Evolution of various concentrations in the moisture and in the outer layer and centre of a model during its and storage in an ammoniacal atmosphere Frédéric Gaucheron, Yvon Le Graët, Françoise Michel, Valérie Briard, Michel Piot

To cite this version:

Frédéric Gaucheron, Yvon Le Graët, Françoise Michel, Valérie Briard, Michel Piot. Evolution of various salt concentrations in the moisture and in the outer layer and centre of a model cheese during its brining and storage in an ammoniacal atmosphere. Le Lait, INRA Editions, 1999, 79 (6), pp.553- 566. ￿hal-00929672￿

HAL Id: hal-00929672 https://hal.archives-ouvertes.fr/hal-00929672 Submitted on 1 Jan 1999

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. l Lait (1999) 79, 553-566 553 © Inra/Elsevier, Paris I Original article Evolution of various salt concentrations in the moisture and in the outer layer and centre of a model cheese during its brining and storage in an ammoniacal atmosphere

Frédéric Gaucheron *, Yvon Le Graët, Françoise Michel, Valérie Briard, Michel Piat

Laboratoire de recherches de technologie laitière, Inra, 65 rue de Saint Brieuc, 35042 Rennes cedex, France

(Received 8 July 1998; accepted 20 May 1999)

Abstract - The evolution of various salt concentrations in moisture and in the outer layer and cen- tre of a model cheese during brining and storage in an ammoniacal atmosphere was studied. During brining, calcium, magne sium, potassium, inorganic phosphate and citrate ions entered the. with a slight decrease in their contents in the moisture. In parallel, sodium and chloride were incorporated in outer layer and in the moi sture of gel. Then, during storage of this model gel in an ammoniacal atmo- sphere, calcium, magnesium, inorganic phosphate and citrate ions migrated to the outer layer of gel and consequently underwent a decrease in their concentrations in the moisture and the centre of the gel. These migrations are related to the formation of pH gradient in the gel which indu ce precipita- tions of these ions in the outer layer. In parallel, sodium, potassium and chloride were rapidly present uniformly in the different parts of the gel. The model cheese potentialities are also discussed. © Inra/Elsevier, Paris.

model cheese / mineraIs / aqueous phase / pH / salt / brining / ripening

Résumé - Évolution des concentrations minérales dans la phase aqueuse, la partie centrale et la surface d'un fromage modèle durant son saumurage et son stockage en atmosphère ammo- niacale. Durant le saumurage, les ions calcium, magnésium, potassium, phosphate inorganique et citrate passent dans la saumure avec une légère diminution de leurs concentrations dans la phase aqueuse. En parallèle, les ions sodium et chlorure sont incorporésà la surface et dans la phase aqueuse du gel. Puis, durant le stockage en ambiance ammoniacale, les ions calcium, magnésium, phosphate inor- ganique et citrate migrent vers la surface du gel avec comme conséquence une diminution de leurs concentrations dans la phase aqueuse et dans la partie centrale du gel. Ces migrations sont dues à la

* Correspondence and reprints. [email protected] 554 F. Gaucheron et al. formation d'un gradient de pH dans le gel qui entraîne une précipitation de ces ions en surface. En paral- lèle, les ions sodium, potassium et chlorure étaient rapidement répartis de façon uniforme dans les dif- férentes parties du gel. Les potentialités de ce fromage modèle sont également discutées. © Inral Elsevier, Paris. fromage modèle 1minéraux 1 phase aqueuse 1 pH 1 saumurage 1 affinage

1. INTRODUCTION [22, 27]. The second approach was the direct determination of the total mineral content Cheese is composed of a paracaseinic in the outer layer and in the centre of the network, dispersed elements (cells, fat glob- gel before and after brining and as a function ules, crystals and air) and interstitialliquid of storage time. containing several dissolved compounds (proteins, sugar, lipids and minerals), Dur- ing its fabrication, physicochemical com- 2. MATERIALS AND METHODS position of cheese varies as a function of acidification, brining, alcalinisation, prote- 2.1. Experimental protocol olysis and lipolysis. Proteolysis and amino acid catabolism, lipolysis and lactose break- The experimental protocol is schematically down are relatively well described in the presented in figure 1. This protocol and subse- quent analyses were carried out in triplicate. literature [4-6, 11,32]. On the contrary, much less is known about the salt fraction evolution during brining and ripening 2.2. Milk epuration by cross-Dow although this fraction plays a considerable microfiltration role in the texture elaboration and in the rhe- ological properties of cheese [6, 16,23,24, Raw skim milk from Triballat (Noyal-sur- 28,31,32]. Moreover, sorne ofthese mod- Vilaine, France) was used. Before microfiltra- ifications are sequential and others can take tion, the skim milk was preheated to 50 oc. Then, place at the same time. So, for simultane- cross-flow microfiltration was used to eliminate ous modifications, it is difficult to under- bacteria [30]. The membrane module in alumine Sterilox (Société des céramiques techniques, stand the fundamental phenomena. Tarbes, France) had 19 channels, each with an The objective of this work was to develop inner diameter of 4 mm, and a total membrane a model gel to study the salt behaviour dur- area of 0.2 m-. The average nominal pore size was 1.4llm. The temperature was 50 Under ing brining and storage under ammoniacal oc. these conditions almost aIl of the whey proteins atmosphere; model gel was chosen to limit and casein micelles pass through the membrane the numerous other changes that occur dur- while 99.99 % ofbacteria are retained [30]. The ing brining and ripening of cheese, i.e. lac- permeate was then concentrated by ultrafiltra- tose degradation, proteolysis and lipolysis. tion (UF). Two approaches were used. The first one was the determination of salt contents in the moisture obtained by gel pressing [22, 27], 2.3. Ultrafiltration after brining and as a function of time of The preparation of the UF retentate was car- storage in ammoniacal atmosphere. It is ried out as previously described by Maubois & noteworthy that most of the few studies Mocquot [20]. The membrane Ml (Carbosep", related to the characterization of the moisture Orelis, France) was used (molecular mass eut- extracted from chee se concem hard off: 150000 g·mol-') at 50 "C. Under these con- Evolution of various salt concentrations in a model cheese 555

Day 0 Skim milk 11

Milk purification by cross-tlow microliltration 11

Milk protein concentration by ultrafiltration

(UF retentate) 11

Addition 01 antibiotics, fungicide and lactic acid 11

Addition of glucono-delta·lactone (40g.kg-') 11

Renneting al 30 ·C

(0.1 g.kg-' of rennet) 11

Cooling at 20 ·C

Day 1

Brining at 12 ·C for 30 min 11 Figure 1. Experi- mentai protocol for Spraying 01 pimaricine solution the preparation of Storage at 12 ·C in ammoniacal atmosphere du ring 13 d the model chee se. Figure 1. Protocole Day 1 before and alter brining, 3 , 6, 9, 11 and 13. expérimental pour la 11 préparation de fro- Extraction and physlco-chemlcal analyses 01 the gel moisture mage modèle. Physico-ehemical analyses of surface and centre of gels

ditions, whey proteins can pass slightly through after GDL addition, 0.10 g-kg'" of rennet solution the membrane (rejection of 75 %) while casein (1/10 000) (laboratoire Granday-Roger, Beaune, micelles are retained. To limit bacterial growth France) was added to the UF retentate. The ren- and contamination, nisin, penicillin Gand strep- net coagulation time was about 20 min. The gels tomycin (Sigma, MO, Saint-Louis, USA) were obtained were left standing ovemight at room added to the UF retentate at the following final temperature. The rate of cooling to 20 "C after gel concentrations: 100 UI ml.:'. 12 UI mL -1 and formation was less than one hour. Afterwards, 12 mg-L':' [17]. A final concentration of a slight syneresis, with a loss of about 20 g for a 10-4mol-L'! OfE amino caproic acid was added total weight of 370 g of gel, was observed. The to inhibit residual plasmin activity. The protein characteristics of the model gel are indicated in concentration was about 19 g-kg'". table J.

2.4. Acidification and renneting 2.5. Brining

A lactic acid (Sigma, France) solution A saturated brine (330 g-kg'" of NaCI) was (1 mol-L -1 dissolved in UF permeate) was added prepared from distilled water and dairy salt under stirring to the UF retentate to have a pH (Compagnie saline du midi, Dax, France). The value of about 6.3. A glucono-delta-Iactone conditions of the brining were 30 min at 12 "C (GDL) (Lysactone, Roquette, Lestrem, France) with a ratio of 2 L of brine per gel. Brining was concentration of 40 g-kg! was added to the UF done under manual stirring to maintain the brine retentate previously incubated at 30 "C to have concentration constant throughout the experi- pH value of about 4.75 after 24 h. Immediately ment. 556 F. Gaucheron et al.

Table I. Characteristics of the model chee se The extractions of moisture were carried out on before brining. day 1 (before and after brining), 3, 6, 9, Il and 13 Tableau I.Caractéristiques du fromage modèle (day 0 = renneting day). Before minerai analyses avant saumurage. of moisture, samples were ultrafiltered on Cen- triflo CF 25 membrane (molecular mass eut-off: 25 000 g-rnol ", Amicon, Epernon, France) Diameter 11.0 cm (800 g, 30 min, 12 OC). Height 3.7 cm Surface 318 cm? Volume 351 cm! 2.8. Zone preparation Weight 370 g PH 4.75 The preparations of the outer layer and centre G1ucono- Delta- Lactone 40 g·kg-I of the gel were carried out on day 1 (before and Total Nitrogen X 6.38 208 g-kg" I after brining), 3, 6, 9, Il and 13 (day 0 = ren- Non Casein Nitrogen X 6.38 28.9 g·kg- neting day) [18, 19]. The depth and weight of I Non Protein Nitrogen X 6.38 4.8 g·kg- each zone of the gel were about 2 mm and 65 g, Fat o g-kg! respectively. Lactose 39 g·kg-I Calcium 5.700 g·kg-I Magnesium 0.300 g-kg " 2.9. Analyses Sodium 0.450 g·kg-I Potassium 1.740 gkg! The pH values were measured with a Por- Chloride 0.800 g·kg-I tamess pH-meter. Inorganic phosphate 6.700 g·kg-1 The contents in total protein were determined Citrate 1.950 g·kg-I according to the lOF method (standard 29.23 % Dry matter 20B:1993). For minerai content determination in different zones of the gel, about 1 g of gel was homogenised 2.6. Storage conditions in 30 g of 0.02 rnol-L -1 nitric acid solution. Then, after standing overnight at room temperature, the solution was filtered on Whatman 42. The gels obtained were sprayed with a 5 gL-1 pimaricine solution (Delvocid'", Gist-brocades, On the filtrates, cation (calcium, magnesium, Seclin, France) and stored in an ammoniacal sodium and potassium) and anion (chloride, phos- atmosphere for 13 days. Storage in ammoniacal phate, citrate) concentrations were determined atmosphere was necessary to create a pH gradi- using atomic absorption spectrometry (Varian, ent in the model gel. The NH, diffusion was done Les Ulis, France) [2] and ion chromatography from a vessel containing 2 mol·L-1 (day 1) and (Dionex, Jouy-en-Josas, France) [7], respectively. then 0.3 mol-Lr! (day 2 to day 13) ammoniacal The accuracies of the cation and anion determi- solution in a ripening box (volume 0.75 m'). nations were about ± 2 %. The concentration Temperature and relative humidity were 12 "C units were expressed in mg per kg of material and 95 %, respectively. (moisture or gel).

2.10. Theoretical calculation 2.7. Moisture extraction of calcium phosphate saturation indexes in a pH range between The extraction protocol used was the one described by Salvat-Brunaud et al. [27] for 4.8 and 6.5. cooked hard cheese of the Emmental type and adapted to soft cheese by Pierre et al. [26]. The From initial composition of moisture (day 1 methodology of this extraction has the advan- after brining) and with a NaCI concentration of tages of preserving the minerai equilibrium and 4 moIL-I, calcium phosphate saturation indexes obtaining sufficient quantities of moi sture for in a pH range between 4.8 and 6.5 were calcu- physicochemical analyses without use of chern- lated. The saturation indexes with respect to a ical compounds or dilution. The quantity of gel given calcium phosphate phase were defined as was 1.2 kg and the ratio sand/gel was about 1.211. the ratio: ionie activity product in solution/solu- Evolution of various salt concentrations in a model cheese 557

bility product. Calculations were performed by an centrations of these ions were unaffected, iterative method described by Holt et al. [13]. indicating that the diffusions of ionie sodium Although calcium is bound more strongly to glu- and chloride were not immediate. Secondly, conate than to lactate, we have assimilated both these ions. It is noteworthy that the results of the pH (figure 4A) and the concentrations of these theoretical calculations indicated tendencies calcium (figure 5A), magnesium (figure 6A), and were of semi-quantitative significance. potassium (figure 7A), inorganic phosphate (figure SA) and citrate (figure 9A) in moisture and the outer layer decreased while those of 3. RESUL TS AND DISCUSSION the centre were not modified. The variations in the outer layer of the ratio concentration 3.1. Brining effects before brining/concentration after brining expressed in % were 32, 36,45, 28 and 36 % The comparison of results before and after for calcium, magnesium, potassium, inor- brining on day 1 of model chee se showed ganic phosphate and citrate, respectively. In that this step has multiple physicochemical parallel, low decreases were observed in the effects. Firstly, the sodium and chloride con- moisture: 9.6, 9.5, 13, 1 and 1.5 % for cal- centrations in the moisture and in the outer cium, magnesium, potassium, inorganic layer increased strongly (figures 2A, 3A, phosphate and citrate, respectively. These respectively). In the centre of the gel, con- differences in variations between the outer

35000 A 30000

~ 25000 ~ 20000 W 15000 ~ 10000 5000 o moisture outer layer centre

12000 B ~Ol 10000 . . • • "'" .... • rncisture Figure 2, Sodium concentra- ~ 8000 tion in the moi sture and in two E 6000 zones of the gel before (white :J ] 4000 histogram) and after brining ~ (black histogram) (A) and as a 2000 function of storage time in o +---t---+--+~-I---+---t---i. ammoniacal atmosphere (H): o 2 4 6 8 10 12 14 moi sture; (C): outer layer (+) Days and centre (.) of gel. 35000 c Figure 2. Concentration en 30000 sodium de la phase aqueuse et ~ 25000 de deux parties du gel avant cil (histogramme blanc) et après E 20000 saumurage (histogramme noir) W 15000 'ë (A) et en fonction du temps de ~ 10000 stockage en atmosphère 5000 ammoniacale (H) : phase centre aqueuse; (C) : surface (+) et O+-.a::..+---t---+---+--+--t---i o 2 4 6 8 10 12 14 centre (.) du gel. Days 558 F. Gaucheron et al. layer and the moi sture showed that the salt ripened cheeses). Thirdly, fat was absent in absorption is essentially a surface phe- our gel. nomenon. In parallel to the absorption of salt, these To approximate the diffusion coefficient decreases of ion concentration in the outer of NaCl, these results were compared with layer are due to ion transfers by diffusion those obtained on model chee se containing from the gel surface towards the brine. It is 70 % of moisture [9]. A value close to noteworthy that using a freshly made NaCI 0.4 cm-/d was found. In previously pub- solution, rather than used brine, leads to a lished works, Geurts et al. [9, 10] and Hardy marked loss of soluble components. Indeed, [12] described the salt diffusion respectively, the mineraI transfers during brining were in model chee se and Camembert cheese, as less important when brine was old or when a process derived from Fick's law and deter- was added to the brine [3, mined a diffusion coefficient of about 8,29]. 0.2 cm-/d. Three reasons can explain these The other consequence of brining is a discrepancies between both determinations. loss of aqueous phase. In our case, no data Firstly, the technologies to prepare our curds concerning water loss was determined. were not those used for traditional cheese. However, this well-known phenomenon is Secondly, our model has a higher moisture described by several authors [3, 8-10, 12, content (about 70 % against 55 % in mould- 29, 32]. Geurts et al. [9] and Hardy [12]

50000 A

7 40000 0) -'" ~ 30000 ID :g 20000 z0 8 10000

0 moisture outer layer centre

20000 B • 70) 16000 -'" ci> • • E 12000 • moisture ID "0 '§ 8000 :c 8 4000

0 0 2 4 6 8 10 12 14 Days

50000 C ., 45000 outer layer 0) 40000 .-'" 35000 ci> E 30000 Figure 3. Evolution of chlo- ID 25000 o§"0 20000 ride concentration. Same sym- :c 15000 bols asfigure 2. 8 10000 centre 5000 Figure 3. Évolution de la 0 concentration en chlorure. 0 2 4 6 8 10 12 14 Mêmes symboles que pour la Days figure 2. Evolution of various salt concentrations in a model cheese 559

found a constant ratio of 2.5 between the was observed [results not shown and 18, absorption of salt and the loss of moisture in 19]. Conversely, during storage in an ammo- a soft cheese. In our case this ratio is prob- niacal atmosphere, the pH of the moisture ably different because: a) our curds have (figure 4B), in the outer layer and in the cen- surfaces different from those of Camembert tre increased (figure 4C). The calcium (fig- cheese (table 1); b) the chemical composition ure 5B), magnesium (figure 6B), inorganic of our brine (only NaCI) was different from phosphate (figure SB) and citrate (figure 9B) those used in chee se making (NaCI, miner- concentrations in the moi sture decreased. ais, organic compounds coming from whey) In parallel, the concentrations of these ions [3, 8]. increased in the outer layer and decreased in the gel centre (figure 5C, 6C, SC, 9C, respectively). 3.2. Effects of storage in an ammoniacal atmosphere These variations of concentrations in moisture, outer layer and centre of gel were 3.2.1. Calcium, magnesium, inorganic related to migrations of calcium, magne- phosphate and citrate ions sium and inorganic phosphate ions from the centre to the outer layer during storage in In the absence of ammonia, no signifi- ammoniacal atmosphere of model gel. cant variation of pH and ion concentrations Moreover, in this work, we first observed

4.90 A

4.80

I 0.

4.70

4.60 moisture outer layer centre

6.00 B

5.50

I a. 5.00 • 4.50 0 2 4 6 8 10 12 14 Days 7.00 C • 6.50 • • outer layer 6.00 I a. Figure 4. Evolution of pH. 5.50 Same symbols asfigure 2. 5.00 centre Figure 4. Évolution du pH. 4.50 Mêmes symboles que pour la 0 2 4 6 8 10 12 14 figure 2. Days 560 F. Gaucheron et al.

a citrate migration. However, it is probable the supersaturation of calcium phosphate that, in cheese, this citrate migration was (octacalcium phosphate OCP and tri- not observed because of its degradation by calcium phosphate TCP) increased strongly lactic acid bacteria [4]. These phenomena and explain the precipitation of calcium of migration take place subsequent to the phosphate salts especially in the outer layer. formation of pH gradient in the gel. The Similar behaviours of both calcium phos- increase in pH value in the outer layer of phate salts were determined. the gel due to storage in ammoniacal atmo- Evolution of calcium was similar to that sphere leads to the precipitation of differ- obtained with magnesium because at 13 d of ent possible salts at the gel surface. These storage in an ammoniacal atmosphere, the salts can be calcium phosphate, magnesium cation concentration ratio between the outer phosphate, calcium citrate and magnesium layer and the centre was about 6 for both citrate which have low solubilities [14]. cations. These similar behaviours are prob- From initial composition of moisture pressed ably due to the same low solubilities for out (after brining) and with a NaCI concen- salts of calcium and magnesium. In milk, tration of 4 moIL-I, theoretical calculations calcium-citrate, by surpassing its solubility of calcium phosphate saturation indices in a limit, is supersaturated and no precipitation pH range between 4.8 and 6.5 were carried of calcium-citrate is observed [32]. How- out (figure 10). Thus, during pH increase, ever, in our case, the citrate concentration

8000 A 7000 ~~ 6000 ~ 5000 E 4000 ::J <:; 3000 êii Q 2000 1000 0 moisture outer layer centre B 8000 7000 .x:œ 6000 œ E 5000 E 4000 • ::J 3000 <:; moisture êii 2000 Q 1000 0 0 2 4 6 8 10 12 14 Days

21000 C • 18000 • ~ 15000 • outer layer œ • • E 12000 Figure 5. Evolution of cal- E 9000 cium concentration. Same ::J centre symbols asfigure 2. ."êii 6000 Q 3000 • Figure 5. Évolution de la 0 concentration en calcium. 0 2 4 6 8 10 12 14 Mêmes symboles que pour la Days figure 2. Evolution of various salt concentrations in a model cheese 561 was higher in the aqueous phase (about massic ratio Ca/P calculated at the gel sur- 3 100 mg-kg:") (table /) than in milk (about face after 13 d of storage in an ammoniacal 1500 mg-kg:') [32]. This high concentration atmosphere was about 2. This ratio sug- of citrate was due to the release of citrate gested the presence of tricalcic phosphate initially bound to casein (about 10 % of total (theoretical massic ratio Ca/P = 1.93) in the citrate) during the acidification of the reten- outer layer of the gel. Brooker [1] identi- tate. So, these high contents in citrate and fied using transmission electron microscopy, calcium contributed to a slight precipitation the presence of calcium phosphate in the of calcium citrate in the outer layer. It is rind of mould-ripened cheese (Coulom- noteworthy that in one-month-old Cheddar miers). However, this Brooker's method cheese, Morris et al. [22] observed the pres- cannot determine the exact nature of this ence of crystals. These authors suggested precipitated salt. In this sense, it would be that these crystals were calcium-phosphate interesting to determine the nature of pre- and calcium-citrate. On the other hand, pres- cipitated salts at the chee se surface by other ence of calcium and magnesium carbonate methods such as polarising microscopy. was unlikely bec ause there is no microbial Taking into account the weight of the metabolism which produces CO2• In the pre- outer layer (about 65 g i.e. 18 % of the total sent work, the nature of the precipitated salts weight), the % of total calcium, magnesium, was not precisely determined. However, the inorganic phosphate and citrate present in

400 A _ 350

~cr, 300 E 250 '[ 200 ,il ~ 150 œ :E1 100 ~ 50 o moisture outer layer centre

350 B ~'" 300 -x: ~ 250 E 200 • .~ 150 Q) §, 100 rrcisture

the outer layer after 13 d of storage in an ion accumulation at the cheese surface coin- ammoniacal atmosphere were 58, 56, 73 cides with the surface flora growth, which and 42 %, respectively. Precipitation ofthese generates a basic pH at the surface of the salts in the outer layer of gel led to deple- cheese by its metabolism [l , 15, 18, 19,21, tions in soluble calcium, magnesium, inor- 25]. ganic phosphate and citrate ion concentra- tions in the moi sture and in the outer layer 3.2.2. Sodium, chloride and consequently induced their migrations and potassium ions towards the gel surface. During their migra- tions, calcium, magnesium, inorganic phos- During storage in an ammoniacal atmo- phate and citrate are not dissociated in the sphere, the sodium, chloride and potassium gel. The forms of association are not deter- concentrations in moisture increased slightly mined but it is probable that soluble salts (figures 2B, 3B, 7B, respectively). In paral- of calcium and magnesium phosphate and lei, the large difference observed in sodium calcium and magnesium citrate existed. chloride content between the centre and the Similar results were observed in different outer layer before and after brining on day 1 mould-ripened cheeses such as Camembert, (figures 2A, 3A, respectively) disappeared Coulommiers, Brie and Pont l'Evêque [l, after 5 d of storage in an ammoniacal atmo- 15, 18, 19,21,25]. In industrial cheeses, sphere because in this period, their concen-

3000 A

-t 2500 ~Cl 0> 2000 E E 1500 ;J ·in U) lU 1000 ë ~ 500

0 moisture ouler layer centre

3000 B r: ~Cl 2500 • . • • 0> • • moisture E 2000 E 1500 ;J ·in U) lU 1000 ë ~ 500 0 0 2 4 6 8 10 12 14 Days 3000 C ., 2500 Cl ~ centre 0> 2000 E ! :==-- • Figure 7. Evolution of potas- E 1500 • • ===::e ;J sium concentration. Same ·in U) :;:er~Yer lU 1000 symbols asfigure 2. ë ~ 500 Figure 7. Évolution de la 0 concentration en potassium. 0 2 4 6 8 10 12 14 Mêmes symboles que pour la Days figure 2. Evolution of various salt concentrations in a model cheese 563 trations decreased in the outer layer and potassium in the cheese probably due to the increased in the centre (figures 2C, 3C, development of surface flora was shown respectively). During this period, the [19]. In our case, as the curd was poor in sodium, potassium and chloride ions dif- microorganisms due to the presence of fused in the gel matrix and at day 5 they antibiotic and fungicide, which limited their were uniformly distributed. This time was in growth [17], no migration of potassium was good accordance with a diffusion coeffi- observed. cient of about 0.4 cm2/d previously deter- mined in this work for NaCl. Then, beyond 5 d and up to the end of storage, the con- 5. CONCLUSION centration of sodium and chloride in the cen- tre increased slightly or was constant for In this study, we have prepared a model potassium. cheese through the combined use of mem- It is noteworthy that the storage in a non brane technologies and different products ammoniacal atmosphere gave similar results (lactic acid, GDL, antibiotics, fungicide). [18, 19] which indicated that the presence of This model has been successfully used to ammonia has no influence on the diffusion quantify ion transfers during its brining and of sodium, chloride and potassium. In storage in ammoniacal atmosphere. The Camembert cheese, a reversible migration of multiple relationships between these trans-

10000 r: Cl -'" 8000 ci> E 6000 Qj' iii s: a. 4000 0 s:'" ~ 2000 0 moisture outer layer centre

10000 B r: Cl -'" 8000 ci> E Qj' 6000 iii s: 4000 a. moisture 0 r.'" 2000 ~ 0 0 2 4 6 8 10 12 14 Days

30000 C -; Cl 25000 -'" • ci> • outer layer Figure 8. Evolution of inor- E 20000 Qj' ganicphosphateconcentration. iii 15000 s: Same symbols asfigure 2. a. 10000 s:'"0 centre Figure 8. Évolution de la s, 5000 concentration en phosphate 0 inorganique.Mêmessymboles 0 2 4 6 8 10 12 14 que pour lafigure 2. Days 564 F. Gaucheron et al.

3500 A 3000 ~ Cl .>< 2500 dl E 2000 ~ 1500 ~ 1000 500 0 moisture ouler layer centre

3500 B _ 3000 . ]' 2500 • • moislure ~ 2000 --- ~ 1500 r: 6 1000 - 500 O+---+--+---+---+---I----I~--I o 2 4 6 B 10 12 14 Days 5000 C • 4000 Figure 9. Evolution of ~ • outer layer citrate concentration. dl3000 E Same symbols as figure ~ 2000 2. ~ Q. • • Figure 9. Évolution de la 1000 iC centre concentration en citrate. 0+---+--+--+---+---+--+----; Mêmes symboles que a 2 4 6 8 la 12 14 pour lafigure 2. Days

1.0E+06 1.0E+05 x 1.0E+04 .gJ 1.0E+03 .S; c 1.0E+02 .2ca 1.0E+01 ::; 1.0E+OO Jj 1.0E-01 1.0E-02 1.0E-03 +----+----+----1------1 4.5 5 5.5 6 6.5 pH

10. Theoretical saturation indices of octacalcium phosphate (OCP) (Â) and tricalcium phosphate (TCP) (.) from pH 4.8 to 6.5. Calculations, from the initial minerai composition of the moisture with a NaCI concentration of 4 moIL-I, were carried out as described by Holt et al. [13]. Figure 10. Index de saturation théoriques des phosphates octacalcique (OCP) (Â) et tricalcique (TCP) (.) entre pH 4,8 et pH 6,5. Les calculs, faits à partir de la composition minérale initiale de la phase aqueuse et une concentration en NaCI de 4 rnol-L -l, ont été réalisés comme décrit par Holt et al. [13]. Evolution of various salt concentrations in a model cheese 565 fers and the physicochemical changes occur- [4] Choisy C; Desmazeaud M.J., Gripon J.c., Lam- ring in the model during brining and stor- beret G., Lenoir J., Le fromage: de la science à l'assurance qualité, 3e éd., Lavoisier Tee & Doc, age were not determined precisely because Paris, 1997. several factors such as pH, structure of [5] Fox F.P., Proteolysis during cheese manufac- matrix, state of water, ionie strength, salt ture and ripening, J. Dairy Sci. 72 (1989) solubilities are involved at the same time. 1379-1400. However, the se results show that global [6] Fox F.P., Cheese: chemistry, physics and micro- biology, vol. 1, 2nd Ed,1993. physicochemical changes of the moisture [7] Gaucheron F., Le Graët Y., Piot M., Boyaval E., can reflect sorne local changes in the cheese. Determination of anions of milk by ion chro- Results are qualitatively in accordance with matography, Lait 76 (1996) 433--443. those previously described [3, 9,12,15,18, [8] Geurts T.J., Walstra P., Mulder H., Brine com- 19,21,25,29] although this model has sorne position and the prevention of the defect "soft rind" in cheese, Neth. Milk Dairy J. 26 (1972) differences in composition with industrial 168-172. soft cheeses. Indeed, it contained a high con- [9] Geurts T.J., Walstra P., Mulder H., Transport centration in gluconate and the fat content of salt and water during of cheese. was close to zero. Moreover the initial lac- 1. Analysis of the processes involved, Neth. Milk Dairy J. 28 (1974) 102-129. tose was not metabolised because it does [10] Geurts T.J., Walstra P., Mulder H., Transport not contain micro-organism. The potential of of salt and water during salting of cheese this model appears very interesting because 2. Quantities of salt taken up and of moi sture it is now possible to study the influence of lost, Neth. Milk Dairy J. 34 (1980) 229-254. different factors on salt migrations. Thus, [II] Grappin R., Rank T.C., OIson F., Primary pro- teolysis of chee se proteins during ripening. the salt migrations can be studied in gel con- A review, J. Dairy Sei. 68 (1985) 531-540. taining more or less proteins, fat, with dif- [12] Hardy J., Étude de la diffusion de sel dans les ferent casein/whey proteins ratios, with fromages à pâte molle type Camembert. Com- phosphopeptides or with different initial pH paraison du salage à sec et du salage en sau- values. The brining and the ripening condi- mure, thèse Di-ingénieur. université Nancy-I, 1976. tions can also be studied. The consequences [13] Holt c., Dalgleish D.G., Jenness R., Calcula- of these physicochemical heterogeneities tion of the ion equilibria in milk diffusate and (pH, minerai composition) in the cheese on comparison with experiment, Anal. Biochem. the texture formation can be also investi- 113 (1981) 154-163. gated. [14] Holt C; Effect of heating and cooling on the milk salts and their interaction with casein, Spe- cial issue: Heat-induced changes in milk, Bull. Int. Dairy Fed. 9501, 2nd ed. (1995) 105-133. ACKNOWLEDGMENT [15] Karahadian C, Lindsay R.C., Integrated roIes of lactate, ammonia, and calcium in texture The authors thank Ll., Maubois for critical development in mold surface-ripened cheese, reading of the manuscript. J. Dairy Sei, 70 (1987) 909-918. [16] Lawrence R.C., Creamer L.K., Gilles J., Tex- ture development during , J. Dairy Sei, 70 (1987) 1748-1760. REFERENCES [17] Le Bars D., Desmazeaud M.J., Gripon J.C., Ber- gère J.L., Étude du rôle des microorganismes [1] Brooker B.E., The crystallisation of calcium et de leurs enzymes dans la maturation des fro- phosphate at the surface of mould-ripened mages. 1. Fabrication aseptique d'un caillé cheeses, Food Microstruc. 6 (1987) 25-33. modèle, Lait 55 (1975) 377-389. [2] Brulé G., Maubois J.L., Fauquant J., Étude de la [18] Le Graët Y., Lepicnne A., Brulé G., Ducruet P., teneur en éléments minéraux de produits obtenus Migration du calcium et des phosphates inor- lors de l'ultrafiltration de lait sur membrane, ganiques dans les fromages à pâte molle de type Lait 54 (1974) 600-614. Camembert au cours de l'affinage, Lait 63 [3] Centeleghe J.L., Millière J.B., Veillet L.. (1983) 317-332. Weber F., Aspects physico-chimiques et micro- [19] Le Graët Y., Brulé G., Migration des macro et biologiques du salage en saumure des pâtes oligo-éléments dans un fromage à pâte molle molles moisies, Tech. Lait. 708 (1971) 13-19. de type Camembert, Lait 68 (1988) 219-234. 566 F. Gaucheron et al.

[20) Maubois J.L., Mocquot G., Préparation de fro- cheese structure, composition and non solvent mage à partir de « pré fromage liquide» obtenu water, Lait 79 (1999) 489-501. par ultrafiltration du lait, Lait 51 (1971) 495-533. [27] Salvat-Brunaud D., Maubois J.L., Le Graët Y., Piot M., Maillard M.B., Corre C., Thierry A., [21] Metche M., Fanni J., Rôle de la flore fongique Extraction et analyse de la phase aqueuse de dans l'accumulation du calcium et du phosphore l'emmental à 4 stades d'affinage, Lait 75 (1995) à la surface des fromages type Camembert, 239-249. Lait 58 (1978) 337-354. [22] Morris HA, Holt c., Brooker RE., Banks J.M., [28] Solorza P.J., Bell A.E., The effect of calcium Manson W., Inorganic constituents of cheese: addition on the rheological properties of a soft analysis of juice from a one-month-old Ched- cheese at various stages of manufacture, Int. dar chee se and the use of light and electron J. Dairy Technol. 51 (1998) 23-29. microscopy to characterize the crystalline phases, [29] Terré E., Le Graët Y., Brulé G., Maubois J.L., J. Dairy Res. 55 (1988) 255-268. Étude du transfert des solutés des fromages à [23] Mpagana M., Hardy J., Effeet of salting on sorne pâte molle dans les saumures. Intérêt du traite- rheological properties of fresh Camembert ment par ultrafiltration sur membrane, Tech. cheese as measured by uniaxial compression, Lait. 997 (1985) 39-47. Milchwissenschaft 41 (1986) 210-213. [30] Trouvé E., Maubois J.L., Piot M., Madec M.N., [24] Noomen A., The role of the surface flora in the Pauquant 1., Rouault A., Tabard J., Brinckman G., softening of cheeses with a low initial pH, Neth. Rétention de différentes espèces microbiennes Milk Dairy J. 37 (1983) 229-232. lors de l'épuration du lait par microfiltration en [25] Oumer A., Dynamique d'évolution des flores flux tangentiel, Lait 71 (1991) 1-\3. microbiennes d'un fromage à pâte molle et à croûte lavée en relation avec quelques aspects [31] Vassal L., Monnet V., Le Bars D., Roux C, Gri- physico-chimiques et biochimiques, thèse doc- pon J.c., Relation entre le pH, la composition torat Institut national agronomique Paris-Gri- chimique et la texture des fromages de type gnon, 1997. Camembert, Lait 66 (1986) 341-351. [26] Pierre A., Michel P., Le Graët Y., Berrier J., [32] Walstra P., Jenness R., Dairy Chemistry and Soft goat cheeses at different ripening stages: . Physics, John Wiley and sons, New York (1984)