Electroactive controlled release thin films

SI Text

Structure of Prussian Blue. SI Fig. 6 depicts the structure of "soluble" Prussian Blue (PB), KFe^III[Fe^II(CN)₆]. Potassium inclusions in this form of PB can dissociate in aqueous solutions, resulting in a net negative charge on the particle surface that renders nanoparticles stable in solution. PB nanoparticles are formed by means of the room-termperature, aqueous-phase reaction that occurs upon the addition of a molar excess of iron (II) chloride to potassium ferricyanide (see Materials and Methods). This synthesis and purification procedure yielded dark blue, aqueous suspensions that were stable in the absence of sonication or chemical stabilizers.

Effect of the Applied Potential on Total Film Thickness. To measure the effect of an applied potential on film thickness, and specifically to determine whether PB loss correlated with deconstruction of the entire film, PB-containing films were exposed to a constant potential of 1.25 V and thickness was measured with respect to time using profilometry (SI Fig. 7). In all PB-containing films, film thickness was observed to decrease rapidly at early times (1-5 min) followed by a more gradual decrease at later times (5-60 min), kinetics which reflect the time scales for PB loss described in Fig. 2. Furthermore, in all systems, thickness was observed to decay until reaching 10-20% of the original film thickness, suggesting that some residual material was remaining on the surface of the substrate. In keeping with this observation, we also observed that 50-70% of the film's incorporated ¹⁴C-DS is released actively, leaving behind a fraction that is likely bound within this portion of the film structure. The amount of material bound at the surface (relative to the total quantity of material in the film) may vary with changes in film composition, deposition conditions, and total film thickness. In (LPEI/PB/LPEI/¹⁴C-DS)₃₀ tetralayer films, thickness decayed to 80% of that of the original film after 1 min at 1.25 V, 50% after 5 min, and 20% after 60 min. In (LPEI/PB)₂₀ systems, destabilization occurred on a more rapid time scale, reaching 43% in under 1 min and 20% in under 4 min. The shorter time scales required for destabilization in (LPEI/PB)₂₀ systems likely reflects the fact that these films lose all cohesive electrostatic interactions after the PB-to-PX transition, resulting in rapid deconstruction relative to (LPEI/PB/LPEI/DS)₃₀ systems, which retain some stable electrostatic interactions (from LPEI and DS) in the presence of an applied potential.

Effect of Applied Potentials Less Than 1.25 V on Film Stability and ¹⁴C-DS Release. SI Fig. 8 shows voltage-dependent release of ¹⁴C-DS after 10 min at the indicated square wave voltage. As expected, the amount of ¹⁴C-DS released from the film increases with an increase in applied voltage. The formal potential for the PB/PX redox pair is ~0.82-0.85 V with a relatively broad peak (by cyclic voltammetry; nearly complete conversion to the PX state occurs at voltages exceeding 1.2 V) (1). At 0.5 V, all of the nanoparticles are in the fully charged PB state and minimal ¹⁴C-DS release is observed. At 0.75 V, a fraction of the Fe(II) centers have been oxidized to Fe(III). Therefore, the surface charge density of the nanoparticles has decreased, inducing partial film destabilization and ¹⁴C-DS release. Likewise, at 1.00 V, even more of the Fe(II) centers have been oxidized to Fe(III) and further film destabilization and ¹⁴C-DS release occurs. Finally, at 1.25 V, nearly all of the iron centers are in the Fe(III) oxidation state, corresponding to the Prussian Brown (PX) state, at which complete film destabilization and ¹⁴C-DS release occurs.

Chronoamperometry and Power Requirements of Thin Films. SI Fig. 9 shows the chronoamperometric response of (LPEI/PB/LPEI/¹⁴C-DS)₃₀ films subjected to various potentials ranging from 0.5 V to 1.25 V vs. SCE for 30 min. In all cases, the current decays rapidly to »0 within 15-20 s.

SI Fig. 10 shows the maximum current flow to (LPEI/PB/LPEI/¹⁴C-DS)₃₀ films subjected to potentials in the range of 0.5-1.25 V (vs. SCE). As expected, there is an increase in maximum current for an increase in the applied oxidative potential. At higher potentials, more of the PB iron centers become oxidized, and thus a higher initial current results. The total capacity (representing the number of redox centers oxidized) can be calculated by integrating under the current versus time curve (SI Fig. 9). We calculated a total capacity of 24,000 mA-s for (LPEI/PB/LPEI/¹⁴C-DS)₃₀ films subjected to a constant potential of 1.25 V (vs. SCE) for 30 min.

SI Fig. 11 shows the power requirements of (LPEI/PB/LPEI/¹⁴C-DS)₃₀ films subjected to various potentials (relative to SCE). Power was calculated by multiplying current response (SI Fig. 9) by the applied potential. Again, as expected, application of a higher oxidative potential has the greatest power requirement. The total energy requirement for electrochemical switching can be calculated by integrating under the power versus time curve. We found a total energy requirement of 30,000 mW-s for (LPEI/PB/LPEI/¹⁴C-DS)₃₀ films subjected to a constant potential of 1.25 V (vs. SCE) for 30 min.

Comparison of Power Requirements for Electroactive Thin Films with Other Common Implantable Systems. This section serves to compare power requirements for the PB-based systems described in this study with those for other implantable devices such as pacemakers. The capacity of a pacemaker battery is typically in the range of 1-1.5 A-h (Medtronic, Inc., personal communication). To apply 1.25 V to a (LPEI/PB/LPEI/¹⁴C-DS)₃₀ film for 30 min requires a capacity of 6.67e-6 A-h, well below the capacity of implanted pacemaker batteries. We can also compare the energy required for typical pacemaker output with that required for dissolution of our films. A typical pacemaker applies a voltage of 1-5 V over a resistance of 500-1,000 ohms for a duration of 0.3-1 ms (Medtronic, Inc., personal communication). Using these values, the range of energy requirements is 0.3-50 mJ. This is in relatively close agreement with Mallela et al. (2). Mallela et al. also report the typical energy requirement of implantable cardioverter defribrillators as 15-40 J. From the chronoamperometry data above, applying 1.25 V to a (LPEI/PB/LPEI/¹⁴C-DS)₃₀ film for 30 min requires 0.03 J of energy. This is 2.5-5 orders of magnitude greater than the energy needed for a pacemaker pulse, but it is 1.5-3 orders of magnitude less than the energy needed for an implantable defribrillator. Again using typical values for pacemakers to give a conservative estimate, the maximum power requirement (1 V over 1,000-ohm resistance) is 1 mW. Typical power requirements for implantable blood pumps range from 3 to 15 W (3). In applying 1.25 V to a (LPEI/PB/LPEI/¹⁴C-DS)₃₀ film, the maximum power requirement is 2.4 mW; this is approximately three orders of magnitude less than that for a pacemaker and six to seven orders of magnitude less than that for implantable blood pumps.

Calculation of Film Loading Capacity. Drug or chemical agent loading in the electroactive LbL thin films described in this work can vary based on factors such as the choice of drug species, film deposition conditions (namely salt concentration and pH), film surface area, and the number of deposited layers. We estimate that films used in the present study, which are deposited onto flat, conducting substrates with a surface area of 2.45 cm² and contain ¹⁴C-dextran sulfate (DS), can be loaded with between 50-300 ng DS/(cm²â€¢tetralayer). Thus, we estimate that a single, 100-nm-thick film (24 tetralayers at 4.2 nm per tetralayer) deposited onto a surface area of 1 cm² can load and release 1.2-7.2 mg of ¹⁴C-DS. Furthermore, in many cases it may be possible to substantially increase loading beyond this degree by increasing the surface area of the deposited film, its thickness, and/or deposition conditions.

Methodology Regarding PB Toxicity Studies. The toxicity assays reported in this study were conducted on PB nanoparticles suspended in solution instead of the film's entire set of constituents (DS, LPEI, and PB). We elected to perform toxicity studies in this way because PB is the electroactive, functional component in these systems, and as such, it is the only essential component that will be present in any electroactive thin film of this type (regardless of drug choice and other architectural components). DS and LPEI, on the other hand, were used in these films only to demonstrate the concept of electroactive controlled release: DS was chosen as a model biomolecule that can be obtained in pure, radiolabeled form, and LPEI was chosen as a counter polyion. Therefore, whereas next generation electroactive films are almost certain to contain PB, they are unlikely to contain LPEI or DS, which will instead be replaced by an active system (for example, a therapeutic molecule). Therefore, toxicity measurements on DS and LPEI were omitted from the present study so as not to obfuscate the more clinically relevant findings regarding the toxicity of PB.

1. DeLongchamp DM, Hammond PT (2004) Adv Funct Mater 14:224-232.

2. Mallela VS, Ilankumaran V, Rao NS (2004) Indian Pacing Electrophysiol J 4:201-212.

3. Geertsma RE, deBruijn ACP, Hilbers-Modderman ESM, Hollestelle ML, Bakker G, Roszek B (2007) New and Emerging Medical Technologies: A horizon scan of opportunities and risks [National Institute for Public Health and the Environment (RIVM), The Netherlands].