Eur J Appl Microbiol Biotechnol (1982) 16:105-109 Eu~ope~ A~nl;~H Jourmlot rtmJiJll~

Microbiology and Biotechnology �9 Springe~Vedag.1982

Foaming in Winemaking

I. A Technique for the Measurement of Foaming in Winemaking

Martin Edwards 1 , Rainer Eschenbruch 1 , and Peter C. Molan 2

1 Ruakura Soil & Plant Research Station, Hamilton, New Zealand 2 School of Science, University of Waikato, Hamilton, New Zealand

Summary. A new technique for the measurement of foam- ing in winemaking is described. Linear relationships are es- tablished between the foaminess of bovine serum albumin solutions and its concentrat ion, between the foaminess of fermented artificial medium and its concentrat ion and be- tween the foaminess and the rate of sparging o f gas. The effect of temperature on the foaminess of a bovine serum albumin solution is also established.

and Clark 1939; Rudin 1957; Bishop 1975; Narziss 1978); none has been described for use in winemaking. We have evaluated some of these methods and found them unsatis- factory for small volumes. Therefore, a new technique had to be devised, which basically is a fundamental modifica- t ion of that developed by Bikerman (1938). Details are described in this paper.

Introduction

Foaming is a problem often experienced in the fermenta- t ion of grape juice. The degree of foaming varies with grape variety, vintage, type of juice extract ion, method of clarification, and the strain o f yeast used. Foaming is par- t icularly pronounced during the first days o f the fermen- tat ion. Usually up to 5% of the tank volume is left empty to accommodate the froth which in some years can attain a height so great that it takes up half of the tank volume. This reduction in fermentat ion capacity increases finan- cial commitments in equipment which is used for only a short period each year. Although some work has been done studying the influence o f the yeast strain on foaming (Ouchi and Nunokawa 1973; Eschenbmch and Rassel 1975; Thornton 1978) no information is available analys- ing the effects o f the composit ion of grape juice on foam formation. At tempts to solve this problem require a re- liable and objective measurement of the foaminess of not only grape juices and wines but also artificial media. Iso- lation of the components responsible for the foaminess demands an assay that can be carried out quickly on frac- tions o f small volume. Almost all the methods have been developed for the measurement o f foaming in beer (Ross

Offprint requests to: R. Eschenbruch, DSIR Wine Research, Box 19, Te Kauwhata, New Zealand

Materials and Methods

Fermentation Conditions. A pure cuRure yeast, Saceharomyces cervisiae, strain 102 of the Ruakura collection, was grown in ster- ile 4 1 lots of complete mineral medium (CMM) with 5 g NH4C1 per litre. The composition of this medium, initially used by Toku- yama et al. (1973), is shown in Table 1. After 12 days of fermen- tation at room temperature the yeast was removed by centrifuga- tion.

Preparations for Assay. The fermented medium was first concen- trated by ultra-filtration from 4 1 to 1 1 with a Millipore High Vol- ume Molecular Filtration Cell fitted with 4,650 cm2 of PTGC membrane (nominal retantion 10,000 daltons), then to 400 ml with an Amicon PM10 membrane (nominal retention 10,000 dal- tons) in a stirred cell. This stock solution of concentrated medium was diluted to the required degree of foaminess by the addition of distilled water.

Bovine serum albumin (Fraction V, Sigma Chemical Company) was dissolved in sodium citrate buffer, 0.1 mol/1, pH 5.0, at initial concentration of 0.1 g/]. Other concentrations were obtained by dilution of this solution with the citrate buffer.

Apparatus for Assay. The apparatus constructed for the measure- ment of foaminess is illustrated in Fig. 1. Gas from a cylinder with a pressure regulator from 0 to 700 kPa passes through high-pres- sure tubing to a flow-regulating restrictor (A), then to a No. 4 (10 /~m pore size) glass sinter (B) fitted in the bottom of a graduated glass tube of 12 mm internal diameter and 55 cm length (C). Oxy- gen-free nitrogen was chosen in preference to carbon dioxide to avoid changing the pH of the solutions being assayed.

The restrictor was made by drawing out a piece of thick-walled capillary tubing of 0.5 mm internal diameter and cutting it back

0171-1741/82/0016/0105]$01.00

106 M. Edwards et al.: Foaming in Winemaking. I.

== t o

Sinter

12 mm

/-Glass column

(B)J Enlarged scale

/-Plastic tubing

(A)

High pressure. tubing

Fig. 1. Diagram of the apparatus for measuring foaminess

Table 1. Composition (per litre) of complete mineral medium (CMM) for growth of yeast cultures

Glucose 100 g NH4C1 5 g CaC12 �9 2H20 0.7 g KCI 0.6 g Na2SO 4 0.3 g MgC12 �9 6H20 0.4 g

Na2HPO 4 �9 2H20 2.5 g Citric Acid 1.8 g Mineral stock solution* 1 ml Vitamin solution** 1 ml

* Composition (per 500 ml): HBO 3 250 mg Thiamine MnC12 �9 4H20 234 mg Inositol FeC13 �9 6H20 100 mg Biotin Na2MoO 4 �9 2H20 100 mg Pantothenate ZnC1 95 mg Folio acid KI 50 mg Pyridoxine CuC12 �9 2H20 13.5 mg Vitamin B12

** Sterilized by filtration and added after autoclav- ing the mineral medium; composition (per 500 ml):

200 mg l g 1 mg

200 mg 200 mg 200 mg

50 mg

until it allowed approximately 30 ml of gas per min to flow at a pressure of 500 kPa. It served to eliminate variations in gas flow rate (hence foam production) caused by changes in back-pressure resulting from differences in the height of the foam produced or in wetting of the sinten A pressure drop of at least 300 kPa through the restrictor is necessary for this.

The porosity for the sinter was found to be important: large pores gave rise to irregular bubbling and allowed liquid to flow back at times, this leading to parts of the disc no longer passing gas. Short tubes fitted with sintered discs are readily obtainable, and after cutting back to just above the disc can be joined to the graduated tube. If the distance between the disc and the graduations (D) is increased a larger sample for the assay is required. The graduations were obtained by affixing to the tube a millimeter scale photo- copied on cellulose acetate film. The scale illustrated was found to be particularly easy to read when concentrating on observing the movement of the foam-liquid interface.

The gas flow set at the start of a series of measurements and was left unaltered throughout. It was determined by means of a bubble flow-meter connected to the open end of the graduated tube. Samples (10 ml) to be assayed for foaminess were run down the side of the graduated tube held inclined: this prevented the liq- uid from getting caught up in rising foam.

The foaminess was measured as the volume of liquid held in the foam. This was determined by observing the level of the liquid remaining and subtracting this from the level of the 10 ml original- ly put in the tube. It was found that the position of the foam/liq- uid interface could be much more easily seen if the interface was illuminated from behind.

Results

The foaminess of solutions of bovine serum albumin of various concentrations was measured with the apparatus described in this paper. The results are shown in Fig. 2. A linear relationship between concentration and the foami- ness was found.

Measurements were also made with various dilutions of the fermented mineral medium. The results are shown in Fig. 3. Again, a linear relationship was found. Repeated measurements were made with one dilution that gave a mean height of the foam/liquid interface of 63.9 mm. In all, 20 readings were obtained. The standard deviation was found to be 2.34, the confidence interval at the 95% level being 1.78-3.42.

Measurements were then made of the foaminess of some finished wines which were noticed to froth a little when the bottles were shaken. The values obtained from these measurements ranged from 25 to 86 mm. These all fell within the middle of the useful range of the apparatus. (The wine samples were degassed for 40 min at room tem- perature in a rotary evaporator and the flow rate was low- ered to 25 ml/min.)

In the case of solutions which were of too low foami- ness, measurements could be made by increasing the gas flow to increase the foaming. The limit to this was found to be the difficulty of discerning the foam/liquid inter- face with a very high gas flow.

-~- 10 E

2o

3o! ~E

.~ 40

~ 60 ~g

90

Gas flow (ml-miff 1)

--~1o

~2o

~3Q " O

"~40

~so

"6 ~e

~:7o

8r

9C

/ / I I ~ I m I I o.o, o.o2 o.~3 0.0, o.os o.oe o.~7 o~8 o.o, o~

Concentration BSA (g mL -~)

Fig. 2. Relationship between the foaminess of solutions of bovine serum albumin and its concentration. Temp. 20 ~ Flow Rate 33 ml. min -1

E 50 E

E

2o

"~ 30

~-4e

}. o 60

~: 70

80 ~

90

Fig. 4. Relationship between the foaminess and the rate of sparging of gas through a solution of bovine serum albumin (BSA) solution. Temp. 20 ~ Concentration of BSA 0.06 g/ml

80

;o 2'o 3'o ,'o s'o ;o 7'0 8'0 % CMM stock solution

Fig. 3. Relationship between the foaminess of the fermented complete mineral medium and its concentration. Temp. 20 ~ C, Flow Rate 33 ml. min -1

i

-~ eo

M. Edwards et al.: Foaming in Winemaking. I. 107

Deg C

"1-

8G 10

Fig. 5. Effect of temperature on the foaminess of a bovine serum albumin solution. Flow Rate 33 ml �9 min -1 , Concentra- tion of BSA 0.04 g/ml

The relationship between the rate of gas flow and the foaminess was determined by measuring the foaminess of 10 ml lots of an appropriate dilution of the bovine serum albumin solution with different flow rates. The results are shown in Fig. 4. The relationship was linear, showing that if measurements are made at different gas flow rates, they can be related by simple conversion.

The effect of temperature on the measurements was also assessed with an appropriately diluted solution of the bovine serum albumin solution. Samples were wanned to

various temperatures, and the sparging gas was passed

through a coil immersed in a water bath of corresponding temperature. As can be seen from Fig. 5 the temperature clearly affects the measurement. This needs to be control- led if room temperatures are not stable.

D i s c u s s i o n

The evaluation of the newly developed technique de- scribed in this paper shows it to be a reliable means of

108 M. Edwards et al.: Foaming in Winemaking. I.

measuring foaminess in winemaking, for which existing techniques are less suitable. These other techniques re- quire the foam to be more stable and higher than normal- ly encountered. They are also too time-consuming and re- quire samples that are too large for conveniently testing large numbers of samples from fractionation studies for which the new technique has been especially developed.

The methods discussed by Ross and Clark (1939) are based on the collapse rate of foams and require the foam to be strong before it can be measured. Although this is an important characteristic of the foam on beer, it is not necessarily crucial in the fermentation of grape juice where a high rate of CO2 evolution can cause frothing without great foam strength. The methods are also slow.

Our technique is shown to successfully measure the quite tow foaminess of finished wine and fermented mineral me- dium. It is also a relatively rapid technique.

The method described by Rudin (1957) also is time- consuming because each assay takes up to 5 rain. The foam from each sample must be produced in the same length of time, so when the degree of foaminess is very variable much trial and error can be involved in obtaining the appropriate dilution of the sample. Furthermore, in this technique also the foaminess has to be quite high be- fore it can be measured, and a sample volume of 500 ml is required compared with the 10 ml used in our tech- nique.

Sample volumes of 100-200 ml are required in the as- say reported by Cumper (1953). A further drawback of this method is the non-linear relationship between concen- tration of foaming agent and foaminess. Similar results were reported by Kalischewski et al. (1979) using the me- thod of Bikerman (1938). Initially our results produced a similar non-linearity. However, after inserting the gas-flow restrictor (A, Fig. 1) the relationship became linear. This suggests that non-linearity is the result of variations in the back pressure caused by varying heights of the foam col- umns obtained with stronger solutions. We also found the experimental error greater than 10% when the gas flow was not regulated sufficiently: this appeared to be due to the varying pattern of gas flow through the sinter.

In Cumper's assay the diameter of the tube has to be very large to prevent the formation of large foam bubbles. In our case this was a particularly marked problem. Also, measuring accurately the height of the foam consisting of irregular bubbles is equally difficult. These problems were overcome by simply measuring the level of the fiat liquid/ foam interface instead. This was confirmed by our mea- suring aliquots of a solution of bovine serum albumin by our method in tubes of internal diameter 12 mm and 42 mm for comparison, and obtaining mean values, from 10 measurements of the drop in level in each, of 46.7 mm (standard deviation = 0.84) and 45.5 mm (standard devia- tion = 0.97) respectively. (The same rate of glas flow per cm 2 across the sinter was maintained in both tubes.) How-

ever, this does mean that our technique is measuring the volume of liquid contained in the foam rather than the vo- lume of the foam itself. Each is a measure of a different aspect of foaminess.

A further complication to be considered in our tech- nique is that the foaming agent is exhausted during the as- say. Kalischewski et al. (1979) found a sample size of at least 200 ml necessary to overcome this problem when measuring the foaming of protein solutions according to the method of Bikerman (1938). In our technique, with a small sample, the foaming reaches a maximum and then decreases. This could be caused by denaturation of the foaming agent by the process itself. (If the gas flow is maintained, foaming eventually ceases, even though no material is lost from the tube.) The volume of a foam at the equilibrium rate of formation and rate of collapse at a constant gas flow is a measure of the stability of the foam. However, if the foaming agent is being denatured, the equilibrium point is constantly changing. Thus the true equilibrium for the initial concentration will not be reached. Nevertheless, as our method produces results that are linearly related to the concentration, these values in- dicate a meaningful measure of foaminess, as it is reasonable to expect that the rate of denaturation would be propor- tional to the concentration of the foaming agent. Further- more, denaturation would be occurring in the fermentation vat as well as in the assay tube.

The method described in this paper is sensitive, precise, rapid and requires only a small sample (10 ml). Although larger samples may be available from fermentations, when fractions from column chromatography are to be assayed the sample size is small. Also a large number of samples can be analysed in a short time. With this method for as- saying foaminess the influence of the grape juice compo- nents and yeast proteins on foaming during winemaking can now be analysed critically. It witl also be of special in- terest to re-evaluate so-called foaming and non-foaming yeast strains.

Acknowledgement. We would like to thank Mr. P. F. Noble, Drs N. R, Cox and J. T. Mclntosh for their constructive suggestions in the preparation of this paper.

References

Bikerman JJ (1938) The unit of foaminess. Transactions Faraday Soc 34:634-638

Bishop LR (1975) Haze- and foam-forming substances in beer. J Inst Brewing 81:444-449

Cumper CWN (1953) The stabilization of foams by proteins. Transactions Faraday Soc 49:1360-1369

Eschenbruch R, Rassel JM (1975) The development of non-foam- ing yeast strains for winemaking. Vitis 14:43-47

Kalischewski K, Bumbullis W, Schiigerl K (1979) Foam behaviottr of biological media. I. Protein foams. Eur J Appl Microbiol Biotechnol 7: 21- 31

M. Edwards et al.: Foaming in Winemaking. I. 109

Narziss L (1978) Der Bierschaum. Kenntnisse und Erkenntnisse iiber ein komplexes Problem. Brauwelt 118:1045-1057

Ouchi K, Nunokawa Y (1973) Non-foaming mutants of sak6 yeasts: Their physieochemical characteristics. J Ferment Tech- no151:85-95

Ross SR, Clark GL (1939) On the measurement of foam stability with special reference to beer. Wallerstein Lab Commun 6 :46- 54

Rudin AD (1957) Measurement of the foam stability of beers. J Inst Brewing 63:506-509

Thornton RJ (1978) Investigations on the genetics of foaming in wine yeasts. Eur J Appl Microbiol Biotechnol 5:103-107

Tokuyama T, Kuraischi H, Aida K, Ueumura T (1973) Hydrogen sulphide evolution due to pantothenic acid deficiency in the yeast requiring this vitamin, with special reference to the effect of adenosine triphosphate on yeast cysteine desulphydrase. J Gen Appl Microbiol 19: 439-466

Received September 22, 1982