Continuous Nanoparticle Production Using Focused Acoustics

Continuous nanoparticle production using focused acoustics

Many drugs currently in development are either water insoluble or lipophilic, thus creating formulation challenges to administer these drugs at a high dose.  In some situations, nanoparticles are an appropriate formulation strategy as a result of their intrinsic properties such as small shape and size (e.g., radius) and large surface area to volume ratios.  Nanoparticles can be generated by grounds up crystallization, which is typically processed in batch mode.  One basic approach of crystallization is to vary the solubility of the solute of interest with different methods, such as vary the ratio to non-solvent vehicle (anti-solvent).  

Low frequency, unfocused ultrasound has previously been demonstrated for improving crystallization kinetics.  However, unfocused ultrasound has efficiency limitations due the inability to focus the energy, creating issues in thermal management such as non-uniform mixing, and uncontrolled thermal fields.  Additionally, these unfocused solutions are not able to provide sufficiently tight, uniform nucleation control at the chemical solvent boundary interface that is required for nanoparticle production.  The Covaris Adaptive Focused Acoustics? (AFA) technology is a highly controlled mechanical based process that provides a practical tool to precisely control both the nucleation formation and the rate of crystal growth.  The high energy efficiency of AFA avoids excessive heat generation, reducing potential thermal damages to the sample and allowing greater control over thermally dependent chemical reactions.  Moreover, the precisely controlled energy delivery improves the reproducibility of the sample processing for higher quality control.

As a model of nanoparticle formation, a Solvent:Anti-solvent crystallization method with an active AFA energy field was tested in different batch volumes as well as in a continuous flow format.  Cinnarizine was used as a model compound.  Nanoparticles were measured using a Malvern Mastersizer 3000, and particle sizes were reported for 10^th percentile, 50^th percentile, and 90^th percentile populations (d10, d50, and d90 populations respectively). 

Initial studies were performed in triplicate on batch scale processing of 2 mL samples.  Parameters evaluated include:

Solvent                                                 Dimethylacetamide (DMA)

Drug solvent concentrations       50 mM, 125 mM, 250 mM

Stabilizer                                              0.2% PVP30 and 0.25 mM SLS

Stabilizer percentage                     0.2% 1.5%, 3.0%

Solvent:Anti-solvent ratio            1:29, 1:14, 1:6.6

For comparison, a control was run at each setting.  The control was mixing 2 mL samples via a magnetic stir bar in a 20 mL scintillation vial for the same time duration of 5 minutes.

In all cases, the particle sizes of the AFA-processed samples were lower and more reproducible than the control.  A typical comparison is shown in the following tables:

Table 1. Results from the stir bar method (Control).

All values are in microns

Table 2. Results from the AFA method.

All values are in microns

d50 particle size populations for the control sample averaged 27.7 microns while the same method processed with AFA yielded average d50 particle size populations of 122 nanometers.

A continuous manufacturing process was evaluated with the following settings:

Drug solvent concentration         125 mM  (46 mg/mL)

Solvent flow rate                             800 µL/min

Anti-solvent flow rate                    10 mL/min

Stabilizer                                              0.2% PVP30 and 0.25 mM SLS

Stabilizer percentage                     0.2%

Solvent:Antisolvent ratios            1:12.5  (10 mL/min anti-solvent, 800 µL/min solvent)

With 125 mM at a 800 µL/min solvent flow rate, the production rate was 36.5 mg/min.

Higher total acoustic power input (Peak Incident Power) yielded both smaller particle sizes and repeatable results.  At PIP values of 500 Watt, particle size distributions are listed in the following table.

Table 3. Particle size distributions at PIP = 500 W.

All values are in microns

Control of particle size was demonstrated by varying PIP.  Particles sizes at 300 PIP are listed in the following table.

Table 4. Particle size distributions at PIP = 300 W.

All values are in microns

The current study demonstrates the technical feasibility of using Adaptive Focused Acoustics (AFA) to control nucleation and crystal growth for continuous flow Solvent:Anti-solvent crystallization.  Nanoparticles were successfully and reliably produced in a continuous flow process with d50 particle size populations below 200 nm.  By using the AFA system to control the crystallization process, a choice of nano- or micron-sized crystals can be easily generated with a simple change of PIP process parameter.  In addition, the isothermal, reproducible AFA process enables control of fluid shear effects on secondary nucleation to minimize crystal size, accelerate the overall rate of crystal formation, and improve yield.  The effect of AFA on the morphology of the crystal is currently under investigation.