For years brewers have used calcium carbonate (chalk, limestone) to neutralize excess malt acid in dark beers. Because it does not dissolve in water they usually add it to the mash directly. More recently it has been discovered that it is not as effective at raising mash pH as it ought to be given the amount of carbonate a given amount of chalk contains. The reason this is so is that a mash contains lots of phosphate – in fact we rely on that phosphate to lower mash pH by reacting with calcium and releasing acid. When calcium carbonate is added to mash its calcium ion also reacts with malt phosphate to precipitate apatite and release acid (protons). A carbonate ion can absorb 2 protons. The precipitation of each calcium ion releases, on average, 1.4 so 1.4/2 = 70% of carbonate’s proton absorption capacity is utilized by protons from apatite precipitation leaving only about 30% available for the job it was intended to do. In this note – preliminary and incomplete at this posting, we explore in some detail how this works.
For best results in brewing mashing must be carried out in a narrow range of temperature and pH in order for the enzymes at play to operate properly. In most cases some acid must be added to the mash, either in the form of mineral or organic acid from a bottle, lactic acid from sauergut or sauermalz, dark malt or a combination of any of these.
Dark beers originated where the water contained high levels of alkalinity from dissolved calcium carbonate (chalk, limestone). Were this alkalinity not neutralized with acid pH would stay too high had the brewers not used the dark malt. Sometimes a brewer has the opposite problem of the original brewer. His water is not alkaline enough on its own to neutralize the acids in the dark malts but he wants the flavors that the dark malts confer. As his water won’t balance the acids in these he must do it by other means either adjusting the water to have alkalinity more like the water with which the beer was originally brewed or by adding alkali to the mash. As calcium carbonate is the natural source of such alkalinity it seems natural to use it to derive any extra alkalinity needed by a brewer and chalk is sold in every home brew shop in the world. There are some problems with it, however. It isn’t very soluble. When it dissolves in nature the process starts with carbon dioxide which is found in the air at very low levels (0.03 – 0.04%) and in the soil at levels perhaps 10 – 100 times this with the source being the respiration of bacteria. Carbon dioxide reacts with water by
to form, on the right, carbonic acid. It is this acid which makes the dissolution of limestone possible because it, as does any acid, give off a proton in the process converting to bicarbonate ion:
Calcium carbonate is not very soluble but a little does dissolve and in so doing separates into calcium and carbonate ions
If carbon dioxide has dissolved and some of the resulting carbonic acid has dissociated the proton from that process is available to the carbonate ion and converts it to a bicarbonate ion
It is the product of the concentrations of calcium and carbonate ions that limit its solubility
in which (calcite, 20 °C) is the solubility product limit and the bracketed quantities are the molar concentrations of the two ions. If carbonate is removed from the solution then clearly the ion product becomes less than K and more can dissolve. At the same time limestone is being consumed by conversion to bicarbonate, carbonic acid is being consumed by conversion to bicarbonate. If carbonic acid is consumed, neutralized, the equilibrium between in the air and carbonic acid in the solution is upset causing more to dissolve in order to replenish the carbonic acid lost to bicarbonate conversion. The overall process is one of flow of ‘carbo’ from the air and from limestone into the bicarbonate pool. An overall equation can be written
Natural water at normal pH contains mostly bicarbonate ions. As pH rises the shift is towards carbonate and as it is reduced the shift is to carbonic. The ultimate control is the partial pressure of to which the water is exposed. A more accurate representation of what happens would be
In this equation the three ’s are fractions which sum to 1
The individual values for depend on pH. Calculation of them is discussed here and curves showing their value as a function of pH can be seen on the alkalinity measurement page here. At pH 7 , and . At pH 8.32 and
The problem with which the brewer is confronted is that he must, if he wishes to attain water with chalk derived alkalinity with natural quality, supply carbon dioxide. This can be done by bubbling through the water or by putting the chalk in a sealed container which is then pressurized by . This is certainly doable but requires some effort. There is one very great advantage to this method (beyond getting water which resembles nature’s) and that is that by controlling the amount of the pH can be set to any desired value and the values of the ’s thereby controlled.
An obvious question is as to whether there is an easier way and there is: use an acid other than carbonic acid. Any acid provides the protons necessary to neutralize carbonate (convert it to bicarbonate). For example sulfuric acid:
A potential difficulty here is that the brewer may not want sulfate in his water. He has, of course, other choices for acid such as hydrochloric, lactic and phosphoric but each of those will leave behind one mEq of its anion for each mmol of carbonate converted to bicarbonate.
We want to brew a dark beer using acid malt and our water has low alkalinity so we must add a base, which is a proton absorber, to vacuum up the protons released by the dark malt. Carbonate ion is a suitable base. Each carbonate ion can absorb 2 protons. First:
and then, as bicarbonate is also a base:
In a natural water, as noted above, most of the carbonate has been converted to bicarbonate so under normal brewing conditions where dark malt acid ‘s protons are being absorbed they are being absorbed by bicarbonate. So we must obtain bicarbonate in our water but this can be difficult or at least bothersome and requires the use of acid. But we are trying to get rid of acid in the first place. It seems counter productive to have to add acid just to get the substance we want to use as a proton absorber into solution and in so doing we are using half the proton absorbing capacity just to get it to dissolve. An obvious solution to this is to add the carbonate directly to the mash.
Adding Carbonate to Mash
For quite a while now people have been advocating adding chalk directly to the mash because it is not soluble in water. The mash is acidic and thus should dissolve the bicarbonate releasing the carbonate ion which captures protons or, in other words, is a base. Quantitative experiments showed that pH increase from chalk additions were not what they should be.
pH is defined by
The entity in brackets is the hydrogen ion activity which we can state is the same as the hydrogen ion concentration with out incurring appreciable error. In pure water the hydrogen ion concentration is approximately 10-7 moles/L and the pH is, accordingly, approximately 7. In an overly acid mash (or other solution) the hydrogen ion concentration is greater than 10-7 and we wish to remove some. To do this we add a base such as bicarbonate ion:
In this equation we have 0.001 mole/L hydrogen ions so the pH is 3. Each carbonate ion has two negative charges and so attracts 2*0.0004995 moles of protons per liter to become 0.0004995 moles per liter of carbonic acid which decomposes into CO2 gas and leaves the solution. What is left is
0.01 – 2*0.0004995 = 0.000,001 moles/L H+ and so the pH is now increased to 6.
It is unusual, but completely valid, to write chemical equations with fractional coefficients as we just did because the coefficients in that expression represent moles per liter. If we wanted an equation in terms of molecules and ions we could just multiply by 107 and get
From this way of writing things it is clear that each carbonate ion absorbs 2 protons but this is a simplified equation. In fact carbonate reacts thus
Because it is clear that the number of carbon atoms on each side of the equation is the same (1) and it is also clear that if, as in our discussion up to this point, all the carbonate is converted to carbonic so that then each carbonate would absorb 2 protons but that only occurs at pH < 4. As another example, if we add carbonate to a solution at pH 8.32 where and each carbonate would absorb (on average) 0.99 protons.
Role of Phosphate
The problem with adding carbonate to the mash is that it doesn’t absorb nearly as many protons as we would hope because malt contains a lot of phosphate and the compound formed from calcium ions and phosphate ions (hydroxyl apatite) is very insoluble compared to calcium carbonate. When calcium and phosphate ions meet apatite precipitates and in so doing releases quite a few protons. These are equally attractive to a carbonate ion and so it winds up neutralizing mostly these protons and not the mash acid protons we want it to. A carbonate ion’s effectiveness as a base (proton absorber) is 30% or less of what it would be were phosphate not present.
We need to present equations similar to (1) and (2) for the phosphate system. We will start with the acid and work toward phosphate (we did it the other way round with carbonate) because we want to emphasize that phosphate is a source of protons, not a sink for them as is the case with carbonate.
Just as is the case for the carbo system, except that there are 4, there are a set of mole fractions, which we shall call , totaling 1, which describe the distribution of phospho species among the four forms it can take. As with carbonate there is a tiny amount of the completely deprotonated species present at any nominal pH. But calcium phosphate is very much less soluble than calcium carbonate expressible through:
At a particular pH if a small amount of calcium carbonate is added to a solution containing sufficient phosphate the following reaction will take place:
Buried in all this is the fact that 10 calcium carbonate molecules release 10 carbonate ions which react with ions containing a total of 6 phosphorous atoms and two hydroxyl ions to form one molecule of hydroxyl apatite while releasing a number of protons that depends on the pH. Some of the 10 carbonate ions absorb protons to become carbonic acid molecules and bicarbonate ions with the relative proportions being determined by pH. The remaining carbonate ions are available to absorb protons from acids in the mash. Even if there are no bases besides carbo and phospho species nearly all the protons will be absorbed by bicarbonate, carbonate and mono, di and tribasic phosphate ions with the result that the pH rises. But clearly it rises less than if all those protons were not released. It is this that makes chalk appear to be less basic than it is when phosphate is not present. Ten molecules of calcium carbonate would be able to absorb 20 protons. When phosphate is present the number is closer to 6. The graph below shows protons which can be absorbed by 10 carbonate ions as a percentage of the 20 which could be absorbed were it not for the protons released when 10 calcium ions are precipitated with phosphate as a function of the solution’s pH. In the region spanned at one end by the overly acidic mash pH’s we might encounter from too much acidic malt and at the other by the pH’s we would most likely want to reach by alkali addition the effectiveness of chalk is seen to be 30% (6/20) or less.
Lime is often recommended as a substitute for chalk. The red curve on the figure shows the effectiveness of lime , , for this application. Note that at higher pH’s it becomes appreciably more effective (because the hydroxyl ion is a much stronger base than the bicarbonate ion) but that at the range of nominal pH interest it is not that much more effective. Where lime has a significant advantage is in the speed at which it works as we shall see.
Reactions involving carbo species are slow. It can take hours for carbonate/bicarbonate to absorb protons whereas hydroxyl ions do it in seconds or less. This is doubtless the biggest advantage to using lime as in both cases the added alkali does not enter the beer. In the case of chalk the calcium precipitates as apatite and is removed by lautering while the carbonate gets converted to CO2 and is expelled from the mash as gas. With lime the calcium again precipitates as apatite but the hydroxyl ions absorb protons to become water.
The graph below shows pH recordings following additions of aliquots of a chalk slurry to water containing monobasic potassium phosphate at about the level one would expect to find dibasic phosphate in a nominal mash. The red trace is the calibrated pH (the grey the meter's uncalibrated reading) and the green is the temperature. As pH is proportional to the number of protons in the solution it will rise as protons are absorbed. Keep in mind that the electrode does have a finite response time but that most of a pH change is reflected within a minute or two. Thus a steep step in the red curve indicates that protons are absorbed immediately upon addition of the chalk slurry. A gradual rise over hours indiates slow consumption of protons.
The aliquots were of different sizes intended to produce noticeable changes in pH (in which we didn’t clearly always succeed) but that is not what we want you to look at here. Rather focus on the length of time it takes the reactions to occur. At lower pH’s it occurs quite quickly. pH settles at it’s new value within minutes. At higher pH’s it takes over 3 hours for an addition to be fully effective.
The graph below is similar but in this case shows what happens when lime, Ca(OH)2 is used instead of chalk. One sees clearly that the reaction times are much faster than with chalk. As was the case with chalk we were adding aliquots of a suspension of lime in water thus, as with chalk, the particles have to dissolve and react. It is also interesting to note that at around pH 5.8 the same phenomenon in which protons are being released by the reaction faster than they are being absorbed occurs. We als not thet there is some slowing of reaction times with lime at higher pH but that it is not nearly as pronounced as is the case with chalk.
Keep in mind that there are multiple reactions going on here simultaneously each with its own rate constants. We do not know what the rate limiting reaction is. It is most interesting to observe that at about 22:30 protons are clearly being released faster than they can be taken up as the pH actually drops somewhat after that time.
In the Role of Phosphate section above we presented evidence based on theoretical calculations of the number of protons which could be absorbed by carbonate ions in the situation where a solution is saturated with respect to apatite. In the graph below we have plotted pH values (circles) taken from the lime data (graph just previous to this one) starting with the first data point after which the solution became visibly turbid. In other words we are certain that it is super saturated with respect to apatite. We have also plotted (squares) the expected pH for a solution with this much phosphate and an eqivalent amount of sodium hydroxide i.e. one in which the precipitation of apatite would not occur and there would be no release of protons from that effect. All added hydroxyl ions would be available to absorb protons. The slopes of the two curves are indicated on the plot. The slope for the lime curve is seen to be 33.7% of the sodium hydroxide curve thus lending further support to our thesis. We reason that if a given addition of hydroxide ought to produce a pH change of ∆pH but turns out to produce a change of 0.337∆pH then that addition was only 33.7% as effective as it ought to be.