Through Vial Impedance Spectroscopy

Transcription

1 (TVIS) A novel process analytical technology for the development of freezedrying processes and products Prof. Geoff Smith Leicester School of Pharmacy, De Montfort University, United Kingdom

2 Outline Introduction o o Description of Measurement System Theory (Dielectric loss mechanisms) TVIS Applications o o o o o o Ice Formation and phase separation in freezing Temperature calibration and prediction Drying rate estimation Heat transfer coefficient calculation Endpoint determination Dry Layer resistance and collapse (time permitting) Acknowledgements Yowwares Jeeraruangrattana GPO Thailand TVIS pass through on GEA Lyophil dryer, Hurth, Cologne 2

3 Introduction to the TVIS System Impedance spectroscopy characterizes the ability of materials to conduct electricity under an applied an oscillating voltage (of varying frequency) Impedance measurements across a vial rather than within the vial Hence Features Single vial nonproduct invasive Both freezing and drying characterised in a single technique Nonperturbing to the packing of vials Stopper mechanism unaffected 3

4 (TVIS) Description of Measurement System 4

5 Freeze drying chamber Junction box Passthrough TVIS measurement vial Resultant current Stimulating voltage LyoView TM analysis software LyoDEA TM measurement software TVIS system (I to V convertor) 5

6 TVIS Measurement System Adelphi VC01020C Adelphi VC00520C Adelphi VCD005 The designs of various vials that have been modified with copper foil electrodes (10 mm in height and 3 mm from the base of each container i. 20 mm crimpneck vial with 10 ml nominal capacity ii. 20 mm crimpneck vial with 5 ml nominal capacity iii. screwneck vial with 5 ml nominal capacity A B The different styles of a bespoke passthrough for TVIS systems A. Connected via the manifold hose on the outside of the dryer B. Connected via the port on top left side of the door on the dryer C. Connected to a port on the top of the drying chamber C 6

7 (TVIS) Dielectric Loss Mechanisms 7

8 Glass Measurement vial 1.1 mm 22 mm 1.1 mm Glass 10 mm Electronic polarization distortion of electrons relative to the nuclei Atomic polarization distortion of nuclei across a heteroatom bond by stretching and bending E OH OH OH OH OH H H H H H E Electronic polarization distortion of electrons relative to the nuclei Atomic polarization distortion of nuclei across a heteroatom bond by stretching and bending Instantaneous polarization dominant mechanism in the glass wall Protonhopping Conduction of protons in liquid water occurs through the Grotthuss "hopturn" mechanism Instantaneous polarization dominant mechanism in the glass wall Space charge polarization (weak frequency dependence) TVIS response for empty vial MW (spacecharge) polarization at glass wall sample interface Conductivity in pure water Dipolar polarization reorientation/alignment of permanent dipoles in liquid water (Debyelike δ relaxation) δ δ Polarization mechanisms in liquid water (relaxation time, ~ 9 ps at 20 o C) TVIS response for liquid water MW (spacecharge) polarization at glass wall sample interface Space charge polarization (weak frequency dependence) TVIS response for empty vial 8

9 Glass Measurement vial 1.1 mm 22 mm 1.1 mm Glass 10 mm Electronic polarization distortion of electrons relative to the nuclei E Dominant at T > 235 K (approx. 40 o C) Generation/migration of L and D orientation defects in ice Ih E Electronic polarization distortion of electrons relative to the nuclei Atomic polarization distortion of nuclei across a heteroatom bond by stretching and bending Atomic polarization distortion of nuclei across a heteroatom bond by stretching and bending Instantaneous polarization dominant mechanism in the glass wall Dominant at T < 235 K (approx. 40 o C) Generation/migration of H3O/OH ion pairs (ionic defects) in ice Ih (similar to the Grotthus mechanism) Instantaneous polarization dominant mechanism in the glass wall Space charge polarization (weak frequency dependence) TVIS response for empty vial MW (spacecharge) polarization at glass wall sample interface Polarization mechanism in ice TVIS response for frozen water (ice) MW (spacecharge) polarization at glass wall sample interface Space charge polarization (weak frequency dependence) TVIS response for empty vial 9

10 C / pf C / pf Frozen Water and Dielectric Relaxation of Ice I. : The polarization of the water dipole in liquid water at 20 C, with a dielectric loss peak frequency of ~ 18 GHz II. : The MaxwellWagner (MW) polarization of the glass wall of the TVIS vial at 20 C, with a dielectric loss peak frequency of 17.8 khz III. : The dielectric polarization of ice at 20 C, with a dielectric loss peak frequencies of 2.57 khz IV. : The dielectric polarization of ice at 40 C with a dielectric loss peak frequencies of 537 Hz (IV) 20 C 40 C 20.3 C khz 20.3 C 20 C 40 C 537 Hz (III) 18 khz (II) Real part Capacitance 18 GHz (I) Imaginary part Capacitance 10

11 (TVIS) Applications 11

12 C / pf C /pf C / pf increase (TVIS) Monitoring Phase Behaviour (ice nucleation temperature and solidification end points by using F PEAK F PEAK temperature calibration for predicting temperature of the product in primary drying Surrogate drying rate (from dc PEAK) dt Solid State F PEAK Liquid State F PEAK o C 18 o C C PEAK Drying time C (~ 100 khz) is highly sensitive to low ice volumes; therefore it could be used for determination end point of primary drying 12

13 TVIS Response Surface (3DPlot) Imaginary Part of Capacitance Real Part of Capacitance Annealing = Reheating and Recooling Liquid state Reheating Frozen solid Recooling Liquid state Frozen solid Reheating Recooling Primary drying Primary drying low frequency low frequency low frequency High frequency Intermediate frequency 13

14 C / pf C / pf Phase Separation in Freezing Step Water (frozen) 0.5 Liquid state Ice peak 0.6 5%w/v Lactose solution (frozen) Liquid state Unfrozen fraction peak Ice peak C ice C unfrozen C G C ice C G R ice R ice R unfrozen ice Unfrozen Fraction Electrode Glass Wall Electrode Glass Wall Microstructure Microstructure 14

15 C /pf Dielectric loss spectrum Data analysing software (LyoView ) identifies the peak frequency (F PEAK ) and peak amplitude (C PEAK ) in the imaginary part of the capacitance spectrum C PEAK F PEAK 15

16 (TVIS) Temperature calibration & Temperature prediction 16

17 Temperature / o C Temperature / o C C / pf Temperature Calibration Shelf Temp. T TC 1 T TC Time /h Polynomial coefficient from Log F PEAK temperature calibration a b c TE BE 5.43 x C 15.5 C y = 0.543x x R² = Bottom electrode y = x x R² = Log F PEAK 17

18 Temperature / o C Temperature / o C Temperature Prediction 5 Temp. constant 28 Temp. constant Shelf Temperature (T s ) T TC 1 T FPEAK BE T TC 2 T FPEAK TE T TC 2 T TC 1 T FPEAK BE T FPEAK TE T (Pi =P C@270μbar ) = Time / h Time / h 18

19 (TVIS) Drying rate estimation 19

20 Ice height (h)/ mm Temperature / o C Drying Rate Estimation 28 I II 32 T avg ~32 C T b Temp. constant T i = 33.1± 5 C T b = 29.8± 3 C Time / h Time / h T i 20

21 Ice height (h)/ mm Temperature / o C Drying Rate Estimation Time / h h (2.8 h) mm T avg ~32 C Temp. constant T i = 33.1± 5 C T b = 29.8± 3 C h (2 h) mm Time / h T b T i Drying rate during the steady state Drying rate ( m t ) = ρ i A h (t1) h (t2) t 2 t 1 Ice density (ρ i ) at 32 C = g cm 3 (Calculated ice temperature between T i & T b ) Internal vial diameter (VC01020C) = 2.21 cm Crosssection area (A) = 3.80 cm 2 Ice height at 2 h (h (2 h) ) = mm Ice height at 2.8 h (h (2.8 h) ) = mm TVIS parameters used for determination: m t = 0.42 g h1 T b = 29.8 C Drying rate = g cm cm cm h = g h 1 21

22 (TVIS) Heat Transfer Coefficient Determination 22

23 Over all heat transfer Coefficient (cals 1 cm 2 K 1 ) Heat Transfer Coefficient (K v )Determination Parameters Drying rate at steady state (g/h) (22.8 h into primary drying) L m t = A ek v (T s T b ) TVIS 0.42 Shelf Temperature, T s (K) Vial s base Temperature, T b (K) K v = L is the latent heat of sublimation of ice (2844 J g 1 or cal g 1 ) and A e is external crosssectional area of the base of the TVIS vial (4.62 cm 2 ) 1.4E3 1.2E3 1.0E3 8.0E4 6.0E4 4.0E4 2.0E4 E0 TVIS Tchessalov S (2017) Kuu W (2009) Brülls M (2002) L m t A e (T s T b ) K v (270 bar) = K v (270 μbar) = cal s 1 cm 2 K 1 K v 270 bar) 5.73 x Pressure (mtorr) L m t A e (T s T b ) = cal g g h cm K = 2.06 cal h 1 cm 2 K 1 = cal s 1 cm 2 K 1 23

24 (TVIS) Endpoint Determination 24

25 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) 2.0 Secondary process Primary process for ice High frequency response High frequency response khz Primary Drying time profile of C' values at a peak frequency of 100 khz, i.e. Log F = Time /h 25

26 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 26

27 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 27

28 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 28

29 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) Ice layer detaches from the wall khz Time /h 29

30 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) Capacitance recovers as ice layer recedes across the base 100 khz Time /h 30

31 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 31

32 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 32

33 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 33

34 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 34

35 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 35

36 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 36

37 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 37

38 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 38

39 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 39

40 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 40

41 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) khz Time /h 41

42 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) Last bit of ice crystal khz Time /h 42

43 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) Last ice crystal disappears khz Time /h 43

44 C' (Real part)/ pf C' (100 khz)/ ff C' (Real part)/ pf End Point Determination (T s 15 C and P c 400 µbar) No change khz C' (100 khz) reaches what looks like a plateau Time /h 44

45 C' (100 khz)/ ff Summary End point determine by high frequency (i.e. 100 khz for ice) real part capacitance (i.e. 100 khz for ice) Last bit of ice crystal Time /h 45

46 Summary End point determine by high frequency (i.e. 100 khz for ice) real part capacitance (i.e. 100 khz for ice) Last bit of ice crystal 0.15 mg 46

47 (TVIS) Product Resistance Determination 47

48 Drying rate / R p (cm 2 Torr h g 1 ) Pressure / torr Temp. / o C Product Resistance (R P ) Determination 5 Rp = P ICE P CHAMBER dm dt A P T(FPEAK) Tcondenser 32 C ln(p ICE ) = Tp or T C Torr Dry Layer Thickness (cm) 0.1 P ICE P CHAMBER 02 Torr 1.75cm2 Torr h g1 Lactose freezedried matrix h g m t Time/h 48

49 R p (cm 2 Torr h g 1 ) Product Resistance (R P ) Determination Rp = P ICE P CHAMBER dm dt 3 mm A P Dry Layer Thickness (cm) 500X 500X Top layer Fine pores Middle layer Microcollapse 3 mm 500X Bottom layer Full collapse 49

50 (TVIS) Collapse Phenomena 50

51 C PEAK / pf C / pf Temperature / o C 5% Sucrose I II III IV T (TC) V Shelf Temperature T (FPEAK ) T g 33.7 C T (Pi )36.4 C dĉ PEAK dt Ĉ peak C peak = h (37.8 C) h (37.9 C) h (37 C) 23.8 h (36.41 C) Time / h 51

52 C PEAK / pf C / pf Temperature / o C 5% Sucrose in 0.26% NaCl I II III IV V Shelf Temperature T (FPEAK ) T (TC) T (Pi )36.4 C 40 T g 38.0 C Ĉ peak dĉ PEAK dt = h (37.8 C) h (37.4 C) h (37 C) h (36.6 C) Time / h

53 C PEAK / pf C / pf Temperature / o C 5% Sucrose 0.55% NaCl I III II IV V Shelf Temperature T (FPEAK ) T (TC) T (Pi )36.4 C 40 T g 41.8 C dĉ PEAK dt = 864 Ĉ peak C peak Time / h h (38 C) h (38.1 C) h (38 C) 23.9 h (36.9 C) 53

54 Summary results of sugar salts solutions Parameters Unit S1 S2 S3 Component Sucrose %w/v Sodium chloride %w/v T g (DSC) C Temperature stabilization period III to IV III to IV III to IV (Before ramping shelf temperature) h T (FPEAK ) C 36.0 to to 37.7 Surrogate drying rate (dĉ PEAK dt) pf/h Structural state Noncollapse microcollapse? Microcollapse VII to VIII VII to VIII VII to VIII Temperature stabilization period h (After shelf temperature constant) C 26.4 to to to 32.7 Surrogate drying rate (dĉ PEAK dt) pf/h Structural State* Microcollapse Collapse Collapse 54

55 C / pf C / pf 5% Sucrose in Salt Solutions h (36.6 C) h (31.8 C) h (30.1 C) h (28.8 C) h (36.9 C) h (33.5 C) h (32.7 C) h (33.4 C) % sucrose in 0.26% NaCl 5% sucrose in 0.55% NaCl 55

56 Limitations C PEAK and F PEAK parameters rely on intimate contact of ice cylinder with glass wall C (100 khz) parameter does not dependent on contact and can be used for end point but relationship between C (100 khz) ice constant is nonlinear Cable length limited to 1m at present CTVIS not compatible with front loading system Incompatible with TCs in same TVIS vial (use fibre optic sensors INFAP) 56

57 Future Work Development mapping a drying characteristics o heat transfer coefficients (K V ) o dry layer resistance (R P ) Instrument Development o Commercial CTVIS (2018) o Noncontact TVIS (201819) Microwell screening Vial clusters in batch FD o TVIS Shuttle (201920) Noninvasive real time information for characterising the freeze drying 57

58 Acknowledgements, Recent Projects & Collaborators De Montfort University, School of Pharmacy o o o o Evgeny Polygalov: coinventor of TVIS instrument Yowwares Jeeraruangrattana. PhD student Bhaskar Pandya. PhD student Irina Ermolina. Senior Lecturer Government Support for industry LyoDEA Lyophilization process analytics By dielectric analysis Biopharmaceutical Stability at Room Temperature Analytical Technologies for the Stabilization of Biopharmaceuticals 58