Aerosol delivery of nanoparticles in uniform mannitol matrices formulated by ultrasonic spray freeze drying

Suzanne M. D’Addio1,2, John Gar Yan Chan1,Philip Chi Lip Kwok1, Bryan Benson2, Robert K. Prud’homme2*, Hak-Kim Chan1*

1Advanced Drug Delivery Group, Faculty of Pharmacy, The University of Sydney, Camperdown, NSW 2006, Australia

2Chemical and Biological Engineering, Princeton University, Princeton, NJ, USA 08854

SUPPLEMENTAL INFORMATION

Characterization of spray freeze dried mannitol/PEG4kparticles

Three formulations of mannitol/PEG4k were prepared by spray freeze drying in order to determine if the presence of PEG5k in the NP corona during spray freeze drying was causing the change in the mannitol polymorph in the dry powder. The weight ratios of 20:1, 3:1, and 0.6:1 were chosen to determine if there was a threshold below which the presence of the alternate mannitol polymorph was not observed. The compositions of the aqueous formulations are detailed in Table S1.

Table S1. Formulations of the aqueous solutions spray freeze dried

Formulation / (Ratio Man:PEG) / Aqueous solution composition
Mannitol (mg mL-1) / PEG4K (mg mL-1)
SFD09 / 20:1 / 10 / 0.5
SFD10 / 3:1 / 10 / 3
SFD11 / 0.6:1 / 10 / 18

After spray freeze drying, a sample of each powder was characterized by DSC, with two thermal scans on each sample. The PEG4k sample had one melting endotherm from 50 – 65 °C on the first scan. After allowing the sample to cool and scanning again, a bimodal peak is observed from 50 – 65 °C, which may be due to polydispersity in the molecular weight of the sample. Each of the samples of spray freeze dried mannitol/PEG4k showed the same unimodal peak on the first scan and a bimodal peak on the second scan. The area under the PEG4k melting peak decreased as the weight fraction of mannitol increased, since the heat flow for each scan is normalized by the total sample weight. The melting and recrystallization transition from 145 – 160 °C, immediately prior to the melting event at 160 °C, was present on the first scan of all samples of mannitol/PEG4k, indicating that the δ polymorph (42) was present in all samples prepared, regardless of the concentration of PEG4k in the sample. This transition was irreversible in all mannitol/PEG4k samples and was not observed on the second scan of each sample.

Figure S1. DSC traces for PEG4k, and spray freeze dried mannitol/PEG4kpowders with weight ratios of 0.6:1 (SFD11), 3:1 (SFD10), and 20:1 (SFD09).

The peaks in each thermal trace were integrated and then normalized by the weight fraction of either mannitol or PEG4k in the sample, as appropriate. The results are summarized in Table S.2. The PEG4k transition was integrated as a single event.The enthalpy of the transition was found to be consistent for all samples, and the average enthalpy for all 8 measurements is -25.9 ± 3.6 W°Cg-1. Similarly, when the transition from 145 – 160 °C was integrated as a single event, and normalized by the fraction of mannitol in each sample, the values are in good agreement, despite the different weight fractions of PEG4k. Finally, the melting enthalpies for the β mannitol were also in good agreement, and the average for the six measurements was -45.7 ± 5.1 W°Cg-1.

Table S2. Integration values for the thermal transitions in the DSC traces in Fig. S.1.

Normalized Enthalpy of Transition (W °C g-1)
Peak / PEG4k / δ mannitol / β mannitol
Scan / 1 / 2 / 1 / 1 / 2
Raw PEG4k / -31 / -27 / - / - / -
SFD09 (20:1) / -19 / -28 / 1.4 / -46 / -46
SFD10 (3:1) / -23 / -26 / 1.5 / -43 / -55
SFD11 (0.6:1) / -28 / -25 / 1.7 / -40 / -44

The powder XRD data further confirms that the δ polymorph is prepared in all samples of spray freeze dried mannitol/PEG4k. The crystalline peaks from PEG4k, seen in the raw powder XRD at 19.2⁰ and 23.5⁰ 2θ, are clearly present at low ratios of mannitol/PEG4kas well. The other peaks in the diffraction patterns for mannitol/PEG4k powders are present in all three spray freeze dried samples, in different relative ratios compared to the PEG4k peaks, but do not correspond to the peaks in the raw mannitol diffraction pattern. A scattering peak at 9.75⁰ 2θ was further evidence that the δ polymorph (42) was present in all samples.

Figure S.2.Powder XRD patterns for PEG4k, SFD mixtures of mannitol/PEG4k, and mannitol.

1