CA-I
VERSUS
0. 0%
CSIDI
0. 02
RG. 5. Undissociated weak acid concentration, [HA], and weak acid anion concentration, [A -1. plotted against [SIDI, for
the same solution as Fig. 4. As indicated, the dashed vertical lines have the same meaning as in Fig. 4.
pH, and causes both the pH and the [H ' ] of the solution
to change more rapidly with [SID] than they did before
the weak acid was added. The effects on (OH-] are just
the opposite. From these results we may conclude that
when [SID] is positive, a weak acid solution "buffers"
[OH-], but not [H'] or pH.
Figure 5 shows plots of [HA] and [Ap] versus [SID],
and demonstrates the surprisingly linear behavior of
[HA] and [A-] with changing [SID]. Note that at the
[SID] = [AToT]/2 point, [HA) = [Au]. From 1141 it
also follows that at this point [H ] = KA,or pH = pKA.
More detailed analysis shows that these conclusions are
only valid so long as KA is within about three orders of
magnitude of Kk. Outside those limits, the [HA] and
[A-] curves are not linear, and do not intersect at the
[SIE)] = [ATm1//2point.
Figure 6 shows a log-log plot of [H'] versus [SID],
to compare with Fig. 3 as well as with the [ H t ] and pH
curves in Fig. 4. It also brings some surprises. The
change in [ H t ] is much steeper over the "buffering"
range near [ATOT1/2than in strong ion solutions, rather
than shallower, as we might have expected from the
term "buffer." This is not merely a distortion produced
by the double log plot, but is consistent with the [PI']
cunre in Fig. 4, which show that [H'] does change
rapidly over this region. The distortions of the inverted
semilog or pH plot in Fig. 4 are the more difficult ones
to understand!
+
WEAK A C I D
EH+I
-10
-2
-10
AND
-4
STRDNG
-IB
-8
[OH-1
-10
-8
SOLUTIONS
CSIDI
VERSUS
18
-a
H ON DIFFERENCE,
18
-8
18
KS I87
-4
10
-2
(esuiv.%b>
FIG.6. Log-Iog plots of [#'I and [OH-] versus [SID] for the same solution as in Figs. 4 and 5, to compare with Fig. 3.
The vertical dashed line is at [ A T ~ T ] / as
~ . indicated.
[H2C031 = Sg12co,X
SH,co,is about 1.5 x
[2]
[Ht] x [OH-] = Kk
Weak acid dissociation equilibrium:
[I4]
iHt]
= K , iHA]
Conservation of mass for "A":
[HCO,] = K2 x [H2C03]
[HCO,] = Kc x pC02
1454
K3 x [HCO;]
Electrical neutrality:
[HA1
BA-1
=
=
BHCQ_II = ~ ~ ~ c o , ( l s I[ATu~Y,
Dl, P C O ~
1co:-1= Fc,,([SIDl,
[OH -1
lHSI
(Am17 pCO2)
FOFI
([SIDl , [AIoTI.p C 0 d
FII([SIDI, lA~o1.1,pCOa)
+ ([SID] + KA) X
[H']~
- Kk - Kc X pCO2)
X [H']'
(KL + Kc x pCOz) - K3 X K, X pCO2)
x [H'j - K, x K, x Kc X p C Q z = O
([SID] -
STANDARD PLASMA
STANDARD P L A S M A
IN+! VERSUS p C 0 2
p H VERSUS p C 0 2
fluids. pCOz therefore changes rapidly and often. Concentrations of strong ions in body fluids, on the other
hand, are detemined by absorption frorra the gut and
reinoval by the kidneys, so that changes in [SID] only
occur slowly, on the order of hours. Proteins are the
only significant weak acids m
i plasilsa, and their total
concentration is determined priinarily by the Iiver.
Changes usually occur only over pcriods of days. It is
therefore physiologically reasonable to examine the dependent variable behaviors in this system by plotting
against pC02, with [AToTl fixed, and [SIT>] set to a
series of values. Figure 7 shows ~ u c ha set of curves for
m'1.
Qualitatively, these curves tell ais that [H ] increases
as pCOz increases, so that C 0 2acts as an acid. [H'] also
increases as [SID] decreases, which is what we shoulel
expect frotrl the definition of [SID].
One striking feature of these curves is that their cur-
EFFECTS O F
LSIDI
STANDARD P L A S M A
CH+3
VERSUS p C 0 2
STANDARD PLASMA
p H VERSUS p C 0 2
i
-
[AgOg]changes, or both. This quantitative analysis permits us to separate and understand these two kinds of
changes in new and very useful ways.
Thus, Fig. 7: or Fig. 8, combined with Eq. 24, provides the means to monitor a particular subject's
acid-base status quantitatively and precisely. If pH and
pC02 values are available for a plasma sample, and if
the subject's plasma protein concentration is known to
is, then the position of the
be near normal, so that [ATOT]
[H'] -pCO, point on Fig. 7, or the pH-pC02 point on
Fig. 8, tells us immediately the effective value for [SID]
in that plasma. If plasma electrolyte values are also
available, so that the inorganic component of [SIB],
([Na'] + [ K f ] - [Cl-]), is known, then the difference
between that measured inorganic component and the
effective [SID] value from the graph is a measure of the
unidentified strong ions which must be present in that
plasma. In "pure" lactic acidosis, for example, this
difference should be just equal to the lactate concentration in the plasma. In a case of diabetic acidosis, this
difference should agree closely with the measured value
for the total keto acids present.
S T A N D A R D PLASMA
CHC03-3
50
E F F E C T S OF
I
STANDARD PLASMA
VERSUS pC02
I
CSIDI
EFFECTS OF pCO2
0
'
LOG pC02
FIG. 11. [HCO,] plotted against pCOz for the Standard
Plasma model. Compare with Fig. 7.
FIG. 12. Log-log plots of all six dependent variables versus pCB2 for the Standard Plasma model, with [SID] held
constant at 42 mequiv./L. K values and [AFOT]the same as in
Fig. 7. Note the decrease in [c@-]
with increasing g o 2 .
The ticks on the horizontal axes between 1 and 2 indicate the
normal operating pC02 of 40 mmHg.
INTERSTITIAL FLUID
STANDARD P L A S M A
EFFECTS OF fSIDl
CH+I
100
VERSUS p C 0 2
-21
- 5 -5 - 4 -3 - 2 - 1
LOG
6SIE3
FIG. 13. Log- log plots of the salnc six dependent variables as in Fig. 1 2, versus [SID], with pC02 held constant at
41 m ~ n H g The
.
tickc on the [SID] axes between - 1 and -2
are at the nonnal plasma operating [SID] of 42 mequiv./L.
FIG. 14. [H'] plotted against pCO,, for the Standard Interstitial Fluid (ISF) mc)del, with [SID] values indicated on the
curves. Same K ' s as in Fig. 7 , but [AT(n] set to 0 throughout.
Temperature e&cts
Temperature elfects on dependent ion concentrations
in isolated sc~lutionscan be calculated easily by this
quantitative approach. We nced only to know how the
various Ks change with te~nperature. All the curves
presented in Figs. 1 through 17 will appear in different
positions when different temperatures, and therefore
Ks, are used in the calculatic~nswhich produce them.
In strong ion solutions, only KI, is involved. As [ 11]
and 1131 show, Kk will affect only [Ht] in alkaline
solutions and only [OW-] in acid solutions. In weak
acid solutions with no C 0 2 , both KL and K,, will change
with temperature, but their effects are only significant
over the [SID] range from 0 to [AT,lT].In solutions with
C 0 2 , all four Ks must be changed when tennperature
changes. K values at different temperatures are available in Robinson and Stokes (19591, and in Butler
19823.
Physiologically, interpretation of temperature effects
is made more ccsmplicated by changes in inetabolic and
cardiovascular f~incticanwhich alter independent variables, notably pC02 and ISID], by surprisingly large
amounts. Equation 25 still applies and still enables us to
calculate the resulting [H' ] changes, but it can not
INTERSTITIAL FLUID
pH VERSUS p C 0 2
E F F E C T S 8F
E F F E C T S QF p C 0 2
CSIDl
L O G pCOZ
Frc;. 15. Same values as in Fig. 14, plotted as pH.
14641,
INTERSTITIAL FLUID
EFFECTS OF CSIDI
c(Z
LOG CSIDl
FIG. 17. Log-log plots of the 4 dependent variables in the
Standard ISF model versus [SID]. pCOz held constant at 47
mmHg.
ACID-EASE
BALANCE
VENTILATION
CA - l
[HA3
CC03=3
CQH-I
LH+1 CpH>