STUDIES ON GUSHING I
INTRODUCTION
by N. S. Curtis, B.Sc, and L. Martindale, B.Sc.
(Whitbread & Co., Lid., Chiswell Street, London, E.C.I)
Received 7th April. 1961
Some of the factors which affect gushing in beer have been examined, particularly in relation to one particular beer, for which no reliable artificial method of removing the tendency to gush appears to be available. At pH values of about 3-3, gushing beer loses its ability to gush even when shaken but, when the pH is restored to normal, shaking will again induce gushing. The need for more knowledge of the brewing factors involved is emphasized.
Introduction
Published work on the problem of gushing beer was reviewed by Beattie in 19511 and by Thorne & Helm7 in 1957 so it is not appropriate to cover this ground again. However, the published work does not seem to explain all examples of gushing, and some instances would seem to require a full-scale research investigation if a satisfactory explanation is to be found. The work now being reported was initiated for this reason; the impetus for its continuation is being provided not by severe outbreaks of gushing, but by the realization that our understanding of the problem is far from complete and, should a severe outbreak occur, no certain cure would seem to be available.
This, and subsequent papers in this series, will be concerned with the problem of gushing beer in a particular brewery. Naturally, this problem is being considered in relation to what has already been published, but the justification for these papers is that in the course of this work some new information has been obtained which might be of more general application.
Incidence of gushing.—In this brewery, incidence of gushing is fortunately rare, and large quantities of beer have never been involved. Over the years, most of the various types of bottled beer have shown some tendency to gush, but the bottled brown ales have given rise to most trouble and the pale ales to least.
Behaviour of gushing beer.—Everyone is familiar with the effect gushing produces, and it is obvious that many bubbles of carbon dioxide are formed in the beer. It has been suggested1 that this evolution of gas may be initiated by a rough bottle wall or by sediment in the bottle. To examine this point, some slow motion cine films have been taken of gushing and non-gushing ale. A record has been obtained of what happens in a bottle of gushing beer when the stopper is released. It is quite clear that there is an immediate formation of small bubbles throughout the bulk of the beer, and it is not possible to associate any special features with the wall or bottom of the bottle. The small bubbles formed throughout the bulk of the beer rise in such a manner as to leave a clear line of demarcation between “bubbly” beer and the clear beer left behind the mass of bubbles. There is no straggling effect, and very little bubble formation occurs once the main effect is over. The bubbles which rise later are larger.
Other bottles of beer, which from the trade point of view would have been considered quite satisfactory as regards gushing, have also been examined. In some of these the release of the stopper again produced a spontaneous release of bubbles throughout the beer, but the number of bubbles was too small to do more than cause some foam to rise in the neck of the bottle. In these cases, where fewer bubbles were formed, the bubbles appeared to be much larger than those formed in bottles of gushing beer. In one example, a selected frame of the film was enlarged so as to enable the bubbles to be counted. The number was estimated to be 2,600 per ml. A repeat determination on another bottle of the same beer gave a value of 2,300 bubbles per ml. This is only a very approximate estimate, and the number of bubbles per ml. of gushing beer would obviously be far greater, but the figures do perhaps serve to show that the bubbles are relatively few in number and, as the nuclei have been reported to be far smaller than anything which is visible microscopically,7 the chances of seeing them even by means of an electron microscope seem to be exceedingly remote.
In the examination of some other bottles of non-gushing beer it was noticed that practically no bubbles were released when the bottle was opened.
Brewers naturally consider gushing beer to be atypical—as most beer does not gush, but from the scientific point of view it would seem to be more accurate to consider such beer to be an extreme example of a fairly general phenomenon. It might be better to regard beer which produces no bubbles when the bottle is opened as being atypical.
Nuclei and bubble formation.—Among the many diversities in the literature, one commonly accepted fact shines out. Krause5 first pointed out in respect of gushing beer that bubble formation requires the presence of a nucleus or active centre. The size of this nucleus varies inversely with the degree of super-saturation of the liquid with the gas and with respect to carbon dioxide in beer the size of the nucleus does not reach molecular dimensions until the degree of super-saturation is much higher than is normally encountered; in beer of normal carbon dioxide content, the size of the nucleus is said to be of the order of I0-7 cm. Although there is general agreement that nuclei must be present, there is a diversity of opinion as to their nature, but to understand the nature of these nuclei will be to understand gushing.
Suspended particles in beer.—Vogel8 linked gushing with the presence of jagged particles in beer, the particles being visible only in the electron microscope. Samples of our gushing beer have been subjected to electron microscopical examination. Sediments obtained by centrifuging gushing beer at 20,000 G have been examined, but particles common only to gushing beer have not been found. From time to time it has been suggested that particles of protein and other matter present in bottled unfiltered stout may cause gushing, but no evidence has been obtained that the suspended matter in gushing stout differs either in quantity, or quality, from that in normal stout.
Examination of almost any freshly-bottled beer provides convincing evidence that even large particles suspended in the beer do not necessarily give rise to bubble formation. Thus, many bottles of filtered beer contain fibres and other particles in suspension which are readily visible to the naked eye. When the bottle is opened, a stream of bubbles is evolved from some particles, but not one single bubble from others. Clearly, some forms of suspended matter in beer are harmless in this respect.
Experiments With Gushing Beer
Krause suggested that the nuclei essential for bubble formation and therefore for gushing might be formed by the aggregation of colloidal substances in beer, this aggregation being brought about by shaking. He conceived a surface-active layer forming on the bubbles, this layer becoming compressed and more stable as the size of the bubble diminished. On the bottle being opened, this compressed bubble would expand. This suggests a fairly unstable nucleus the components of which might be readily dispersed. It is also possible that the compressed surface-active aggregate could act as a dry spot on which bubbles might form.
In contrast, Vogel’s jagged particles8 and Brenner’s minute crystals of calcium oxalate2 seem much more stable and could be expected to be less readily dispersed.
Stability of the nuclei.—The nuclei present in the beers which we have examined were very unstable as is shown by the results of the following experiments:—
(i) A bottle of gushing beer was opened and the contents were collected. The beer was filtered through a Whatman No. 12 paper to remove most of the carbon dioxide; it was then chilled and re-carbonated in a laboratory carbonator, consisting of a stainless-steel cylinder with wash-bottle-type fittings, chilled beer in the cylinder being subjected to an appropriate pressure of carbon dioxide, while the cylinder was shaken; the carbonated beer could readily be transferred under counter-pressure to bottles or other containers. The re-carbonated beer was allowed to stand at room temperature overnight before being opened, when, it was found not to gush.
(ii) A bottle of gushing beer was opened and about half the contents shot out. The bottle was then filled by adding, under counter-pressure, non-gushing beer. This mixture did not gush. The portion which was initially ejected from the bottle was collected, mixed with some non-gushing beer and re-carbonated. This mixture did not gush.
(iii) The screw stopper on a bottle of strongly gushing beer was released just sufficiently to allow foam to reach the stopper. It was then tightened. When the foam had collapsed the operation was repeated two or three times. This involved little escape of gas. When finally the bottle was opened normally the beer hardly gushed at all.
(iv) A bottle of gushing beer was kept at 100° F. for 24 hr. On being opened at that temperature it did not gush.
Inhibitors.—It has been suggested that non-gushing beer may contain an inhibitor. To examine this possibility, a bottle of gushing beer and a bottle of quiet beer from a different brew were chilled and half the contents of each was transferred under counter-pressure to another bottle. The mixed beer was left overnight at room temperature before being opened. It did gush, thereby suggesting that the quiet beer did not contain an inhibitor.
Effect of heat.—Experiment (iv) above gave results which are in agreement with some observations on the effect of pasteurization. However, in certain undefined circumstances, high temperatures seem able to favour gushing. Thus, a certain stout in a tropical market has given rise to frequent complaints of gushing, but the behaviour of this gushing beer seems to be different from gushing beer infrequently encountered in the home market. In the latter case, gushing is characterized by a fairly sudden and brief ejection of beer from the bottle. In the tropics, it appears that gushing is not so violent but is more prolonged so that, if the bottle is not immediately poured, there is a substantial loss of beer.
Effect of cooling.—Cooling to temperatures below 40° F. has always been found by us to be associated with a quietening of gushing beer, though a rise in temperature will restore the defect. It should also be noted that bottles which are shaken at low temperatures develop gushing, as do bottles shaken at room temperature, but the defect will not be noticed unless the bottles are opened at room temperature.
Effect of shaking.—Krause8 and all subsequent workers have emphasized that shaking is a most potent agent for inducing gushing. Schmith6 has drawn attention to the necessity of having an interface if shaking is to induce gushing, and Thome & Helm7 have pointed out that beer in completely filled bottles will not gush under conditions of agitation that would cause beer in normally filled bottles to develop gushing. We have made similar observations in experiments in which quart bottles, filled to varying heights, were transported for several hundred miles before being opened. Beer in bottles which had been completely filled did not gush, and beer in bottles which had been filled to the collar, thereby leaving only a little gas space in the sealed bottle, gushed only to a slight degree. Beer in bottles, which had been filled to the normal height, or which had been deliberately short-filled, gushed considerably.
Natural and induced gushing.—Although shaking can induce gushing in beer, and although many trade complaints have referred to beer which had obviously received considerable agitation in transit, it has not been possible to prove conditions of excessive shaking in all trade complaints of gushing. It has also been very clear that many beers, which gushed in various stability tests which involved shaking, were quite stable in trade. This raised doubts as to the usefulness of shaking tests as a means of distinguishing potentially stable from potentially unstable beers. Many reports from trade involving many different beers revealed the fact that some seemed to be exceptionally stable. These beers when examined in various shaking tests gave practically no signs of gushing. It was, therefore, clear that beers which were quite stable even when shaken severely were commonly brewed and could be accepted as a reasonable standard for the comparison of other beers. Consequently, it is the aim of this work to examine factors which will lead to the production of a beer which will show no signs of gushing when vigorously shaken for a prolonged period.
Standard shaking test.—The method of Thorne & Helm7 has been adopted. This involves rotating bottles on a shaker at 30 r.p.m. for 24 hr. at room temperature. The bottles are allowed to stand for 1 hr. before being opened. The volume of beer expelled from the bottle has been taken as a measure of the intensity of gushing. Generally speaking we have used quart bottles in this test, as these bottles are associated both in trade and in this shaking test with a higher degree of gushing which is out of all proportion to the larger amount of beer involved.
Cures for gushing.—Several substances, some of which have been mentioned in the literature, have been added to beer which was subsequently subjected to the shaking test. Cobalt7 added at 2 p.p.m. and ethylene-diamine tetra-acetic acid3 added at rates from 20 to 60 p.p.m. were without effect. (In parenthesis it should be mentioned that the nickel content of many beers examined in this work did not exceed 0-3 p.p.m. and therefore cannot be considered exceptionally high). Pepsin and papain were separately added to beer at the rate of 3-6 g. per brl. and the beer was kept for 24 hr. at 37° C. before being shaken at room temperature; both enzymes had no quietening effect.
The colloids Manucol SS/LM, Manucol SS/LD and Manucol Ester EA/KM have separately been added to beer at a concentration of 0-1%; although the beers gushed after shaking, the amount expelled from the bottle was only about one-tenth of that expelled from the control beer. Gum arabic was without effect when used at 0-l–0-8%. Pasteurization was not able in these experiments to suppress gushing induced by shaking.
Adsorbents.—The aim of this work was originally to see if the substances which cause gushing in beer could be removed in a manner that would lead to their being concentrated and at least partially separated from other substances so as to facilitate their eventual identification. Adsorbents seemed likely to achieve this end.
The treatment was effected by adding the reagent to bottled carbonated beer in quart bottles which were then shaken as in the shaking test. To test the effect on induced gushing, the beer was then filtered (Whatman No. 12), carbonated and subjected to the shaking test.
Charcoal at 0-1% was without effect, but when it was used at 0-6 or 1% the treated beer could not be induced to gush. Charcoal removes a wide range of materials from beer and therefore was not suitable for the purpose in mind, various other materials were examined. Silica gel, Deglutan (a proprietary aluminium silicate), and cellulose powder were all used at 1% and were found ineffective, but fuller’s earth, Tansul (a proprietary bentonite), kaolin and activated alumina when used at the same rate were quite effective in removing the causes of induced gushing. Kaolin and alumina were of particular interest, as they were without effect on either the head retention or flavor of the beer.
Kaolin, being cheaper than activated alumina, has been examined further with a view to seeing if a commercially acceptable process of treating potentially gushing beer could be devised. Beer after fining and carbonated beer just before filtration have been shaken with kaolin and generally there was a positive response to the treatment. In these experiments there were two controls. One was prepared by carbonating the beer, if necessary, and then subjecting it to the shaking test; all samples gushed in this test. For the other control the beer was subjected to all stages of the kaolin treatment except that the kaolin was omitted; in other words the beer as received was put into quart bottles, shaken for 24 hr., filtered, carbonated and then submitted to the shaking test. In the 16 experiments carried out, only one beer gushed after this treatment and that only a little. When kaolin was added, results were similar except that the one beer which gushed after kaolin treatment was not from the same brew that gushed in the second control test. These results are curious, as they imply that the kaolin is not essential for this treatment to be effective. On the other hand filtration alone cannot be considered a means of removing from beer the substances which cause gushing. Most gushing ales have been filtered, and in the laboratory many beers which have gushed when they have been subjected to the shaking test have in the course of preparation been filtered through Whatman No. 12 paper. The explanation of these results may be that filtration of beer can remove the substances which cause gushing if these substances have first been made to aggregate into the nuclei which are so important a feature of gushing. The function of material such as kaolin or activated alumina may be no more than to aid the formation of the nuclei. This conclusion is perhaps supported by the fact that no means has yet been found of desorbing the substances which at one time it was thought might have been adsorbed onto the kaolin. It is also interesting to note that filtration through a bed of alumina gave a beer which gushed in the shaking test, whereas the same beer when shaken with alumina was quiet in the shaking test.
In considering how to make use of these observations in commercial practice, it is necessary to find an alternative to the prolonged shaking. Stirring (with kaolin) both on a laboratory scale and on a 12-gal. scale has been tried, but the proportion of failures was too high to justify further development.
Survey of various brown ales.—Half-pint bottles of 12 different brands of brown ale have been subjected to the shaking test. This examination was repeated on four different occasions, six bottles being examined each time. Six of these beers were associated with some degree of gushing, whereas the other six beers were generally quiet. Thus, although gushing as a trade defect may happily not be prevalent, a number of beers would seem to be potentially unstable.
Influence of pH.—Beers of fairly different pH values can be induced to gush, so that pH—within the range common to beer—cannot be considered to be an important factor. Some experiments have now been carried out to observe the effect of reducing the pH of beer to abnormally low values. It was found that beer which could be induced to gush at normal pH values (3-9—4-1) suddenly lost this property at pH values below about 3-3. Restoration of the pH to its normal value also restored the ability of the beer to gush when shaken. This effect has been observed with three different brown ales, and has also been observed to occur in beers which were the subjects of trade complaints as well as in beers which only gushed when shaken.
This observation is naturally of no practical importance, but it does perhaps give a little information on the nature of the substances involved in gushing. These substances would appear to have an iso-electric point at a pH of about 3-3. This is a low value for proteins, and does perhaps suggest that if the substances which cause gushing are proteins or peptides they must be of an acidic nature.
Conclusions
The examples of gushing beer which have been investigated provide additional support for Krause’s hypothesis5 that gushing is due to the presence in beer of nuclei or active centres for the formation of bubbles of gas. These nuclei, Krause suggested, may be composed of colloidal organic material which has been caused to aggregate. A powerful agent causing the formation of these nuclei from then- unidentified precursors is shaking.
Many beers which do not gush do give rise to the evolution of some gas bubbles when the bottle is opened. These beers are only different in degree from gushing beers.
Of greater distinction are those beers which even after severe shaking give rise to practically no bubbles when the bottles are opened. In these beers there would seem to be an almost complete absence not only of the nuclei themselves but also of their precursors. It is important to know that such stable beers exist, even if one is not aware of how this desirable property has been obtained.
The ease with which these nuclei can be temporarily dispersed or inactivated would seem to eliminate the possibility that they are minute crystals, but does perhaps suggest a loose association of colloidal particles.
No reliable method of suppressing gushing has been found. Of the methods tried the most successful has been a new procedure which involved vigorous agitation of the beer followed by filtration.
The apparent lack of any reliable cure for gushing emphasizes the need for more knowledge of the brewing factors involved. It is always more desirable to prevent than to cure, and in this case it might be a great deal more effective. An examination of some brewing factors is reported in Part II of this series.
Acknowledgements.—We are indebted to the Directors of Whitbread & Co., Ltd., for permission to publish this paper. We also wish to thank Commercial Motion Pictures, Ltd., of London for taking the cine’ films, and Aeon Laboratories, Ltd., of Egham, Surrey, for the examination by electron microscope.
References
1. Beattie, G. B., Wallerstein Lab. Commun., 1951, 14, 81.
2. Brenner, M. W., Proc. A.M. Amer. Soc. Brew. Chem.. 1957, 5.
3. British Patent, No. 815,377 (1959).
4. Helm, E., & Richardt, O. C, Woch. Brau., 1938. 55. 80.
5. Krause, B., Svcnska Bryg. Manads., 1936, 51, 221.
6. Schmith, T., Acta chem. Scand.. 1952, 223.
7. Thornc, R. S. W., & Helm, E., this Journal, 1957, 415.
8. Vogel, E. H., Wallerstein Lab. Commun., 1950, 13, 316.
Crescent City Brew Talk 