Green Amorphous Nanoplex as a New Supersaturating DrugDelivery SystemWean Sin Cheow and Kunn Hadinoto*

School of Chemical and Biomedical Engineering, Nanyang Technological University, Singapore 637459

*S Supporting Information

ABSTRACT: The nanoscale formulation of amorphous drugs represents a highly viable supersaturating drug-delivery system forenhancing the bioavailability of poorly soluble drugs. Herein we present a new formulation of a nanoscale amorphous drug in theform of a drug−polyelectrolyte nanoparticle complex (or nanoplex), where the nanoplex is held together by the combination of adrug−polyelectrolyte electrostatic interaction and an interdrug hydrophobic interaction. The nanoplex is prepared by a trulysimple, green process that involves the ambient mixing of drug and polyelectrolyte (PE) solutions in the presence of salt.Nanoplexes of poorly soluble acidic (i.e., ibuprofen and curcumin) and basic (i.e., ciprofloxacin) drugs are successfully preparedusing biocompatible poly(allylamine hydrochloride) and dextran sulfate as the PE, respectively. The roles of salt, drug, and PEin nanoplex formation are examined from ternary phase diagrams of the drug−PE complex, from which the importance ofthe drug’s charge density and hydrophobicity, as well as the PE ionization at different pH values, is recognized. Under theoptimal conditions, the three nanoplexes exhibit high drug loadings of ∼80−85% owing to the high drug complexation efficiency(∼ 90−96%), which is achieved by keeping the feed charge ratio of the drug to PE below unity (i.e., excess PE). The nanoplexsizes are ∼300−500 nm depending on the drug hydrophobicity. The nanoplex powders remain amorphous after 1 month ofstorage, indicating the high stability owed to the PE’s high glass-transition temperature. FT-IR analysis shows that functionalgroups of the drug are conserved upon complexation. The nanoplexes are capable of generating prolonged supersaturation upondissolution with precipitation inhibitors. The supersaturation level depends on the saturation solubility of the native drugs, wherethe lower the saturation solubility, the higher the supersaturation level. The solubility of curcumin as the least-soluble drug ismagnified 9-fold upon its transformation to the nanoplex, and the supersaturated condition is maintained for 5 h.

1. INTRODUCTIONThe development of various methods of transforming crystallineparticles of an active pharmaceutical ingredient (API) into theirmore soluble forms is driven by the realization that a large fraction(∼66%) of new promising drug candidates exhibit low saturationsolubility in the aqueous phase.1 The low aqueous solubility of theAPI translates to its low oral bioavailability, which necessitates highand frequent dosing, causing high financial and pill burdens forpatients. The three major formulation strategies for bioavailabilityenhancement include (1) transforming the API into its highlysoluble salt form, (2) formulating the API into crystallinenanoparticles (or nanoAPI), and (3) delivering the API in itsamorphous form.2

Because a majority of drugs are either weak organic acids orbases,2 salt formation is the first approach in increasing thesaturation solubility of acidic and basic drugs because of its

preparation simplicity.3 However, acidic and basic drugs may notnecessarily form salts, and salt formation does not necessarilyguarantee enhanced saturation solubility.1 The nanoAPI strategy,however, is not limited to acidic and basic drugs because itssaturation solubility enhancement relies on a particle sizereduction to the nanoscale following the Ostwald−Freundlichsolubility theory.4 However, the Ostwald−Freundlich equationsuggests that the saturation solubility enhancement viananoionization is significant only for size ≪100 nm.4 NanoAPIwith sizes of ∼150−200 nm has been found to exhibit only a15% higher saturation solubility than the microscale counter-parts.5 Presently, established nanoAPI preparation techniques

Received: December 5, 2011Revised: February 9, 2012Published: March 22, 2012

Article

pubs.acs.org/Langmuir

© 2012 American Chemical Society 6265 dx.doi.org/10.1021/la204782x | Langmuir 2012, 28, 6265−6275

(e.g., high-pressure homogenization, wet milling) are not yetcapable of consistently producing nanoAPI at sizes below 100 nm.6

Therefore, the potential of nanoAPI as a bioavailability-enhance-ment strategy depends on advancements in its preparationtechniques.A different strategy is to exploit the metastable state of the

amorphous form of the API to generate an apparent temporaryincrease in the solubility. The dissolution of amorphous APIresults in a highly supersaturated solution producing apparentsolubility that is much higher than the saturation solubility ofthe crystalline counterparts.7 The high supersaturation levelwould drive drug absorption across the gastrointestinal lumen,resulting in enhanced bioavailability, provided that thesupersaturation can be maintained for a time period sufficientfor absorption. It is significant that the in vitro generation ofhigh supersaturation levels by amorphous API has been shownto translate to enhanced bioavailability in vivo.8,9

Amorphous API is typically prepared in the form ofmicroscale solid dispersions of the drug stabilized by highglass-transition-temperature polymers (e.g., hydroxypropylme-thylcellulose (HPMC), poly(ethylene glycol)) by cogrinding,hot melt extrusion, or antisolvent precipitation techniques.10,11

The roles of the stabilizer, which usually makes up at least 50%(w/w) of the formulation,12 are twofold: (1) to preventrecrystallization of the amorphous API during storage byoccupying its high-energy surfaces and (2) to function asprecipitation inhibitors of the dissolved drug to prolong thesupersaturation. The typical amorphous API formulations,however, suffer from drawbacks of (1) low drug loading due tothe high stabilizer content and (2) a high propensity torecrystallize upon dissolution, attributed to slow dissolutionrates of the amorphous API, particularly when slowly dissolvingstabilizers (e.g., HPMC) are used,13 resulting in low apparentsolubility.The in-solution recrystallization propensity of the amorphous

API can be solved by having a nanoscale formulation,12 where thehigh specific surface area of nanoparticles ensures a high dissolutionrate. A number of nanoscale amorphous APIs (or amorphousnanoAPIs) having drug loadings of up to ∼90% (w/w) have beensuccessfully prepared by a wide range of techniques (e.g., anti-solvent precipitation and sonoprecipitation).14−16 These techniques,however, exhibit several weaknesses, such as intricate processes, theheavy use of organic solvents, high-energy expense, and wide sizedistributions of the products.In an earlier study,17 we have developed a new method of

preparing stable amorphous nanoAPIs having uniform sizes viaa self-assembly drug−polyelectrolyte complexation processwhere the said method is simple, solvent-free, energy minimal,and fast. Using dextran sulfate (DXT) as the polyelectrolyte(PE), we have shown that a poorly soluble amphoteric APIsuch as ciprofloxacin (CIP), after being transformed into itscations in an acidic environment below its pKa1, self-assemblesupon interactions with oppositely charged DXT to form theCIP−DXT nanoparticle complex (or CIP nanoplex) in thepresence of salt. The formation of the CIP nanoplex, whichexhibits drug loading of up to ∼80% (w/w), is driven by CIP−DXT electrostatic interactions coupled with hydrophobicinteractions of the amphiphilic CIP molecules. The stabilityof the dry-powder form of the amorphous CIP nanoplex after1 month of storage has also been demonstrated.Herein we investigate the feasibility of extending the self-

assembled drug−PE complexation process to prepare nano-plexes of poorly soluble weak acid APIs such as ibuprofen (IB)

and curcumin (CCM). Poly(allylamine hydrochloride) (PAH),a synthetic polycation commonly used in drug and genedelivery,18 is used as the PE. Ternary phase diagrams of thedrug−PE complex are constructed to determine the range ofcompositions of salt, drug, and PE required for successful IBand CCM nanoplex formation. The findings are compared withthat of the CIP nanoplex prepared under the same conditions.The phase diagrams are also used to elucidate the roles of PE,drug, and salt in nanoplex formation. The optimal formulations,in terms of the resultant production yield and complexationefficiency, are subsequently identified. The drug loading, size,zeta potential, supersaturation profile, and storage stability ofthe optimal nanoplex formulations are examined. The resultsdemonstrate the amorphous nanoplex as a very capable andgreener alternative to a supersaturated drug delivery system.

2. MATERIALS AND METHODS2.1. Materials. CIP, sodium chloride (NaCl), potassium bromide

(KBr), potassium hydroxide (KOH), hydroxypropylmethylcellulose(HPMC), glacial acetic acid, and phosphate-buffered saline (PBS, pH7.4) were purchased from Sigma-Aldrich (USA). PAH (MW 120 000−200 000 Da), IB (an anti-inflammatory drug), and CCM (a naturalagent with various therapeutic functions) were purchased from AlfaAesar (USA). DXT (MW 5000 Da) was purchased from Wako PureChemical Industries (Japan). The hydrophobicities of the amphiphilicAPIs, as measured by their octanol/water partition coefficient values(log P),19 are presented in Table 1. Saturation solubilities of the nativeAPI crystals and pKa

19 are also presented in Table 1.

2.2. Preparation of the Amorphous Drug Nanoplex. Depend-ing on its solubility as a function of pH, the drug was dissolved in anacidic or basic aqueous solution to form its cations or anions,respectively. CIP, being an amphoteric drug,20 forms cations when it isdissolved in aqueous acetic acid solution (AA) (pH < pKa1), whereasacidic drugs IB and CCM21,22 form anions when they are dissolved inaqueous KOH solution (pH > pKa). The ionized drug solution wassubsequently added to the oppositely charged aqueous PE solution,upon which drug−PE electrostatic interaction took place to form asoluble drug−PE complex as illustrated in Figure 1. At a certain criticalconcentration whose value depends on the drug hydrophobicity,drug−PE complexes aggregated by means of interdrug hydrophobicinteractions and finally formed the insoluble drug nanoplex in thepresence of salt. The role of the salt is to provide a charge-screeningeffect to minimize repulsions of the like-charged PE chains, which caninhibit complex aggregation. The strong drug−PE electrostaticinteraction prevented the drug molecules from assembling intoordered crystalline structures upon precipitation, resulting in theamorphous nanoplex.

The detailed procedures for the CIP nanoplex preparation havebeen presented in Cheow and Hadinoto17 and are not repeated herefor brevity. To prepare the IB and CCM nanoplexes, PAH wasdissolved in 0.0, 0.1, or 0.2% (v/v) aqueous AA to form a solution ofpolycations, whereas the drug was dissolved in a 0.1 M KOH aqueoussolution to form anionic drug molecules. Next, 1 mL of the drugsolution was added to equal volume of PAH solution, with a final salt(NaCl) concentration of 0.1 M. The mixture was left for 3 h underambient conditions to allow the complexation to equilibrate. Toremove the excess drug and PE (i.e., those that do not formnanoplexes) as well as NaCl and KOH, the nanoplex suspension waswashed with three cycles of centrifugation and resuspension in

Table 1. Relevant Properties of the Model APIs

IB CCM CIP

log P 5.2 2.5 1.32pKa 4.6 8.31, 10, 10.2 6.1, 8.6saturation solubility (mg/mL) 1.75 0.021 0.14

Langmuir Article

dx.doi.org/10.1021/la204782x | Langmuir 2012, 28, 6265−62756266

deionized water. The nanoplex suspension was then transformed intodry powder by freeze drying. On this note, because free CCM is proneto degradation in alkaline22 solution, the CCM solution was added tothe PAH solution immediately after its preparation to form the CCM−PAH complex.The ternary phase diagram of the drug−PE complex was

constructed by varying the mass compositions of the drug, PE, andsalt (Mdrug/PE/salt) at different AA concentrations while keeping thetotal salt concentration constant at 0.1 M. The complexation productswere separated into four categories: (1) a nanoplex, (2) precipitates,(3) a combination of the nanoplex and precipitates, and (4) a solublecomplex. To identify nanoplex suspensions, removal of the excess saltand resuspension in DI water were necessary because the charge-stabilized nanoplex was flocculated in the presence of salt. Tocategorize the complexation products, the samples were centrifugedfor 15 min at 14 000 rpm. The sediments obtained at the bottom ofthe centrifuge tubes were nonredispersible pellets in the case ofprecipitates but were redispersible suspensions in the case of thenanoplex. Drug−PE mixtures were categorized as soluble complexeswhen a clear solution was observed.The drug to PE charge ratios (Rdrug/PE) were calculated from the

MW and the number of charges per molecule of the drug and PE. Asshown in Figure 2, IB and CIP have one charge each when fullyionized in base and acid, respectively, whereas CCM has three chargeswhen fully ionized. For IB, the single −COO− group per IB molecule(MW 206.3 g/mol) results in a charge density of 4.8 × 10−6 molcharge/mg. The charge densities of CCM and CIP are similarlycalculated, yielding 8.1 × 10−6 and 3.0 × 10−6 mol charge/mg,respectively. For the PEs, there is one −NH3

+ group per PAHmonomer (MW 93.5 g/mol) and 24 −OSO3

− groups per DXTmolecule, resulting in charge densities of 10.7 × 10−6 and 4.8 × 10−6

mol charge/mg, respectively.2.3. Physical Characterizations of the Amorphous Drug

Nanoplex. The complexation efficiency (CE) is defined as the masspercentage of drug that forms a nanoplex relative to the initial amountof drug added. The amount of drug that forms the nanoplex was

calculated as the difference between the initial amount of drug added andthe amount of drug remaining in the supernatant after the firstcentrifugation. IB, CCM, and CIP concentrations in the supernatant weremeasured by UV−vis spectrophotometry (UV Mini-1240, Shimadzu,Japan) at absorbance wavelengths of 263, 463, and 291 nm, respectively.The nanoplex size and zeta potential were measured by photoncorrelation spectroscopy (PCS) using a Brookhaven 90Plus nanoparticlesize analyzer (Brookhaven Instruments Corporation, USA).

The production yield is defined as the total nanoplex massproduced relative to the total mass of drug and PE initially added. Thetotal nanoplex mass produced was determined by freeze drying analiquot of the nanoplex suspension. Drug loading, which is defined asthe percentage of drug making up the nanoplex, was determined bymeasuring the amount of drug released when a known amount ofnanoplex was completely dissolved in PBS. The CE, yield, and drug-loading characterizations were performed using at least two replicates.To verify size measurement by PCS, the freeze-dried nanoplex wassputter coated with platinum and imaged using scanning electronmicroscope (SEM) model JSM-6700F (JEOL, USA).

The stability of the nanoplex powders was examined after storage at25 °C and 55% relative humidity for 1 month. Powder X-raydiffraction (PXRD) patterns of the nanoplex powders were examinedimmediately after preparation and after one month, using a D8Advance X-ray diffractometer equipped with Cu Kα radiation (Bruker,Germany) from 10 to 60° (2θ) with a step size of 0.02°/s. The PXRDpattern of the nanoplex powders was compared with that of the nativeAPIs. Differential scanning calorimetry (DSC) analysis was performedusing SDT Q600 (TA Instruments, USA), where 5 mg of powder wasplaced in an alumina pan and heated from 30 to 400 °C at 10 °C/min.Infrared (IR) spectra were recorded using a Perkin-Elmer SpectrumOne FT-IR system with a spectral resolution of 4 cm−1 from 4000 to450 cm−1. The sample pellets for FT-IR analysis were prepared bypressing a mixture of 1 mg of nanoplex powder and 100 mg of groundKBr in a die at 10 tons for 1 min.

2.4. Generation and Quantification of Supersaturation. Thedrug-saturation solubility presented in Table 1 was determined byadding native drugs in excess (∼100 mg) to 20 mL of PBS. After 24 hof incubation in a shaking incubator at 37 °C, a 2 mL aliquot of theincubated sample was centrifuged and filtered through a 0.1 μm filter(Puradisc, Whatman, USA), after which the dissolved drug

Figure 1. Amorphous nanoplex formation via self-assembly to yielddrug−PE complexation.

Figure 2. Three model APIs used are ionized when dissolved in acid(CIP) or base (IB, CCM).

Langmuir Article

dx.doi.org/10.1021/la204782x | Langmuir 2012, 28, 6265−62756267

concentration was measured. Experiments with CIP and CCM wereperformed in opaque bottles because these compounds are prone tophotolytic degradation.23 For IB and CCM, the supersaturated drugsolution was generated by adding the nanoplex at a concentrationequal to 25 times the saturation solubility to 50 mL of PBS at 37 °Cunder constant stirring. At specified time points, 2 mL of PBS waswithdrawn and centrifuged, and the supernatant was filtered anddiluted. To be more specific, 0.5 mL of the 2 mL filtered drug solutionwas immediately diluted 10-fold with PBS to prevent drugprecipitation from the supersaturated solution. Afterward, the drugconcentration was measured to obtain the supersaturation profile as afunction of time. For comparison, the supersaturation profile fromdissolution in PBS containing 1 mg/mL HPMC, an establishedprecipitation inhibitor known to prolong and intensify the super-saturation level, was also characterized. The supersaturation profilesreported were obtained from a minimum of two replicates. For CIP,the supersaturation generation was examined at 10 times the saturationsolubility because at 25 times the dissolved drug precipitates almostinstantaneously, preventing the supersaturated drug solution frombeing properly quantified.

3. RESULTS

3.1. Ternary Phase Diagrams and Optimal Formula-tions. 3.1.1. IB Nanoplex. Ternary phase diagrams of theIB−PAH complexation at three different AA concentrations(i.e., 0.0, 0.1, and 0.2% (v/v)) of the PAH solution arepresented in Figure 3. The results indicate that in the absenceof AA approximately two-thirds of the MIB/PAH/salt investigatedproduces a soluble IB−PAH complex. The percentage ofsoluble complex decreases to half at AA = 0.1 and 0.2% (v/v).MIB/PAH/salt that produces the soluble complex is foundpredominantly in the triangular regions at the bottom left ofthe phase diagrams as indicated by the dashed lines in Figure 3.The triangular regions are demarcated at ∼20, 25, and 30%(w/w) salt for AA = 0.0, 0.1, and 0.2% (v/v), respectively,above which only the soluble complex is formed. Outside thetriangular regions, a minimum of 30% (w/w) PAH is requiredto produce the nanoplex below which a soluble complex orprecipitates are formed.The formulation that results in the highest yield (∼45%) and

CE (∼90%) is identified at 35% IB, 50% PAH, and 15% salt(w/w) for AA = 0.1% (v/v) (arrow in Figure 3B). The high CEsuggests that most of the IB is recovered in the form of thenanoplex; therefore, the low yield is caused by the presence ofexcess PAH in the feed (i.e., RIB/PAH < 1). However, decreasingthe PAH composition while simultaneously increasing the IBcomposition, in an attempt to improve the yield, results inprecipitate formation as indicated by the solid circles below theoptimal point.3.1.2. CCM Nanoplex. Ternary phase diagrams of the

CCM−PAH complexation at two different AA concentrations(i.e., 0.0 and 0.1% (v/v)) of the PAH solution are presented inFigure 4. Only a small fraction of the MCCM/PAH/salt investigated(<15%) result in nanoplex formation. Specifically, only 5 out of39 data points result in nanoplex formation at AA = 0.0% (v/v)(Figure 4A). The number is further reduced to three at AA =0.1% (v/v) (Figure 4B). The ternary phase diagram at AA =0.2% (v/v) is therefore not constructed because even fewernanoplexes are produced. Approximately 55 and 40% of theMCCM/PAH/salt investigated produce a soluble complex at AA =0.0 and 0.1% (v/v), respectively, which are lower than thepercentages observed in the IB−PAH complexation atcorresponding AA concentrations. Similar to the IB−PAHmixtures, the soluble complex is predominantly formed in the

bottom left triangular regions above 30 and 35% salt (w/w) forAA = 0.0 and 0.1% (v/v), respectively.The minimal amount of nanoplex produced is therefore not

caused by an increase in soluble complex formation. Instead, it

Figure 3. Ternary phase diagrams of the IB−PAH complex in thepresence of 0.1 M NaCl at (A) 0.2, (B) 0.1, and (C) 0.0% (v/v) AA.The arrow denotes the optimal formulation in terms of the CEand yield.

Langmuir Article

dx.doi.org/10.1021/la204782x | Langmuir 2012, 28, 6265−62756268

is due to an increase in the amount of precipitates formed.Approximately 35 and 55% of the MCCM/PAH/salt investigatedproduce precipitates at AA = 0.0 and 0.1% (v/v), respectively,which are significantly higher than the ∼13 and 20% observedfor the IB−PAH complexation at the same AA concentrations.Outside the triangular regions, a minimum of 60% (w/w) PAHis required to produce the nanoplex, below which precipitatesor a soluble complex is are produced, depending on the drugcomposition.The optimal formulation with yield ≈ 48% and CE ≈ 90% is

identified at 20% CCM, 60% PAH, and 20% salt (w/w) forAA = 0.1% (v/v) (arrow in Figure 4A). Similar to the IB nanoplex,the combination of low yield and high CE suggests that excessPAH is present in the feed (RCCM/PAH < 1). Decreasing the

PAH composition, however, results in a combination of thenanoplex and precipitates as indicated by the solid trianglesbelow the optimal point.

3.1.3. CIP Nanoplex. In contrast to the IB and CCM−PAHcomplexation, the CIP−DXT complexation at 0.1 M NaClproduces the nanoplex for all of the MCIP/DXT/salt valuesinvestigated at AA = 0.2% (v/v). Hence, the ternary phasediagram of the CIP−DXT mixture is not presented here butinstead is presented in Figure S1 of the SupportingInformation. Note that for the CIP−DXT complexation theAA concentration refers to that of the aqueous solution used todissolve the drug, not the solution used to dissolve the PE as inthe IB and CCM−PAH complexes. Compared to our earlierstudy,17 the range of RCIP/DXT investigated is extended in thepresent work to 0.05−3.0 from 0.4−1.3. For the optimalformulation, multiple MCIP/DXT/salt values can produce a highyield and CEs of around 80 and 95%, respectively: (i) 50:30:20,(ii) 20:10:70, and (iii) 40:20:40% (w/w). The optimalformulation is selected at 50% CIP, 30% DXT, and 20% salt(w/w) (arrow in Figure S1) because its calculated RCIP/DXT isthe closest to unity.

3.2. Physical Characteristics of the Amorphous DrugNanoplex. Physical characteristics of the optimal IB, CCM,and CIP nanoplex formulations are presented in Table 2, wherethe formulations are presented in terms of the drugconcentration and Rdrug/PE instead of Mdrug/PE/salt. The optimalRdrug/PE values for the IB, CCM, and CIP nanoplexes are foundto be at 0.7, 0.6, and 1.0, respectively. That the three optimalformulations have Rdrug/PE ≤ 1 is not a coincidence but rather aprerequisite to achieving high CE (∼90−96%). The less thanoptimal yield (50−80%) is not a concern because the high CEdenotes a high recovery of the valuable API in the finalproducts, hence minimal drug wastage. Furthermore, the excessPE in the feed that contributed to the low yield can berecovered and recycled.Attributed to the high CE, the drug loading is high at ∼83,

79, and 84% for the IB, CCM, and CIP nanoplexes,respectively, which are comparable to the loading exhibitedby amorphous nanoAPI prepared by conventional techniques.The nanoplex zeta potential reflects that of their correspondingPE values, suggesting the role of PE as a colloidal stabilizer ofthe nanoplex. The IB and CCM nanoplexes possess highlypositive zeta potentials of ∼42 and 58 mV, respectively, whichare attributed to the cationic PAH, whereas the CIP nanoplex isnegatively charged at ∼−47 mV owing to the anionic DXT.The nanoplex sizes measured by PCS are in the order of IB >

CCM > CIP, with all sizes <500 nm. The decrease in theparticle size from IB (∼480 nm) to CIP (∼320 nm) correlateswith the decrease in their log P values, where the size is largestfor IB as the most hydrophobic drug (i.e., highest log P). Thehigher hydrophobicity leads to intensified interdrug hydro-phobic interactions resulting in larger assemblies of the drug−PE complex being formed, which ultimately leads to a largernanoplex size. Figure 5 shows SEM images of the threenanoplexes after freeze drying, where nanoparticles withrelatively uniform sizes of around 400, 200, and 180 nm are

Figure 4. Ternary phase diagrams of the CCM−PAH complex in thepresence of 0.1 M NaCl at (A) 0.1 and (B) 0.0% (v/v) AA. The arrowdenotes the optimal formulation in terms of CE and yield.

Table 2. Physical Characteristics of the Optimal Nanoplex Formulations

nanoplex drug (mg/mL) Rdrug/PE CE (%) yield (%) drug loading (%) size (nm) zeta potential (mV)

IB 17 0.7 90 60 83 480 ± 60 +42CCM 9 0.6 90 48 79 400 ± 40 +58CIP 17 1.0 96 80 84 320 ± 40 −47

Langmuir Article

dx.doi.org/10.1021/la204782x | Langmuir 2012, 28, 6265−62756269

observed for the IB, CCM, and CIP nanoplexes, respectively. TheIB and CCM nanoplexes are spherical in shape and relativelyuniform in size (Figures 5A,B), whereas the CIP nanoplex is not asspherical and is more irregular in size (Figure 5C). The PCSmeasurement returns nanoplex sizes that are on average ∼100−150 nm larger than those observed in the SEM images, which canbe attributed to the PE swelling in an aqueous environmentcontributing to larger nanoplex hydrodynamic sizes reported.3.3. Stability of Amorphous Drug Nanoplex Powders.

PXRD patterns of the native APIs and their nanoplexequivalents are presented in Figure S2 of the SupportingInformation. In contrast to the sharp crystalline peaks observedin the PXRD patterns of the native APIs, the nanoplexequivalents, which are made up of ∼80−85% (w/w) drug,exhibit low, diffuse peaks denoting their highly amorphousstate. Importantly, the nanoplexes remain in the amorphousstate after 1 month of storage, signifying their on-shelf stability.In accordance with the PXRD results, DSC thermograms of thenanoplex similarly show characteristics of an amorphous form.The crystalline melting peaks present in the native APIthermograms (Figure 6A) are replaced by small step changesindicative of glass transition in the nanoplex thermograms

(Figure 6B). The melting points of the native APIs increase inthe order of IB < CCM < CIP at 74, 166, and 264 °C,respectively, whereas the glass transition temperatures (Tg) of thenanoplex increase in the same order, at 126, 132, and 230 °C,respectively. The nanoplex stability is evident in the absence ofrecrystallization peaks associated with amorphous-to-crystallinetransitions upon heating. The nanoplex stability can be attributedto the high Tg of DXT and PAH (>100 °C).17,24

3.4. FT-IR Spectroscopy Analysis. FT-IR spectroscopyanalysis is performed to confirm the presence of API in thenanoplex formulation and to identify the API functional groupsinvolved in the complexation. IR spectra of the IB, CCM, andPAH nanoplexes, with their corresponding native equivalentsand PEs, are presented in Figure 7.The presence of IB in the nanoplex is evident because the IR

spectra of both the IB nanoplex and native (Figure 7A) displaypeaks at 1618−1617 and 1638 cm−1 that are attributed to thearomatic CC double bond,25 an identifying functional groupin the IB molecule. The presence of IB in the nanoplex isfurther confirmed by the peaks at 2850−2960 cm−1, which arecharacteristics of the stretching and deformation of methylgroups25 present in both the IB nanoplex and native but not inthe PAH. With regard to the interacting IB functional group, alarge peak at 1722 cm−1 present in the IB native spectrum,which is attributed to the carbonyl stretching of the carboxylicacid group (−COOH),25 is conspicuously absent from the IBnanoplex spectrum. Instead, a new peak at 1551 cm−1 due tothe −COO− asymmetric stretching25 is present in the IBnanoplex spectrum. Taken together, the disappearance andemergence of the −COOH and −COO− peaks, respectively,point to the −COO− group of IB as the one interactingelectrostatically with the NH3

+ group of PAH to form the IB−PAH complex.For the CCM nanoplex, the phenolic C−CH groups in the

CCM molecule are selected as the identifying functional group

Figure 5. SEM images of (A) IB, (B) CCM, and (C) CIP nanoplexesafter freeze drying.

Figure 6. DSC thermograms of (A) native API crystals with meltingendotherms and (B) amorphous drug nanoplexes with small stepchanges demarking the glass-transition temperatures.

Langmuir Article

dx.doi.org/10.1021/la204782x | Langmuir 2012, 28, 6265−62756270

because the same group is not present in PAH. The in-planevibration of the phenolic C−CH groups contributes to thepeaks at 1282, 1027, and 808 cm−1 in the CCM native spec-trum (Figure 7B).26 The same peaks are shifted in the CCMnanoplex spectrum to 1276, 1027, and 816 cm−1, respec-tively, denoting the presence of CCM in the nanoplex.To identify the interacting CCM functional group, the 1627and 1603 cm−1 peaks in the CCM native spectrum, attributedto the stretching of the CC and CO bonds,26 arecontrasted with the peaks in the CCM nanoplex spectrum. The1627 cm−1 peak is red-shifted to 1618 cm−1, but the 1603 cm−1

peak is not detected in the CCM nanoplex spectrum.In addition, a strong peak at 1510 cm−1 due to the stretchingand in-plane bending of CO26 is noticeably absent in theCCM nanoplex spectrum. The disappearance of the 1603 and

1510 cm−1 peaks points to the CO group of CCM as the oneresponsible for complexation, where the ketone groupstransform into the anionic keto−enol form and interactelectrostatically with the NH3

+ groups of PAH.We have shown previously that the antimicrobial efficacy of

the CIP nanoplex is comparable to that of the native CIP, thusdemonstrating the presence of CIP in the nanoplex.17 Theinteracting CIP functional group, however, has not beenidentified. The pyridone group of the CIP molecule is uniqueto CIP and thus can be used as the identifying functional group.The peak corresponding to the CO stretching of the pyri-done group27 is shifted from 1618 cm−1 in the CIP native spec-trum to 1629 cm−1 in the CIP nanoplex spectrum (Figure 7C).The shift is attributed to the transformation of the zwitterionicform of CIP to its cationic form having the NH2

+ group. Theabsence of 1595 cm−1 corresponding to NH2

+ bending28 in theCIP nanoplex spectrum indicates that the NH2

+ group interactselectrostatically with the anionic SO3

− groups of DXT.3.5. Supersaturation Profiles. The in vitro supersatura-

tion profiles of the three nanoplexes in PBS with or withoutHPMC are presented in Figure 8. The supersaturation levelplotted on the y axis is defined as the ratio of the apparentsolubility of the amorphous nanoplex to the saturationsolubility of the native crystals. On this note, the drug isreleased only from the nanoplex in an aqueous salt solutionsuch as PBS, which contains 0.14 M NaCl, where too little saltresults in minimal drug release. The charge screening by the saltis believed to decomplex the drug−PE electrostatic interaction,resulting in the release of the drug.17

Figure 8A shows that transforming IB into the IB nanoplexresults in only marginal supersaturation generation, where theapparent solubility is only ∼1.4 higher than the saturationsolubility, with or without HPMC. Despite the low super-saturation level generated, the IB nanoplex dissolves rapidlywithin 5 min to reach the maximum apparent solubilityattributed to its nanoscale size and amorphous state. Themaximum supersaturation level is sustained for more than 2.5 hwith or without HPMC. The minimal impact of HPMC on thesupersaturation level indicates that the precipitation of thedissolved drug from the supersaturated solution is not the causeof the low supersaturation level.The CCM nanoplex, in the absence of HPMC, dissolves

almost instantaneously as visually observed by the instanta-neous color change of the dissolution medium upon nanoplexaddition, which is followed within seconds by a color reversalback to a clear solution as the dissolved CCM precipitates outof the supersaturated solution. The rapid dissolution andprecipitation prevent the supersaturation level from beingquantified. As a result, the recorded apparent solubility is equalto the saturation solubility (Figure 8B). In contrast, rapiddissolution and precipitation of the CCM nanoplex are nolonger observed in the presence of HPMC. The supersaturationlevel reaches the maximum point of ∼9 in 1 h before itgradually decreases to ∼2 in 5 h, providing ample time for drugabsorption. The slower dissolution rate to reach the maximumsupersaturation level is caused by the adsorption of HPMConto the surface of the dissolving nanoplex, whereas the slowerprecipitation of the dissolved drug is attributed to theadsorption of HPMC onto the nuclei surfaces.As a side note, CCM is known to undergo significant

hydrolytic degradation of up to 75% after 2.5 h of dissolution inPBS;29 therefore, various encapsulation methods have beeninvestigated to minimize CCM degradation.30 Although we do

Figure 7. IR spectra of (A) the IB nanoplex and native, (B) theCCM nanoplex and native, and (C) the CIP nanoplex and native andtheir PEs.

Langmuir Article

dx.doi.org/10.1021/la204782x | Langmuir 2012, 28, 6265−62756271

not specifically study CCM degradation, a reduction in theCCM degradation rate upon complexation with PAH is impliedfrom the present results. The rate of CCM degradation istypically quantified from the decrease in the CCM absorbanceat a single wavelength. The fact that a high CCM super-saturation level can be maintained over 5 h, as opposed to arapid decrease within 2.5 h if CCM is degraded, suggests thatCCM released from the nanoplex undergoes minimal or lessdegradation. It is postulated that after its release from thenanoplex, PAH electrostatically interacts with CCM in thedissolution medium, hence minimizing CCM degradation.Indeed, electrostatic interactions of CCM with cationic micellesin solution have been found to reduce the CCM degradationrate.30

For the CIP nanoplex, without HPMC, the maximumsupersaturation level (∼8) is achieved within 5 min, denotinga very rapid dissolution (Figure 8C), which is in agreementwith the observations for the IB and CCM nanoplexes. Notunexpectedly, the supersaturation level decreases rapidlyto ∼2 within 30 min because of the precipitation of the dis-solved drug. In the presence of HPMC, a slower dissolution rateis observed, where the maximum supersaturation level (∼5) isachieved only after 30 min. The supersaturation is maintained

at a level of ∼3 to 4 for a period of 90 min, which isconsiderably longer than 25 min without HPMC, henceproviding a longer time window for absorption.

4. DISCUSSION4.1. Elucidating the Roles of Salt, PE, and Drugs in

Nanoplex Formation. The use of drug−PE complexation fornanoscale drug delivery is not completely new. A number ofrecent studies have taken advantage of the drug−PE electro-static interaction for controlled drug release purposes, wherethe release is prolonged by encapsulating the API intonanoparticles made of PE as the matrix.31,32 The drug loadingfor such a nanoplex, however, is low (<25%) because asubstantial amount of PE is needed to modulate the drugrelease. The role of PE in the present work is thus vastlydifferent from that in the previous studies. Here, the role of PEis to ensure colloidal stability and stabilize the amorphous stateof the nanoplex. Furthermore, the nanoplex exhibits high drugloading (∼80−85%), hence the nanoplex is more aptlyregarded as a drug nanoparticle itself, rather than a drug-encapsulating nanoparticle.Because a majority of orally administered drugs are

amphiphilic and they can be dissolved in aqueous acid orbase33 such that they can be transformed into an amorphousnanoplex, the potential for the nanoplex to become thestandard formulation for a supersaturating drug delivery systemseems promising, particularly as boosted by its simple, greenpreparation method. Nevertheless, despite the success inpreparing an amorphous nanoplex with supersaturationgeneration capability, we have not fully dissected the exactroles of the salt, drug, and PE in nanoplex formation, which iscrucial to the nanoplexation method being developed further.The results of the IB−PAH and CCM−PAH complexations

indicate that the percentage of soluble complex formeddecreases with increasing AA concentration in the PAHsolution. Becauase PAH is a weak polyelectrolyte whose degreeof ionization is dependent on the solution pH (pKa ≈ 9),34

increasing the AA concentration in the PAH solution results ina lower pH of the final drug−PE mixture, leading to a higherdegree of ionization of PAH. As a result, there are morecomplexed drug molecules present on the PAH chains at lowerpH, which in turn leads to enhanced complex aggregationforming an insoluble drug−PE complex, which is driven byhydrophobic interactions of the complexed drug molecules.A lower percentage of the drug−PE complex formed thus remainsin solution with decreasing pH. On this note, the lower percentageof the CCM−PAH soluble complex formed compared to IB−PAHat the same AA concentrations (50 and 44 compared to 66 and50% for 0.0 and 0.1% (v/v) acetic acid) can be attributed to thestronger CCM−PAH electrostatic interaction as a result of thehigher charge density of CCM (i.e., 3 versus 1 charge/moleculefor CCM and IB, respectively).The higher charge density of CCM similarly contributes to

the higher percentage of precipitates formed, as compared tothat for IB. Precipitates are predominantly recovered from theregion having simultaneous low PE and high drug compositions(i.e., toward the bottom right corner of the phase diagrams)because of the extensive interdrug hydrophobic interactionsacross a limited number of PE chains. For the IB−PAHcomplex, the percentages of precipitates are 39, 42, and 53% atAA = 0.0, 0.1, and 0.2% (v/v), respectively. A similar trend isobserved for the CCM−PAH complex, though at higherpercentages (i.e., 72 and 87% at AA = 0.0 and 0.1% (v/v),

Figure 8. Supersaturation profiles of (A) IB, (B) CCM, and (C) CIPnanoplexes upon dissolution in PBS with or without HPMC.

Langmuir Article

dx.doi.org/10.1021/la204782x | Langmuir 2012, 28, 6265−62756272

respectively). The increased percentage of precipitates formedat higher AA concentrations is attributed to the increasedhydrophobicity and lower solubility of the drugs at lower pH.In contrast to the soluble complex and precipitates formed by

IB and CCM complexed with PAH, CIP complexation withDXT results only in nanoplex suspensions. Owing to its strongpolyelectrolyte nature, DXT molecules are fully charged underall conditions investigated. The full ionization of DXT underthe present operating condition is in contrast to that of PAH,where the addition of the basic drug solution results in thepartial ionization of PAH. Despite the successful nanoplexformation at all MCIP/DXT/salt values investigated, this by nomeans implies that the CIP nanoplex can be produced underany conditions. More importantly, it does not mean that all ofthe CIP nanoplexes produced have the same characteristics(e.g., CE and size).4.2. Significance of Rdrug/PE on CE and Yield. Although

successful nanoplex formation is governed by the pH-dependent ionization of the PE and Mdrug/PE/salt, the CE andyield are more or less governed by Rdrug/PE. In this regard,Rdrug/PE > 1 denotes the presence of more drug than PEavailable for complexation, resulting in low CE. However,Rdrug/PE < 1 denotes excess PE in the feed, which ensures highCE but with the trade-off of having a low yield. Therefore, theoptimal formulation in terms of both the CE and yield isgenerally found at Rdrug/PE values equal to or slightly lower thanunity, which is demonstrated in the present results.Figure 9 presents the CE and yield of all of the IB, CCM, and

CIP nanoplexes (precipitates are excluded) plotted against their

Rdrug/PE values. With the exception of the IB nanoplex preparedat low drug concentrations, where very few nanoplexes are re-covered, the CE is consistently high at ∼80−100% for Rdrug/PE < 1(Figure 9A). Increasing Rdrug/PE to values equal to 1 orgreater leads to a gradual decrease in CE due to insufficient PE

available for complexation. In contrast, the yield is shown toincrease with increasing Rdrug/PE for Rdrug/PE < 1 and reach amaximum at Rdrug/PE ≈ 1, above which the yield starts todecrease because of the lower CE. Hence, the optimal Rdrug/PEvalues for the IB, CCM, and CIP nanoplexes as expected lie atvalues equal to or less than unity. Importantly, the overlap inthe CE and yield variations as a function of Rdrug/PE across thethree different APIs suggests that the observed trends in CEand yield are applicable to all drug−PE pairs. The high yieldachievable at Rdrug/PE near unity implies an efficient nano-plexation process with a potential for scaling up, whichcurrently remains uninvestigated.

4.3. Supersaturation Characteristics of an AmorphousDrug Nanoplex. A comparison of the supersaturation profilesof the three nanoplexes obtained in the presence of HPMCindicates that the CCM nanoplex generates the best solubilityenhancement, as evidenced by its high supersaturation level(∼9) that is accompanied by a prolonged supersaturationperiod. In contrast, the IB nanoplex displays the least solubilityenhancement with a maximum supersaturation level of only 1.4,whereas the CIP nanoplex exhibits an intermediate super-saturation level of ∼5. The supersaturation level generatedtherefore follows the order of CCM > CIP > IB, coincidingwith the saturation solubility order of the native APIs that isCCM < CIP < IB. In other words, the lower the saturationsolubility of the API, the higher the solubility enhancementgained from its transformation into the nanoplex. Nevertheless,the higher apparent solubility generated by the nanoplex isobtained at the expense of a slower dissolution rate. Therefore,amorphous nanoplex formulation is most suitable for poorlysoluble APIs, whose therapeutic effectiveness requires them tobe maintained at high concentrations for a prolonged period.The supersaturation behavior of the nanoplex in an acidicenvironment mimicking that of the stomach will be investigatedin future studies.

5. CONCLUSIONSWe have demonstrated the feasibility of employing anamorphous drug nanoplex as a supersaturating drug deliverysystem to improve the bioavailability of poorly soluble drugs.Both anionic (i.e., IB and CCM) and cationic (i.e., CIP) drugmolecules can be transformed into the nanoplex using a simple,green process, which involves only mixing of the drug and thePE solution in the presence of salt under ambient conditions.The roles of the salt, PE, and drug in nanoplex formation havebeen elucidated by a close examination of the drug−PEcomplex phase diagrams. Specifically, the importance of (1) thedrug’s hydrophobicity and charge density and (2) the PE’sdegree of ionization has been discussed. The optimal nanoplexformulations exhibit high drug loading in the range of ∼80−85%that is attributed to the high CE achieved at Rdrug/PE ≤ 1.The nanoplex powders remain amorphous after 1 month ofstorage, denoting its stability attributed to the high Tg of thePE. The supersaturation level generated by the nanoplex in thepresence of HPMC depends on the saturation solubility of thenative API, where the lower the saturation solubility, the higherthe supersaturation level. For CCM as the least-soluble API,high supersaturation levels of between ∼3 and 9 times thesaturation solubility are generated by the CCM nanoplex over a5 h period. The future direction is to extend the amorphousnanoplexation to poorly soluble APIs (<0.1 mg/mL) that arenot freely soluble in either acid or base, as well as nanoplex sizecontrol. In addition, we are also looking for alternative cationic

Figure 9. Variations in the (A) CE and (B) yield of the nanoplex as afunction of Rdrug/PE.

Langmuir Article

dx.doi.org/10.1021/la204782x | Langmuir 2012, 28, 6265−62756273

PEs with higher pKa values for complexation with acidic drugssuch that complete ionization in basic solution can be achieved,hence reducing the amount of PE required to form nanoplexes,resulting in a higher yield.

■ ASSOCIATED CONTENT*S Supporting InformationTernary phase diagram of CIP−DXT complexation. PXRDpatterns of crystalline native drugs and the amorphousnanoplex. This material is available free of charge via theInternet at http://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Author*Tel: (65) 6514 8381. Fax: (65) 6794 7553. E-mail: kunnong@ntu.edu.sg.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSWe thank Xue Tong for her assistance with the PXRDmeasurement and gratefully acknowledge the funding from theMinistry of Education of Singapore AcRF Tier I Fund (grantno. RG 76/10).

■ REFERENCES(1) Serajuddin, A. T. M. Salt formation to improve drug solubility.Adv. Drug Delivery Rev. 2007, 59, 603−616.(2) Ando, H. Y.; Radebaugh, G. W. Property-based drug design andpreformulation. In Remington: The Science and Practice of Pharmacy,21st ed.; Troy, D. B., Ed.; Lippincott Williams & Wilkins: Philadelphia,2005.(3) Rabinow, B. E. Nanosuspensions in drug delivery. Nat. Rev. DrugDiscovery 2004, 3, 785−796.(4) Grant, D. J. W.; Brittain, H. G. Solubility of PharmaceuticalSolids. In Physical Characterization of Pharmaceutical Solids; Brittain, H.G., Ed.; Drugs and the Pharmaceutical Sciences; Marcel Dekker: NewYork, 1995; Vol. 70, pp 321−386.(5) Van Eerdenbrugh, B.; Vermant, J.; Martens, J. A.; Froyen, L.;Humbeeck, J. V.; Van den Mooter, G.; Augustijns, P. Solubilityincreases associated with crystalline drug nanoparticles: methodologiesand significance. Mol. Pharmaceutics 2010, 7, 1858−1870.(6) Shegokar, R.; Muller, R. H. Nanocrystals: industrially feasiblemultifunctional formulation technology for poorly soluble actives. Int.J. Pharmaceut. 2010, 399, 129−139.(7) Brouwers, J.; Brewster, M. E.; Augustijns, P. Supersaturating drugdelivery systems: the answer to solubility-limited oral bioavailability?J. Pharm. Sci. 2009, 98, 2549−2572.(8) Yang, W.; Johnston, K. P.; Williams Iii, R. O. Comparison ofbioavailability of amorphous versus crystalline itraconazole nano-particles via pulmonary administration in rats. Eur. J. Pharm. Biopharm.2010, 75, 33−41.(9) Tam, J. M.; McConville, J. T.; Williams, R. O.; Johnston, K. P.Amorphous cyclosporin nanodispersions for enhanced pulmonarydeposition and dissolution. J. Pharm. Sci. 2008, 97, 4915−4933.(10) Yamada, T.; Saito, N.; Imai, T.; Otagiri, M. Effect of grindingwith hydroxypropyl cellulose on the dissolution and particle size of apoorly water-soluble drug. Chem. Pharm. Bull. 1999, 47, 1311−1313.(11) Verreck, G.; Six, K.; Van den Mooter, G.; Baert, L.; Peeters, J.;Brewster, M. E. Characterization of solid dispersions of itraconazoleand hydroxypropylmethylcellulose prepared by melt extrusion - part I.Int. J. Pharm. 2003, 251, 165−174.(12) Matteucci, M.; Paguio, J.; Miller, M.; Williams Iii, R.; Johnston, K.Flocculated amorphous nanoparticles for highly supersaturatedsolutions. Pharm. Res. 2008, 25, 2477−2487.

(13) Sertsou, G.; Butler, J.; Hempenstall, J.; Rades, T. Solvent changeco-precipitation with hydroxypropyl methylcellulose phthalate toimprove dissolution characteristics of a poorly water-soluble drug.J. Pharm. Pharmacol. 2002, 54, 1041−1047.(14) Chen, J. F.; Zhang, J. Y.; Shen, Z. G.; Zhong, J.; Yun, J.Preparation and characterization of amorphous cefuroxime axetil drugnanoparticles with novel technology: high-gravity antisolvent precip-itation. Ind. Eng. Chem. Res. 2006, 45, 8723−8727.(15) Matteucci, M. E.; Hotze, M. A.; Johnston, K. P.; Williams, R. O.Drug nanoparticles by antisolvent precipitation: mixing energy versussurfactant stabilization. Langmuir 2006, 22, 8951−8959.(16) Dhumal, R. S.; Biradar, S. V.; Yamamura, S.; Paradkar, A. R.;York, P. Preparation of amorphous cefuroxime axetil nanoparticles bysonoprecipitation for enhancement of bioavailability. Eur. J. Pharm.Biopharm. 2008, 70, 109−115.(17) Cheow, W. S.; Hadinoto, K. Self-assembled amorphous drug-polyelectrolyte nanoparticle complex with enhanced dissolution rateand saturation solubility. J. Colloid Interface Sci. 2012, 367, 518−526.(18) Zahr, A. S.; de Villiers, M.; Pishko, M. V. Encapsulation of drugnanoparticles in self-assembled macromolecular nanoshells. Langmuir2005, 21, 403−410.(19) Takagi, T.; Ramachandran, C.; Bermejo, M.; Yamashita, S.; Yu,L. X.; Amidon, G. L. A provisional biopharmaceutical classification ofthe top 200 oral drug products in the United States, Great Britain,Spain, and Japan. Mol. Pharmaceutics 2006, 3, 631−643.(20) Lee, D. S.; Han, H. J.; Kim, K.; Park, W. B.; Cho, J. K.; Kim,J. H. Dissociation and complexation of fluoroquinolone analogues.J. Pharm. Biomed. Anal. 1994, 12, 157−164.(21) Avdeef, A.; Berger, C. M.; Brownell, C. pH-metric solubility. 2:Correlation between the acid-base titration and the saturation shake-flask solubility-pH methods. Pharm. Res. 2000, 17, 85−89.(22) Tonnesen, H. H.; Karlsen, J. Studies on curcumin andcurcuminoids 0.5. alkaline degradation of curcumin. Z. Lebensm.-Unters. Forsch. 1985, 180, 132−134.(23) Cheow, W. S.; Hadinoto, K. Factors affecting drugencapsulation and stability of lipid-polymer hybrid nanoparticles.Colloids Surf., B 2011, 85, 214−220.(24) Shao, L.; Lutkenhaus, J. L. Thermochemical properties of free-standing electrostatic layer-by-layer assemblies containing poly-(allylamine hydrochloride) and poly(acrylic acid). Soft Matter 2010,6, 3363−3369.(25) Kolhe, P.; Misra, E.; Kannan, R. M.; Kannan, S.; Lieh-Lai, M.Drug complexation, in vitro release and cellular entry of dendrimersand hyperbranched polymers. Int. J. Pharm. 2003, 259, 143−160.(26) Kolev, T. M.; Velcheva, E. A.; Stamboliyska, B. A.; Spiteller, M.DFT and experimental studies of the structure and vibrational spectraof curcumin. Int. J. Quantum Chem. 2005, 102, 1069−1079.(27) Turel, I.; Bukovec, P.; Quiros, M. Crystal structure ofciprofloxacin hexahydrate and its characterization. Int. J. Pharm.1997, 152, 59−65.(28) Bodo, E.; Ciavardini, A.; Giardini, A.; Paladini, A.; Piccirillo, S.;Rondino, F.; Scuderi, D. Infrared multiple photon dissociationspectroscopy of ciprofloxacin: investigation of the protonation site.Chem. Phys. 2012, DOI: 10.1016/j.chemphys.2011.05.018.(29) Sun, D.; Zhuang, X.; Xiang, X.; Liu, Y.; Zhang, S.; Liu, C.; Barnes,S.; Grizzle, W.; Miller, D.; Zhang, H.-G. A novel nanoparticle drugdelivery system: the anti-inflammatory activity of curcumin is enhancedwhen encapsulated in exosomes. Mol. Ther. 2010, 18, 1606−1614.(30) Leung, M. H. M.; Colangelo, H.; Kee, T. W. Encapsulation ofcurcumin in cationic micelles suppresses alkaline hydrolysis. Langmuir2008, 24, 5672−5675.(31) Ranjan, A.; Pothayee, N.; Seleem, M.; Jain, N.; Sriranganathan,N.; Riffle, J.; Kasimanickam, R. Drug delivery using novel nanoplexesagainst a Salmonella mouse infection model. J. Nanopart. Res. 2010, 12,905−914.(32) Ding, D.; Li, K.; Zhu, Z.; Pu, K.-Y.; Hu, Y.; Jiang, X.; Liu, B.Conjugated polyelectrolyte-cisplatin complex nanoparticles forsimultaneous in vivo imaging and drug tracking. Nanoscale 2011, 3,1997−2002.

Langmuir Article

dx.doi.org/10.1021/la204782x | Langmuir 2012, 28, 6265−62756274

(33) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.Experimental and computational approaches to estimate solubility andpermeability in drug discovery and development settings. Adv. DrugDelivery Rev. 1997, 23, 3−25.(34) Choi, J.; Rubner, M. F. Influence of the degree of ionization onweak polyelectrolyte multilayer assembly. Macromolecules 2004, 38,116−124.

Langmuir Article

dx.doi.org/10.1021/la204782x | Langmuir 2012, 28, 6265−62756275