va

ceion

widts)anccanclesfordenitude attaining submicron sizes. In this paper, results are presented with a panel

ibed as3]. Nanmolecles suc

based on their morphology: nanospheres, presenting a homogeneous, This paper will focus on the generation of nanoparticles by

Journal of Controlled Release 147 (2010) 304310

Contents lists available at ScienceDirect

Journal of Contro

j ourna l homepage: www.e lsev

NANOMEDICINEmatricial structure inwhich the drugs are uniformly dispersed [4], andnanocapsules, showing a typical core-shell structure. The formulationof nanoparticles involves certain challenges: size control anddistribution, morphology, the therapeutic goal of drug delivery,biocompatibility of the polymer and compatibility of the physico-chemical properties of the drug.

The methods used to produce nanoparticles can be divided intothree main groups [2,5]: (i) physicochemical methods e.g. thecreation of nanoparticles using preformed polymers and inducingtheir precipitation by emulsicationsolvent evaporation, diffusion

mechanical methods, that is to say, the third of the above-mentionedmethods. The originality of this study lies in its use of new technologydeveloped by Bchi, the so-called Nano Spray Dryer B-90. This givesrise to a fundamentally new concept in spray drying technology, thatof producing submicron particles from a solution.

A traditional spray dryer is generally used to transform liquidsubstances into powders rapidly and efciently. The speed of theprocess and the consequently short drying time enables the drying ofeven temperature-sensitive products without degradation [6,7]. Thisspray drying process is particularly used to improve productor reverse salting-out (ii) in situ chemicamacromolecules, giving rise for instance

Corresponding author. Tel.: +33(0)3 68 85 42 13; fE-mail addresses: xiang.li@etu.unistra.fr (X. Li), nant

arpagaus.c@buchi.com (C. Arpagaus), belleteix.f@buchi.thierry.vandamme@pharma.u-strasbg.fr (T.F. Vandamm

0168-3659/$ see front matter 2010 Elsevier B.V. Adoi:10.1016/j.jconrel.2010.07.113h as drugs are dissolved,attached to the externalinto two main groups

increase the absorption rate, improve bioavailability, enable targetdrug delivery for cancer therapy and intravenous delivery systemsand so forth [3].entrapped, encapsulated, or even adsorbed orinterface. Nanoparticles can be categorized1. Introduction

Nanoparticles are commonly descrranging in size from 10 nm to 1 m [1delivery are designed from macroassemblies, in which the active princip(e.g. from furosemide) are also shown. 2010 Elsevier B.V. All rights reserved.

solid colloidal particles,oparticles used for drugular and/or molecular

interfacial polycondensation reactions [4] (iii) mechanical meth-ods e.g. use of high-energy devices like high pressure homo-genizers, soniers [2], or high-energy wet mills [3]. The relativeinterest of using nanoparticulate rather than microparticulatesystems in pharmaceutical applications is in their potential tol synthesis methods ofto polymerization or

conservation incompound andcosmetics and fbeen used for ena matrix systemexhibiting a sphnature of theconditions such

ax: +33(0)3 68 85 43 06.on@unistra.fr (N. Anton),com (F. Belleteix),e).

ll rights reserved.ncapsulate nano-emulsions, or to formulate nano-crystals

Nano-emulsionNano-crystalsof ve representative polymeric wall materials (arabic gum, whey protein, polyvinyl alcohol, modiedstarch, and maltodextrin) and the potentials to eBchiNanoparticle

revolutionary new sprayerparticles by an order of magNanoparticles by spray drying using innoSpray Dryer B-90

Xiang Li a, Nicolas Anton a,, Cordin Arpagaus b, Fabria Universit de Strasbourg, Facult de Pharmacie, UMR CNRS 7199 Laboratoire de Concept74 route du Rhin BP 60024 F-67401 Illkirch Cedex, Franceb Bchi Labortechnik AG, Meierseggstr. 40, Postfach CH-9230 Flawil 1, Switzerland

a b s t r a c ta r t i c l e i n f o

Article history:Received 13 May 2010Accepted 17 July 2010Available online 24 July 2010

Keywords:Spray drying

Spray drying technology isslurries, pastes or even melmaterials industries to enhetc. However, spray dryingfor drug delivery (e.g. partidrying technology, notablytive new technology: The Bchi Nano

Belleteix b, Thierry F. Vandamme a

et Application de Molcules Bioactives, quipe de Pharmacie Biogalnique,

ely known and used to transform liquids (solutions, emulsions, suspension,into solid powders. Its main applications are found in the food, chemical ande ingredient conservation, particle properties, powder handling and storagealso be used for specic applications in the formulation of pharmaceuticalsfor pulmonary delivery). Bchi is a reference in the development of spraylaboratory scale devices. This study presents the Nano Spray Dryer B-90, aveloped by Bchi, use of which can lower the size of the produced dried

lled Release

i e r.com/ locate / jconre ldried solid form. With the development of activeemulsion encapsulation used in pharmaceutics,unctional food preparation, this method has alsocapsulation purposes. The powder thus generated isin the form of microparticles (i.e. microspheres [5]),erical or hollowed morphology depending on thewall material used and the operational dryingas the inlet temperature, solid concentration, gas

aqueous phase inducing the demixtion of the oil (through spinodaldecomposition) in the form of nanometric emulsion droplets.Surfactants immediately stabilized the nano-droplets formed.Vitamin E acetate was chosen as a model oily phase (0.09 g),mixed with the nonionic surfactant (0.06 g) and then brought intocontact with the aqueous phase (0.35 g), spontaneously generatingnano-emulsions. The surfactant oil weight ratio (SOR) was set at40%. 1 wt.% of wall material aqueous solution was separatelyprepared and mixed with the nano-emulsion, in order to obtain aratio of 1 :4 between the nano-emulsion and the wall material (drycompounds). The size distribution and polydispersity of nano-emulsions were assessed by dynamic light scattering using aMalvern Nano ZS instrument (Malvern, Orsay, France). TheHeliumNeon laser (4 mW) operated at 633 nm with the scatterangle xed at 173, and the temperature was maintained at 25C.The polydispersity index (noted PDI) appears merely as amathematical denition, accounting for the relative error betweencurve t and experimental values. The PDI discloses the quality ofthe dispersion, from values lower than 0.1 for suitable measure-

305X. Li et al. / Journal of Controlled Release 147 (2010) 304310

NANOMEDICINEow rate or feed rate. The powder samples are generally heteroge-neous and amorphous. Overall yields of spray drying on a laboratoryscale are limited, reaching around 50% to 70% [8].

The strength of the Bchi Nano Spray Dryer B-90 lies in itsvibration mesh spray technology, creating tiny droplets (beforeevaporation) in a size range of a smaller order of magnitude than inclassical spray dryers. This is a revolution in spray dryingtechnology, making it possible to produce powders in thesubmicron size range with very narrow distributions and highformulation yields [9].

Our aim in this study was to evaluate the potentials and limits ofthis new spray dryer. Firstly, by determining particle size, distribution,homogeneity, morphology, formulation yields and general aspects ofthe samples, the inuence of both the wall materials (of polymericnature) and the different experimental conditions were investigated.The second part of the study focused on the encapsulation of a modellipid dispersion (i.e. a nano-emulsion formulated by a low-energymethod [10]) to generate dry submicron particles. The nal part dealswith the use of the Nano Spray Dryer B-90 to dry pure activemolecules (solubilized in the aqueous or organic phase) in order toproduce nano-crystals. Sample characterization was performed byscanning electron microscopy (SEM).

2. Experimental section

2.1. Materials

The so-called wall materials used in this work refer to supportmaterials for the spray drying process. The following materials weresuccessively used: arabic gum (kindly provided by CNI ColloidesNaturels International, Rouen, France), whey protein (kindly providedby DAVISCO Foods international Inc., USA), polyvinyl alcohol (PVA,from Sigma, Saint-Louis, USA), and Cleargum CO 03psy226 andGlucidex IT 12psy226 (which are, respectively, modied starch andmaltodextrin with DE 12, kindly donated by Roquette Frres, France).Nonionic surfactants from BASF (Ludwigshafen, Germany) werekindly provided by Laserson (Etampes, France), and used as received.They are Cremophor ELP from BASF (announced by BASF as puriedand compatible with the parenteral administration, with HLB around1214), which are polyoxiethylated-35 castor oil. Furosemide, anactive and hydrophobic model molecule for the formulation of nano-crystals, was kindly provided by Solmag (Garbagnate, Italy). Vitamin Eacetate, sodium chloride, and acetone were purchased from Sigma,and ultrapure water was obtained using the MilliQpsy226 ltrationsystem (Millipore, Saint-Quentin-en-Yvelines, France).

2.2. Sample preparation

2.2.1. Wall material solutionsThe different wall material solutions were prepared in ultrapure

water a day prior to performing the experiments (to ensure suitablepolymer swelling), and at three different concentrations: 0.1 wt.%,1 wt.% and 10 wt.%.

2.2.2. Nano-emulsion formulationLipid nano-emulsions were chosen as lipid model suspensions,

with controllable size and polydispersity. The chosen size range ofnano-emulsions was a few orders of magnitude smaller than thesprayed droplets and dried particles, that is, smaller than 100 nm.Nano-emulsions were formulated according to the low-energyemulsication process published elsewhere [10]. Briey, the nano-emulsion droplets are generated by bringing into contacts twophases: (i) one is composed of oil plus the hydrophilic surfactanttotally miscible; and (ii) the second phase is the aqueous one whichcan be for instance pure water or buffer. Once these two phases are

mixed, the hydrophilic species are immediately solubilized by thements and good quality of the colloidal suspensions, to values closeto 1 for poor quality samples, which in concrete terms, either do notpresent droplets sizes in the colloidal range, or exhibit a very highpolydispersity. Measurements were performed in triplicate, beforeand after performing the spray drying process (ltered on 0.45 min this latter case).

2.2.3. Spray drying of pure low molecular weight materialsIn order to formulate nano-crystals, different compounds (hydro-

philic or not) were tested and spray dried to obtain nanoparticles ofpure active ingredients. Sodium chloride nano-crystals were formu-lated from aqueous solutions at 0.1 wt.% and 1 wt.% concentrations.Furosemide was spray dried at 1.25 wt.% concentration in pureacetone. In this case the Nano Spray Dryer B-90 apparatus wasconnected to a cooling unit, the Inert Loop B-295, for safe operation ofsolvents in a closed-mode conguration. Nitrogen was used as aninert gas to prevent an explosive gas mixture. The O2 concentration inthe closed loop was kept below 4 vol.%.

2.3. The new Bchi technology: the Nano Spray Dryer B-90

The Nano Spray Dryer B-90 is based on a new spray dryingconcept. A complete diagram of the apparatus is illustrated in Fig. 1.The drying gas enters the apparatus from the top, heating up to thesetting inlet temperature, then ows through the drying chamber,exiting the spray dryer at the bottom outlet, and is ne ltered beforeFig. 1. Schematic of the Nano Spray Dryer B-90.

306 X. Li et al. / Journal of Controlled Release 147 (2010) 304310

NANOMEDICINEleaving the instrument. The inlet temperature, Tin, and outlettemperature, Tout, are respectively measured just after the heatingand just before the ne ltering. The liquid sample was fed to thespray head by a pump. As illustrated in the gure, the generation ofdroplets was based on a piezoelectric driven actuator, vibrating a thin,perforated, and stainless steel membrane in a small spray cap. Themembrane (spray mesh) features an array of precise, micron-sizedholes (4.0, 5.5 or 7.0 m). The actuator is driven at around 60 kHz,causing the membrane to vibrate, ejecting millions of precisely sizeddroplets per second with a very narrow distribution. These extremelyne droplets are dried into solid particles which are collected byelectrostatic charging and are deected to the collecting electrode.Finally the resulting powder is collected using a rubber spatula. Theoperating conditions for the experiments were kept constant at:Tin=100C, feed rates range from 3 to 25 mL/h (depending on thesolution viscosity, composition, etc.), the drying gas ow rate=100 L/min, and the spray mesh used in this study was the membrane with4.0 m sized holes.

2.4. Scanning electron microscope (SEM) and size distribution

The size and structure of spray dried nanoparticles were evaluatedusing a scanning electron microscope Philips XL20 (University ofStrasbourg, Plateforme de Microscopie lectronique, Institut de Gnt-ique et de Biologie Molculaire et Cellulaire). The specimens werecoated with platinum/gold and examined at 20 kV. The particle sizedistributions were disclosed from the SEM micrographs using asoftware-assisted method that we developed, whereby only theparticle number and their diameter are collected from the rawimages (even for these SEM pictures on which the particles appearedas agglomerated, sticked to the carbon support and metalized).Even if particles form aggregates, our method, by isolating eachparticle from the rest and dening its surface area, allows individualidentication, thus disclosing the particle size distribution. A log-normal extrapolation was applied, with a probability density functionof the form f x; ; = explnx2 = 22= x

2p

(where and are the mean and standard deviation), thus allowing aquantitative comparison of the SEM pictures, and hence between thespray dried samples.

3. Results and discussion

3.1. Formulation with pure wall materials

This section deals with the impact of the formulation parameterson the resulting properties of the spray dried powders, i.e., theparticle size, polydispersity and morphology, which are determineddirectly from the raw SEM pictures. These so-called formulationparameters are (i) the nature of the wall materials, (ii) the preciselocation of powder collection on the cylindrical collecting electrodeand (iii) the concentration of the wall material solutions beforespray drying.

As mentioned in Section 2.1, a representative range of vedifferent wall materials were spray dried; arabic gum, wheyprotein, polyvinyl alcohol, modied starch and maltodextrin. Fig. 2reports SEM micrographs of spray dried powders made from 1 wt.% of wall material solution. The particle size distributions next tothe photographs were directly derived from the SEMpictures. Table 1 summarizes the quantitative values of the sizepeaks and SD, as well as the formulation yields. The inuence ofthe sample collection locations in the cylinder-shaped electrode isalso highlighted: samples collected from the bottom half of thecollecting cylinder (the lower 50% indicated in Fig. 1) are labeledsubscript 1 and those from the top half (the top 50% indicated inFig. 1) are labeled subscript 2. The reason for this comparison lies

in the basic law of electrostatic physics, which should inuencethe magnitude of the active electrostatic forces on the particles infunction of their size. This could result in the segregation ofpotential particle size along the cylinder length, and therefore invariations in particle size distribution along the cylinder electrode.

The spray dried samples were generally powdery and fragile.The SEM pictures appeared homogeneous, presenting spherical-shaped particles, regardless of the wall materials and collectinglocations. Furthermore, the peak maxima were clearly below 1 m.For modied starch, the peak was actually below 500 nm, which, fora spray drying technique is a noteworthy result. Along with theneness of the distribution, this is a very novel result for suchtechnology.

Besides these homogeneous global results, signicant differ-ences between the maxima and size distribution standarddeviations of the samples are highlighted by the quantitativeapproach reported in Table 1. The comparison demonstrates threemain points: (i) the values of the peak size and SD of the bottompart (for reference, this part should normally have smallerparticles and a narrower distribution than the top part) (ii)variation of the particle size between the top and the bottom part,and (iii) variation of the SD between the top and the bottom part.The samples of arabic gum (a1) and (a2) showed similar maximaof 565 and 581 nm respectively. The difference was more apparentin the peak widths, with a standard deviation of 287 nm for thebottom part and 363 nm for the top part of the collecting cylinder.This result indicates that bigger particles are more rapidly trappedby the high voltage collection system than smaller ones, resultingin a particle segregation effect over the cylinder length. Thisphenomenon is further intensied by the surface charge density ofthe macromolecules, as conrmed by the second example withmaltodextrin, a neutral sugar, in which no difference was notedbetween the two parts of the cylinder. Indeed, there was nosignicant difference between the two collection parts; neither forthe maxima (657 and 629 nm) nor for the SD values (459 and435 nm). Polyvinyl alcohol (c) showed a similar trend tomaltodextrin, with a very slight rise of both size maxima (from673 to 729 nm) and peak width (from 386 to 425 nm), and (d)show the results for modied starch, a clearly charged molecule[11]. Here the segregation effect between the two collection partswas signicantly enhanced. This result corroborates the proposedparticle separation idea, given that the maxima increased from457 to 617 nm and the SD from 289 to 396 nm.

The next series of experiments were designed to evaluate theeffect of the solid concentration in the sample solution on the spraydried particle size distributions. Representative cases were found forarabic gum and whey protein solutions at 0.1, 1 and 10 wt.%concentrations. The results are presented in Fig. 3 and include thequantitative values of the size maxima and SD. The formulation yieldsare reported in Table 2.

At 0.1 wt.%, outstandingly small particle sizes were achieved: thepeaks shifted to lower values of 353 and 421 nm with standarddeviations of 107 and 144 nm for arabic gum and whey proteinrespectively. When the sample concentration was increased to 1 wt.%,the size and polydispersity also increased (see Table 2). Furtherincrease of the concentration to 10 wt.% resulted in an enlargement ofthe peakwidth, while the peak location remained roughly unchanged.Nevertheless it resulted in severe changes to the general powdermorphology (see Fig. 3).

These experiments with pure wall materials illustrate thepromising potentials of the Bchi Nano Spray Dryer B-90. Thegeneration of spray dried particles is greatly facilitated, with highfabrication yields and very narrow size distributions, with peakswell below 1 m. Size and distribution depend on the nature of thewall materials, the powder collection location (also inuenced andenhanced by the applied material), and mainly on the sample

concentration.

307X. Li et al. / Journal of Controlled Release 147 (2010) 304310

NANOMEDICINE3.2. Encapsulation of nano-emulsions

Low-energy nano-emulsions, i.e. simply formed by spontaneousemulsication, were chosen as a model to investigate the potential ofthe Nano Spray Dryer B-90 for lipid encapsulation. The oil-in-waternano-emulsion formulation procedure was followed, as described in2.2.2 The nano-emulsion thus obtained was then mixed with 1 wt.%wall material aqueous solution (at the weight ratio of 1:4). In order to

Fig. 2. SEM micrographs of pure wall materials spray dried particles, the concentration forpolyvinyl alcohol, and (d) modied starch. The subscripts 1 indicates that the samples wererespective size distribution is reported for each one of the samples (see Table 1). The scale

Table 1Size distribution: values of maxima (curve peaks) and standard deviation of thesamples presented in Fig. 2. Formulation yields and representative experimental dataare also presented: dried initial mass (mini), collected mass (mcoll), inlet temperature(Tin) and outlet temperature (Tout).

Wall materials Peak(nm)

SD(nm)

Yield(%)

mini/mcoll(mg)

Tin/Tout(C)

(a1) Arabic gum (bottom) 565 287 g74.7 188.8/141.0 100/4358(a2) Arabic gum (top) 581 363(b1) Maltodextrin (bottom) 657 459 g43.0 541.0/232.8 100/4158(b2) Maltodextrin (top) 629 435(c1) Polyvinyl alcohol (bottom) 673 386 g80.9 357.2/288.9 100/4257(c2) Polyvinyl alcohol (top) 729 425(d1) Modied starch (bottom) 457 289 g94.5 229.6/216.9 100/4358(d2) Modied starch (top) 617 396facilitate encapsulation with the new spray drying process, the lipidnano-droplets were specically formulated, with a size well below100 nm (84.9 nm, PDI = 0.117). The nano-emulsion sizes weremeasured before and after spray drying following a redispersion in avolume of water equivalent to that removed (results presented inTable 3). The emulsion/wall material ratio was adjusted in order toobtain suitable powdery samples after spray drying. The SEM resultsare reported in Fig. 4, and the quantitative values of the size maximaand SD are summarized in Table 4.

According to the data obtained, the Nano Spray Dryer B-90 isobviously able to encapsulate emulsions smaller than 100 nm intoseparated (i.e. not aggregated) submicron solid particles. Inaddition, re-dissolving the spray dried powder in water resultedin a redispersion of the encapsulated nano-emulsions without anysignicant degradation and/or size increase. This result demon-strates the capacity of this spray drying process to preserve theintegrity of nano-structured systems, a fundamental prerequisitefor research in nano-medicine and nano-pharmaceutics. It is alsoworth noting the importance of the nature of the wall materialsused. For instance, when using arabic gum, the PDI dramaticallyincreased (from 0.117 to 0.511). This is attributed to the specicsurface active features of the arabic gum and its afnities with theoily components of the emulsion, which can induce dropletocculation and destruction.

all samples before spray drying was xed at 1%. (a) arabic gum, (b) maltodextrin, (c)collected from the bottom part of the cylinder, and subscripts 2, from the top part. Thebars are 10 m.

umes w

308 X. Li et al. / Journal of Controlled Release 147 (2010) 304310

NANOMEDICINEFig. 3. Inuence of the sample concentration on the powder size distribution. (a) arabic gconcentrations of the aqueous solution before spray drying: 0.1, 1 and 10 wt.%. The samplreported for each one of the samples (see Table 2). The scale bars are 10 m.Comparing the SEM pictures (Fig. 4, and Table 4) of encapsulatedlipid emulsion to those obtained with pure wall materials (Fig. 2,Table 1) under identical experimental conditions (i.e. WM concen-tration xed at 1 wt.% collected in the top part of the electrode), wenoted that encapsulated lipid nano-emulsions showed a slightincrease in peak size along with a slight decrease in peak width, thishowever shows a non-negligible impact of the presence of the oildroplets on the generation of the dried particles. With an emulsion/wall material ratio of 1:4, the composition of the particles wasestimated to contain 80% of hydrosoluble wall material, 12% of oil oractive lipophilic compound (vitamin E acetate in this study), and 8% ofnonionic surfactant. We can conclude that the presence of surfaceactive compounds such as lipid nano-droplets and nonionic surfac-

Table 2Size distribution: values of maxima (curve peaks) and standard deviation of thesamples presented in Fig. 3. Formulation yields and representative experimental dataare also presented: dried initial mass (mini), collected mass (mcoll), inlet temperature(Tin) and outlet temperature (Tout).

Wall materials Peak(nm)

SD(nm)

Yield(%)

mini/ mcoll(mg)

Tin/Tout(C)

(a1) 0.1 wt.% Arabicgum

353 107 70.2 60.3/42.3 100/4260

(a2) 1 wt.% Arabicgum

581 363 74.7 188.8/141.0 100/4358

(a3) 10 wt.% Arabicgum

549 545 54.0 365.4/197.3 100/4158

(b1) 0.1 wt.% Wheyprotein

421 144 50.8 54.5/27.7 100/4156

(b2) 1 wt.% Wheyprotein

593 374 76.0 416.1/316.0 100/3858

(b3) 10 wt.% Wheyprotein

537 618 89.4 1053.5/941.5 100/4760, and (b) whey protein. The subscripts 1, 2 and 3 respectively indicate the three differentere collected at the top part of the collecting cylinder. The respective size distribution istants do not have a signicant impact on the particle fabricationprocess. The powders obtained exhibited size distributions withmaxima still below the 1 m scale and with a relatively narrow sizedistribution (compared to the results in Tables 1 and 2).

3.3. Nano-crystals produced by spray drying

The fabrication of spray dried submicron particles made fromsolutions using pure active ingredients, i.e. without so-called wallmaterials (excipients), is a challenge and is still in abeyance. It ishowever of prime interest in pharmaceutic and cosmetic formula-tion. The main advantage of nano-sized API crystals for drugdelivery systems is the enhanced physicochemical properties ofdrug dispersion at the nanometric scale. Two model substanceswere tested in this study. The rst was simply hydrosoluble salt,NaCl, and the second was furosemide, a low molecular weight active

Table 3Nano-emulsion size distribution, before mixing with wall materials (WM), mixed withWM before spray drying, and after spray drying when the dried particles areresolubilized in the same water volume.

Wall materials Nano-emulsion+WM

Nano-emulsion+WM

Raw nano-emulsion

Before spraydrying

After spraydrying

dh (nm) PDI dh (nm) PDI dh (nm) PDI

Arabic gum g84:86 0:117 95.49 0.155 105.9 0.511Modied starch 91.49 0.229 161.5 0.243Maltodextrin 83.49 0.088 140.6 0.167Whey protein 88.54 0.261 110.3 0.268

) modied starch; (c) maltodextrin; and (d) whey protein. Wall material concentrations ofal (dry compounds) weight ratio xed at 1:4 (see details in the text). The respective sizem.

309X. Li et al. / Journal of Controlled Release 147 (2010) 304310

NANOMEDICINEmolecule with poor solubility in water, used as a diuretic drug totreat congestive heart failure and edema. The results are presentedin Fig. 5, and the values of the size maxima and SD are summarizedin Table 5.

Spray drying a pure species be it hydrophilic or lipophilic with a low molecular weight, results in homogeneous andpowdery samples. In the case of sodium chloride, similar resultsas in Fig. 3, were achieved; the size distribution was in keepingwith the sample concentration, and the narrow peaks shifted from517 to 993 nm by increasing the solid concentration from 0.1 to1 wt.%. As previously observed when increasing the WM concen-tration (Fig. 3 and Table 2), this is typically due to an increase ofthe substance amount within each droplet formed before drying.The furosemide powder size measured 1.24 m using a 1.25 wt.%concentration, which was still considered to be a noteworthyresult. The SEM photograph of furosemide (b) shows theformation of nanometric needle-like crystals (see insert inFig. 5, two of these crystals, indicated by the arrows, were

Fig. 4. Encapsulation of low-energy nano-emulsions by spray drying. (a) arabic gum; (bthe aqueous solution before spray drying: 1 wt.%; Nano-emulsion and the wall materidistribution is reported for each one of the samples (see Table 4). The scale bars are 10evaluated at a thickness of 66.9 and 152.0 nm) forming only partof the predominant sphere-shaped crystals constituting theparticles. This means that the evaporation of discrete dropletstotally controls furosemide crystallization, overriding the freeneedle-like crystallization of the product.

Here again, this novel spray drying technology shows signicantpotential in formulating nano-materials and this time without theapplication of wall materials. Avoiding the use of excipients is aconsiderable advantage in the vast domain of pharmaceutical formu-lation. In addition to diluting the samples, wall materials can induceadverse effects. The spray drying experiment described above couldconstitute an optimized and efcient method of API formulation.

Table 4Size distribution: values of maxima (curve peaks) and standard deviation of thesamples presented in Fig. 4.

Wall materials Peak (nm) SD (nm)

(a) Arabic gum 709 253(b) Modied starch 625 275(c) Maltodextrin 773 352(d) Whey protein 801 301

Fig. 5. Powder of pure low molecular weight molecules of (a) a model of water solublesalt (sodium chloride), and (b) of a model of poorly soluble drug (furosemide). (a1) and(a2) respectively refer to concentrations of 0.1 and 1 wt.%, and the concentration offurosemide in (b) is 1.25 wt.%. The star indicates the correspondence with an enlargeddetail in picture (b). The respective size distribution is reported for each one of thesamples (see Table 5). The scale bars are 10 m.

4. Conclusion

The Bchi Nano Spray Dryer B-90 appears to provide verysatisfactory results for the formulation of submicron particles, withrelatively high yields (70% to 90%) for small sample amounts (50 mgto 500 mg). Five representative polymeric wall materials (arabicgum, whey protein, polyvinyl alcohol, modied starch and mal-todextrin) were spray dried and the resulting size distributions wereshown to be mainly below the 1 m scale, attaining sizes as low as~350 nm with a standard deviation of ~100 nm for arabic gum(0.1 wt.% solid concentration), which is a very noteworthy result forspray drying technology. The size and standard deviation depend onthe nature of the wall material used, the collection location of thepowder samples on the collecting electrode (depending on thechemical structure and intrinsic molecule charge) and mainly on theconcentration of the spray dried solution. The preliminary results ofencapsulated nano-emulsions and formulated nano-crystals usingthis novel spray drying technology appear to be of extreme interestfor API formulation and offer promising perspectives for newpharmaceutical applications using spray drying.

Acknowledgments

We are very grateful to Bchi for making it possible to carry outthis study with the Nano Spray Dryer B-90. Many thanks for theirexcellent technical support.

References

[1] J. Kreuter, Encyclopedia of Pharmaceutical Technology, volume 10, MarcelDekker, New York, pp. 219342.

[2] N. Anton, J.P. Benoit, P. Saulnier, Design and production of nanoparticlesformulated from nano-emulsion templates a review, J. Control. Release 128(2008) 185199.

[3] R. Singh, J. Lillard, Nanoparticle-based targeted drug delivery, Exp. Mol. Pathol. 86(2009) 215223.

[4] C. Vauthier, K. Bouchemal, Methods for the preparation and manufacture ofpolymeric nanoparticles, Pharm. Res. 26 (2009) 1002511058.

[5] J. Richard, J.-P. Benoit, Microencapsulation, Techniques de l'ingnieur J 2 (210)(2000) 120.

[6] R. Vehring, Pharmaceutical particle engineering via spray drying, Pharm. Res. 25(2008) 9991021.

[7] P. Schuck, A. Dolivet, S. Mjean, P. Zhu, E. Blanchard, R. Jeantet, Drying bydesorption: a tool to determine spray drying parameters, J. Food Eng. 94 (2008)199204.

[8] C. Arpagaus, N. Schafroth, Laboratory scale spray drying of biodegradablepolymers, Respir. Drug Delivery Eur. (2009) 269274.

[9] Bchi Incorporation, Nano Spray Drier online brochure, http://www.buchi.com/Nano-Spray-Dryer-B-90.12378.0.html

[10] N. Anton, T. Vandamme, The universality of low-energy nano-emulsication, Int. J.Pharm. 377 (2009) 142147.

[11] C. Preetz, A. Rbe, I. Reiche, G. Hause, K. Mder, Preparation and characterizationof biocompatible oil-loaded polyelectrolyte nanocapsules, Nanomedicine 4(2008) 106114.

Table 5Size distribution: values of maxima (curve peaks) and standard deviation of thesamples presented in Fig. 5.

Materials Peak (nm) SD (nm) Yield (%)

(a1) 0.1 wt.% Sodium chloride 517 182 81.1(a2) 1 wt.% Sodium chloride 993 256 85.4(b) 1.25 wt.% Furosemide 1245 482 69.3

310 X. Li et al. / Journal of Controlled Release 147 (2010) 304310

NANOMEDICINE

Nanoparticles by spray drying using innovative new technology: The Bchi Nano Spray Dryer B-90IntroductionExperimental sectionMaterialsSample preparationWall material solutionsNano-emulsion formulationSpray drying of pure low molecular weight materials

The new Bchi technology: the Nano Spray Dryer B-90Scanning electron microscope (SEM) and size distribution

Results and discussionFormulation with pure wall materialsEncapsulation of nano-emulsionsNano-crystals produced by spray drying

ConclusionAcknowledgmentsReferences