and Models to Predict Spray-Dried Dispersion Particle Size John - - PowerPoint PPT Presentation
Application of Fundamental Relationships and Models to Predict Spray-Dried Dispersion Particle Size John Baumann, Sr. Principal Engineer Drug Product Development and Innovation Lonza - Bend, OR Lonza Pharma & Biotech | Drug Product
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 2
The Process
Nozzle Feed Solution Drug is dissolved with excipient(s) in a common solvent Drying Gas Drying Chamber The solution is spray-dried to remove the solvent
Atomizati tion
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 3
Solvent, Polymer and API
Solvent Solvent
Dr Droplet Formation Solv
Evaporation Skin in Formation Dr Drie ied Part rticle le
Ho Holl llow Sp Sphere
Tparticle > > Tboil PparticleSolventVapor > > PsprayD
yDryer
Col Colla lapsed Ho Hollo llow Sp Sphere
Tparticle < < Tboil PparticleSolventVapor < < PsprayD
yDryer
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 4
Manufacturability – Spray-drying Scale-up Performance Manufacturability – Powder Flow
20 40 60 80 100 120 140 500 1000 1500 2000 Droplet D[3,2] (µm) Flowrate (g/min)
PSD SD-1 PSD SD-2 PSD SD-4
y = 0.0807x + 0.9123 R² = 0.9214 2 4 6 8 10 20 40 60 80 100 ffc at 3kPa D(50) [µm]
10 20 30 40 50 60 70 30 60 90
Time (min) Drug Concentration (µg/mL)
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 5
Experimental Tools Empirical Models First Principles Models / CFD Increasing Complexity / Resource
Ashgriz, ed. Handbook of Atomization and Sprays. 2011.
Belhadef, et al. Pressure-swirl Atomization: Modeling and Experimental Approaches. Int J Multiphase Flow. 2012.
TSI I - PDP DPA Malv lvern - Spraytec
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 6
Pressure Swirl Nozzles
Arthur Lefebvre, Atomization and Sprays (1989)
Lefebvre Jones 𝐸 50 = 2.47𝜏0.25𝜈𝑀
0.16𝜈𝐻 −0.04𝜍𝑀 −0.22 ሶ
𝑛𝑀
0.315∆𝑄 𝑀 −0.47 𝑚0 𝑒𝑝 0.03 𝑀𝑡 𝐸𝑡 0.07 𝐵𝑡 𝐸𝑡𝑒𝑝 −0.13 𝐸𝑡 𝑒𝑝 0.21
𝐸 3,2 = 2.25𝜏0.25𝜈𝑀
0.25 ሶ
𝑛𝑀
0.25∆𝑄 𝑀 −0.5𝜍𝐵 −0.25
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 7
Assumptions:
concentration
particle size (no impact of morphology) Dr Droplet For
Skin in Formation
Solvent, Polymer and API
Dr Drie ied Part rticle le 𝐸 3,2 = 2.25𝜏0.25𝜈𝑀
0.25 ሶ
𝑛𝑀
0.25∆𝑄 𝑀 −0.5𝜍𝐵 −0.25
Cdrop*Vdrop=Cski
skin*Vskin
dski
skin=dSD SDD
BLD BLD-35 35 BLD BLD-200 200
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 8
Solu Solution Property Mea easurement te technique Res esult lts Viscosity Low shear (Hydramation ReactaVisc) 0.6 – 14 cP Surface Tension Tensiometer (Wilhelmy plate) 23-25 mN/m Skinning Concentration Thermogravimetric Analysis 21wt% Var ariable Mea easurement te technique Solvent Acetone Polymer HPMCAS-M Concentration range of HPMCAS 1wt% – 9wt% Dryer Scale BLD-35, BLD-200
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 9
By scale / nozzle type
y = 0.7562x - 1.1302 R² = 0.9773 y = 0.7498x - 5.9343 R² = 0.8554 10 20 30 40 50 60 10 20 30 40 50 60 Actual SDD D[3,2] (µm) Predicted SDD D[3,2] (µm) BLD-200 (Spraying Systems SK) BLD-35 (Schlick 121)
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 10
Combined data
y = = 0.6 .6812x - 2.4 .4577 R² ² = = 0.8 .8374 10 10 20 20 30 30 40 40 50 50 60 60 10 10 20 20 30 30 40 40 50 50 60 60 Act ctual l SDD D[3 D[3,2] (µm (µm) Predicted SDD D[3 D[3,2] (µm (µm)
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 11
Unrealistic values required (viscosity < pure acetone) to make predicted = actual
𝐸 3,2 = 2.25𝜏0.25𝜈𝑀
0.25 ሶ
𝑛𝑀
0.25∆𝑄 𝑀 −0.5𝜍𝐵 −0.25
y = 0.3666e0.4029x R² = 0.9929 y = 0.1202e0.2959x R² = 0.508 0.01 0.1 1 10 2 4 6 8 Viscosity (cP) [HPMCAS] in acetone (wt%) Measured viscosity Fit Viscosity Ace cetone Vis iscosity
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 12
Unrealistic values (>>100%) required to make predicted = actual
drop
kin*Vskin
100 200 300 400 500 600 700 800 900 2 4 6 8 10 Skinning concentration (wt%) Polymer concentration (wt%)
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 13
Adjusting model constant = 1.33, shows improved accuracy
𝐸 3,2 = 2.25𝜏0.25𝜈𝑀
0.25 ሶ
𝑛𝑀
0.25∆𝑄 𝑀 −0.5𝜍𝐵 −0.25
y = 1.149x - 2.4577 R² = 0.8374 10 20 30 40 50 60 10 20 30 40 50 60 Actual SDD D[3,2] (µm) Predicted SDD D[3,2] (µm)
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 14
❑ Application of empirical models and fundamental relationships can be applied for predicting SDD particle size as demonstrated by HPMCAS case study
✓ Sensitivity analysis of Lefebvre model identified gaps with viscosity value (e.g. as f(shear rate)) and constant – this was anticipated from the conditions which this model was developed ✓ Accuracy of Lefebvre droplet size prediction needs further refinement for polymeric spray solutions ✓ Skinning concentration was not found to have a large impact on the model prediction
❑ Ongoing work focuses on a range of formulation and process considerations to broaden the applicability to additional dispersion polymers
Lonza Pharma & Biotech | Drug Product Development and Innovation | John Baumann | 30 Oct 2018 15