developing smart control strategies for freeze-drying of ... · developing smart control strategies...

Developing smart control strategies for

freeze-drying of lactic acid bacteria

EFFoST Annual Meeting, 20-23 November 2012, Montpellier, France

1

(1) Joint Research Unit of Microbiology and Food Process Engineering

AgroParisTech, INRA, Thiverval-Grignon, France

Stéphanie Passot (1), Ioan Cristian Tréléa (1) Miquel Galan (2)

(2) R&D Process and Application Development Manager

Telstar Technologies, S.L., Terrassa, Spain

BackgroundBackgroundProduction of freeze-dried concentrates of lactic acid

= a complex process

Fermentation

Cooling

Concentration

PAT“Operations that monitor, analyze

and control critical quality attributes of processes and

2

Functionality: acidification activity

(CinAc®)

Formulation

Freeze-dryingFreezing

Sublimation

Desoprtion

Storage

attributes of processes and products, while manufacturing is in progress, i.e. continuous quality

verification”

BackgroundBackgroundProduction of freeze-dried concentrates of lactic acid

= a complex process

Fermentation

Cooling

Concentration

A pluridisciplinary team

Process understanding

Experimental and modeling study

Development of new sensing

3

Formulation


Sublimation

Desoprtion

Storage

Development of new sensing

technologies

Implementation of monitoring &

control tools on the pilot plants

Demonstration on the final end-

user (cheese)

Bacteria Freeze-Drying: The approachBacteria Freeze-Drying: The approach

LAB concentrate

Protective medium:

(200 g/L sucrose)

Freezing (tray)

Primary dryingSecondary drying

Product

3 Process variables3 Process variables

� Shelf temperature

� Chamber pressure

Critical Quality attributesCritical Quality attributes

�Acidification activity

� Structure of the cake

Raw material attributesRaw material attributes

� Physical properties

Storage

4

� Time� Structure of the cake

� Storage stabilityPAT PAT


LAB concentrate

Protective medium:

(200 g/L sucrose)

Freezing (tray)


Product









Storage

5


� Storage stability

Critical Process ParametersCritical Process Parameters

� Product temperature

� Water activity, moisture content

� End point of drying steps

Quantitative Relationship

Sensing method

Measure or Predict in real

time the CPP

Modeling & advanced

optimization tools

Fermentation

Cooling

Concentration

Formulation

Relationships between CPP and bacteria qualityRelationships between CPP and bacteria quality

� Study the whole process

� Quantify the biological degradation after each

step of the freeze-drying process

�Impact of the sublimation conditions

6

Formulation


Sublimation

Desoprtion

Storage

�Impact of the sublimation conditions

� Impact of the residual moisture content

ON

� The bacteria acidification activity


Functionality: acidification activity

(CinAc®)

1- Viability � Plate counts

2- Technological property: Acidification activity � CINAC

6

7

-6

0

dpH/dt, pHu.min-1

Samples analysisSamples analysis

7

The higher the tm value, the lower the acidification activity

Corrieu et al. 1988 (Patent FR 2629612)

4

5

0 200 400 600

Time (minutes)

pH

-24

-18

-12

103 dpH/dt, pHu.min

tm

3- Residual moisture content: Karl Fisher titration, water activity, Tg

Fermentation

Cooling

Concentration

Shelf T°

Chamber pressure-60

-40

-20

0

20

40

0 20 40

Time (h)

Temperature (°C

)0

20

40

60

80

100

Pressure (Pa)

-60

-40

-20

0

20

40

0 20 40

Time (h)

Temperature (°C

)

0

20

40

60

80

100

Pressure (Pa)

Conservative Aggressive

Pilot plant

(Usifroid, SMH 15)

Freeze-drying protocol

Shelf T°

Chamber P

Relationships between CPP and bacteria activityRelationships between CPP and bacteria activity

8

Formulation


Sublimation

Desoprtion

Storage

Freezing

DesorptionSublimation

(Usifroid, SMH 15)

� Pilot plant equipped with a sample thief and a hygrometer

Effect of water content on the acidification activityEffect of water content on the acidification activity

Analysis of samples removed at various times of the desorption step

320

340

Conservative

Aggressivet = 25 hr

Low activity

9

240

260

280

300

0 2 4 6 8

Water content (%)

tm (min)

t = 0

t = 6 hr

Desorption Sublimation

High activity

Effect of water content on storage stabilityEffect of water content on storage stability

8

12

16

20

k (min/day)

Old data

Conservative

AggressiveTg = 25 °C

Optimal range of water content for bacteria stability

10

0

4

8

0 2 4 6 8

Water content (%)

k (min/day)

Slope k = loss rate of acidification activity

Tcoll: the minimum temperature at which the ice sublimation is

accompanied with the collapse of the dried structure

Frozen region

Su

blim

ati

on

fro

nt

Collapse-7°C

Freeze-drying microscopy:

(FDCS 196, Linkam Sci. Inst.)

Direct observation of the structure during

primary drying (sublimation)

Properties of the raw materialProperties of the raw material

11

Tcoll = -7°°°°C ≈ Tg’� Maltodextrin DE 5-8: 10% solution

Dried regionSu

blim

ati

on

fro

nt

-10°C

-7°C

-9°C

BUT for bacteria suspension Tcoll = -27°°°°C >> Tg’ = -36°°°°C

Formulations with low value of Tg’ and Tcoll >> Tg’

� Performing primary drying at product temperature higher than Tg’

0

Temperature (°C

)

Shelf T°

Tg

0

Temperature (°C

)

Shelf T°

Product T°°°° close to Tg’ Product T°°°° close to Tcoll

0

Temperature (°C

)

Shelf T°

Tg

0

Temperature (°C

)

Shelf T°

Tg

0

Temperature (°C

)

Shelf T°

0

Temperature (°C

)

Shelf T°

Product T°°°° close to Tg’ Product T°°°° close to Tcoll


12

Differences in biological activity recovery after freeze-drying ?

Probably Yes

-60

-40

-20

0 10 20 30Time (h)

Temperature (°C

)

Tcoll

-60

-40

-20

0 10 20 30Time (h)

Temperature (°C

)

Shelf T°Tcoll

Tg

OR ??

-60

-40

-20

0 10 20 30Time (h)

Temperature (°C

)

Tcoll

-60

-40

-20

0 10 20 30Time (h)

Temperature (°C

)

Tcoll

-60

-40

-20

0 10 20 30Time (h)

Temperature (°C

)

Shelf T°Tcoll

Tg

-60

-40

-20

0 10 20 30Time (h)

Temperature (°C

)

Shelf T°Tcoll

Tg

OR ??

Some experimental results

Primary drying (DI) -20°C/20 Pa 0°C/20 Pa

tm1 before FD 261 min 240 min

tm2 after FD 475 min 366 min

dtm=tm2-tm1 214 min 126 min

Loss of biological activity during FD

Close to Tg’ Close to Tcoll-10

0

10

Temperature (°C

)

0°°°°C / 20 Pa

-20°°°°C / 20 Pa-10

0

10

Temperature (°C

)

0°°°°C / 20 Pa

-20°°°°C / 20 PaClose to Tg’

Close to Tcoll


13

Loss of biological activity during storage

FD: freeze-drying

-50

-40

-30

-20

0 5 10 15 20 25

Time (hr)

Temperature (°C

)

Tg’

Tcoll

-50

-40

-30

-20

0 5 10 15 20 25

Time (hr)

Temperature (°C

)

Tg’

Tcoll

Close to Tg’

200

400

600

800

0 20 40 60 80 100

Storage time (days)

tm (min)

-20°C/20 Pa 0°C/20 Pa

k=2.8

k=2.1

Hypothesis: Degradation rate K = f (Product T°-Tg)

Freeze-dried samples of LAB

Equilibration at various

Kdt

dtm=

0

400

800

1200

1600

2000

tm (min)

aw = 0.73

aw = 0.33

aw = 0.09

K

Can we integrate in the model the degradation of “biological activity”?


14

Equilibration at various

water activity values

Storage at 25°°°°C under vacuum

Measurement of biological

activity at 7, 10 and 29 days

k0=3.89

k1 = 0.032

( )TgTk storagee−

= 1

0kK

0

20

40

60

-50 0 50 100

Tstorage - Tg (°C)

K (min/day)

0

0 10 20 30

Storage time (days)

aw = 0.09

Integration of the equations describing the

degradation rate in the lyo-model

Simulation of the evolution of tm during

the process

-40

-20

0

20

0 10 20 30

Time (h)

Temperature (°C

) Close to Tg’

DI = -20°°°°C, 20 Pa

80

Loss of acidfication activity (min)

19 min


15DI: Primary drying

Time (h)

-40

-20

0

20

0 10 20 30

Time (h)

Temperature (°C

) Close to Tcoll

DI = 0°°°°C, 20 Pa

0

20

40

60

0 1 2 3 4 5 6

Thickness of dry layer (mm)

Loss of acidfication activity (min)

Close to Tg’

Close to Tcoll

19 min


LAB concentrate

Protective medium:

(200 g/L sucrose)

Freezing (tray)


Product









Storage

16








Sensing method


time the CPP

Modeling & advanced

optimization tools

Chamber

Vacuum

Ice

Condenser

Mass transfer Heat transfer

Modelling and optimization toolModelling and optimization tool

A one-dimensional model of heat and mass transfer

17

Dry product

Vial bottom

Tray

Frozen

product

Dry product

Vial wall

Vacuum

Shelf

Product top

Sublimation front

Product bottom

Shelf

Product top

Sublimation front

Vacuum

Sensor

Parameter estimation using a specific tool developed by IIM-CSIC

Modelling and optimization toolsModelling and optimization tools

18www.iim.csic.es/~amigo

Modelling and optimization toolsModelling and optimization tools

Experimental description

� Six experiments

(Primary drying)

Estimation of parameters of the freeze-drying model

Mass transfer Heat transfer

19www.iim.csic.es/~amigo

(Primary drying)

� Product temperature, vapor

pressure

� No replica: experimental

errors were fixed at 10%


LAB concentrate

Protective medium:

(200 g/L sucrose)

Freezing (tray)


Product









Storage

20








Sensing method


time the CPP

Modeling & advanced

optimization tools

To measure or predict in real time the critical process parameters or product quality

Development of sensing methodsDevelopment of sensing methods

Common sensors

� Product temperature probe

� Vapor pressure

� Pressure rise test

Electronic nose

21

New sensors

� Electronic nose

� Acoustic impedance probe

Water activity Acidification activity



Common sensors


� Vapor pressure


Acoustic impedance probe

22

New sensors

� Electronic nose




Common sensors


� Vapor pressure


Acoustic impedance probe

23

New sensors

� Electronic nose



LAB concentrate

Protective medium:

(200 g/L sucrose)

Freezing (tray)


Product









Storage

24








Sensing method


time the CPP

Modeling & advanced

optimization tools

Acknowledgments:This work was financially supported by the European Community:

CAFÉ Project: Computer-Aided Food processes for control Engineering

25

Thank you for your attention

developing smart control strategies for freeze-drying of ... · developing smart control strategies...

Documents