deposition of antimicrobial coatings on ... - Springer Link

structural, Functional, and Biological thin Films

Overview

deposition of antimicrobial coatings on microstereolithography-fabricated microneedles Shaun D. Gittard, Philip R. Miller, Chunming Jin, Timothy N. Martin, Ryan D. Boehm, Bret J. Chisholm, Shane J. Stafslien, Justin W. Daniels, Nicholas Cilz, Nancy A. Monteiro-Riviere, Adnan Nasir, and Roger J. Narayan Microneedles are small-scale needle-like projections that may be used for transdermal delivery of pharmacologic agents, including protein-containing and nucleic acid-containing agents. Commercial translation of polymeric microneedles would benefit from the use of facile and cost effective fabrication methods. In this study, visible light dynamic mask microstereolithography, a rapid prototyping technique that utilizes digital light projection for selective polymerization of a liquid resin, was used for fabrication of solid microneedle array structures out of an acrylate-based polymer. Pulsed laser deposition was used to deposit silver and zinc oxide coatings on the surfaces of the visible light dynamic mask microstereolithography-fabricated microneedle array structures. Agar diffusion studies were used to demonstrate the antimicrobial activity of the coated microneedle array structures. This study indicates that light-based technologies, including visible light dynamic mask microstereolithography and pulsed laser deposition, may be used to fabricate microneedles with antimicrobial properties for treatment of local skin infections. introduction Skin infections, including abscess, cellulitis, erysipelas, and impetigo, are an issue of growing medical importance; for example, skin infections are associated with 7–10% of hospitalizations.1,2 Gram-positive cocci bacteria (e.g., Staphylococcus aureus and group A streptococci) are most commonly associated with infection of normal skin. A 2001 study indicated that the microorganisms most frequently associated with skin infections were (arranged by Vol. 63 No. 6 • JOM

How would you… …describe the overall significance of this paper? In this manuscript visible light dynamic mask microstereolithography was shown to be an appropriate technique for scalable production of microneedle devices. The combination of dynamic mask stereolithography and pulsed laser deposition was used to create acrylate-based polymer microneedle arrays with antimicrobial properties, which may be used for treatment of local skin infections. …describe this work to a materials science and engineering professional with no experience in your technical specialty? Antimicrobial silver and zinc oxide thin films were deposited on visible light dynamic mask microstereolithography-fabricated microneedles by means of pulsed laser deposition. The results of agar diffusion assays showed that the silver thin films and zinc oxide thin films provided the microneedle devices with antibacterial properties. Significant effects against Staphylococcus epidermidis and Staphylococcus aures, two common pathogens, were noted. …describe this work to a layperson? Microneedles devices were fabricated in a layer-by-layer manner from computer models using a rapid prototyping process known as visible light dynamic mask microstereolithography. Antimicrobial silver and zinc oxide thin films were deposited on the visible light dynamic mask microstereolithography-fabricated microneedles using a physical vapor deposition process called pulse laser deposition. These coated microneedle devices demonstrated significant activity against two bacteria associated with skin infections, Staphylococcus epidermidis and Staphylococcus aures.

www.tms.org/jom.html

decreasing frequency) Staphylococcus aureus, Enterococcus species, coagulase-negative staphylococci, Escherichia coli, and Pseudomonas aeruginosa.3 Empiric antibiotic therapy for staphylococci and streptococci is the most common approach for skin infection treatment; however, topical use of zinc oxide and silver may also find use in skin infection treatment.4 There has been significant interest in recent decades in the medical use of silver.5–7 As noted by Nadworny and Burrell, silver nitrate was introduced for burn treatment in 1965.8 In 1967, silver sulfadiazine cream was introduced for burn and wound treatment. Ag+ ions interact with sulfur-containing proteins (e.g., proteins in bacterial membranes) and phosphorus-containing compounds (e.g., DNA).5 Ag+ ions can also cause cell death by interacting with the respiratory chain within bacteria. Schaller et al. showed that silver-containing Contreet-H® wound dressings released silver and provided activity against methicillin-resistant Staphylococcus aureus and Candida albicans in an in vitro cutaneous infection model that contained reconstituted human epithelium; furthermore, no major morphological effects were noted in treated epithelial cells.9 More recently, Bhattacharyya and Bradley demonstrated the use of nanocrystalline silver (Acticoat 7®) for decreasing the spread of cutaneous necrosis and reducing methicillin-resistant Staphylococcus aureus loading in a clinical activity involving two patients.10 They suggested use of silver dressings as a “preemptive antimicrobial therapy” for management of difficult-to-treat and high-risk acute skin infections. Silvercontaining ointments and moisturizers 59

ExpErimEntal procEdurE The first hole resolution structure, shown in Figure Ba, is an arA Perfactory III SXGA+system (EnvisionTEC GmbH, Gladray of circular holes having five replicates in the y-direction and beck, Germany) was used to produce microneedle devices from in the x-direction; processing of 800 µm, 700 µm, 600 µm, 550 a commercially available photosensitive acrylate polymer, eSµm, 500 µm, 450 µm, and 400 µm diameter holes was attempted. hell 200 (Envisiontec GmbH, Ferndale, MI), which is used in The second hole resolution structure, shown in Figure Bb, is an fabrication of thin-walled hearing aid shells.58 According to the array of square holes having 5 replicates in the y-direction and in manufacturer, eShell 200 is a Class-IIa biocompatible, waterthe x-direction; processing of 800 µm, 700 µm, 600 µm, 550 µm, resistant material. It contains 0.5–1.5% wt. phenylbis(2,4,6 500 µm, 450 µm, and 400 µm hole side-lengths was attempted. A trimethylbenzoyl)-phosphine oxide photoinitiator, 15–30% wt. third structure was produced to determine the line resolution of the propylated (2) neopentyl glycoldiacrylate, and 60–80% wt. urevisible light dynamic mask micro-stereolithography system; this thane dimethacrylate. This acrylate-based polymer a exhibits a structure is shown in Figure Bc. This structure consisted of lines glass transition temperature of 109ºC (E1545-00 test method), with input widths ranging from 140 µm to 30 µm in 10 µm increa flexural strength of 103 MPa (D790M test method), a tensile ments; in this structure, gaps of 500, 300, 250, 200, and 150 µm strength of 57.8 MPa (D638M test method), and an elongation were designed. After processing, the test structures were imaged at yield of 3.2% (D638M test method); in addition, it exhibits a with an EZ4 optical microscope (Leica Microsystems, Wetzlar, water absorption value of 0.12% (D570-98 test method). Previous Germany) using LAS EZ software (Leica Microsystems, Wetzlar, nanoindentation studies provided hardness and Young’s modulus Germany). values of 93.8 ± 7.25 MPa and 3050 ± 90 MPa, respectively.59 Microneedle arrays with four different geometries were proThe Perfactory III SXGA+ system (EnvisionTEC GmbH, duced using the visible light dynamic mask micro-stereolithograGladbeck, Germany) equipped with a Digital Micromirror Device phy system. The heights and base geometries of these microneedle SXGA+ (1280 × 1024-pixel resolution) guidance chip (Texas Indevices are shown in Table A. The microneedle devices exhibited struments, Dallas, TX) and a halogen bulb was used for layerrectangular pyramid shapes. Two different input pyramid heights, by-layer polymerization of the acrylate-based polymer. Structures with heights of 1,000 µm and 1,250 µm, and two different base were fabricated within a 96.54 mm × 72.41 mm build envelope. geometries, with dimensions of 500 µm × 250 µm and 750 µm × Microneedle devices were fabricated at 550 mW using an expo250 µm, were produced. The tips of all four microneedle device sure time of 3.5 seconds and a z-direction step size of 30 µm. geometries exhibited dimensions of 90 µm × 30 µm. Four identical Perfactory® RP software (Envisiontec GmbH, Ferndale, MI) was 3 × 3 microneedle arrays were attached by support structures to a used to specify the layout of the microneedle devices in the build 1.0 mm thick and 18 × 18 mm substrate (Figure Ac).The STL dearea. An image containing several input STL files is presented in signs for the microneedle devices were created using Solidworks Figure Aa. Structures were produced on the build platform acsoftware (Dassualt Systemes S.A., Velizy, France). The support cording to the layout that had been specified using Perfactory® structures were produced using Magics RP 13 software (MateriRP software. After fabrication, the build platform containing atalise NV, Leuven, Belgium). Eight substrates, each containing four tached microneedle devices was removed from the resin-filled microneedle devices, were produced in a single batch. Microneebasin. An image of microneedle devices attached to the build dles with heights of 1,000 µm and base dimensions of 750 µm × platform, which corresponds to the layout shown in Figure Aa, 250 µm were also produced on a cross-shaped substrate. is provided in Figure Ab. The microneedle devices were subseSince porcine skin has similar morphology and thickness to huquently removed from the build platform; support structures were man skin, it is an acceptable model for human skin. A skin penetraalso removed. The microneedle devices were then rinsed in isotion study was performed with full-thickness weanling Yorkshire propanol (Fisher Scientific, Waltham, MA) and rinsed in acetone skin. The ability of the microneedle devices to penetrate porcine (Fisher Scientific, Waltham, MA). The microneedle devices were skin was examined using trypan blue (Mediatech, Inc., Manassas, subsequently placed in an Otoflash Post Curing System instruVA), a toluidine-based dye. The skin was stored at 3ºC until the ment (EnvisionTEC GmbH, Gladbeck, Germany) for 50 seconds; skin penetration study was performed. A cross-shaped microneethis instrument provides light exposure over a 300–700 nm wavedle device was inserted into the porcine skin. Immediately after length range for post-build curing. microneedle device removal, trypan blue was applied to the miFive different types of structures were produced using the viscroneedle insertion site. Residual dye was subsequently removed ible light dynamic mask micro-stereolithography system in orusing isopropanol swabs. The microneedle devices were examined der to obtain resolution data. Two different test structures were before and after porcine skin insertion using a Leica EZ4 optical produced to determine the hole resolution, which we will define microscope (Leica Microsystems, Wetzlar, Germany). Optical mias resolution of a hole produced in a 2 mm thick flat substrate.

a b c Figure A. (a) Layout of input STL files, which were used for fabrication of microneedle array structures by means of visible light dynamic mask micro-stereolithography. (b) Optical image of the microneedle array structures on the build platform. (c) Drawing of the input STL file of a microneedle array structure; this structure consists of four microneedle arrays that are attached to the substrate.

60


JOM • June 2011

Figure B. Drawings of input files that were used for fabrication of (a) circular holes, (b) square holes, and (c) line patterns. Optical images of (d) circular holes, (e) square holes, and (f) line patterns, which were fabricated by means of the visible light dynamic mask micro-stereolithography system.

a

b

d

e

c

f

croscopy was also used to examine the microneedle-fabricated, trypan blue-treated pores in skin. Pulsed laser deposition of silver thin films and zinc oxide thin films was performed using a KrF excimer laser, which was operated at a wavelength of 248 nm, a pulse duration of 25 ns, and a repetition rate of 10 Hz. The coatings were deposited at room temperature and under 10–5 Torr vacuum for five minutes. For deposition of silver thin films, a 99.99% purity target was obtained from a commercial source (Alfa Aesar, Ward Hill, MA). For deposition of zinc oxide thin films, a 99.99% purity ZnO powder (Alfa Aesar, Ward Hill, MA) was pressed into round 2 inch diameter pellets and subsequently sintered at 900ºC in an oxygen atmosphere for 12 hours. Eight 250 × 500 µm base, 1,250 µm height microneedle devices were coated with each of the two coating materials. Silver and zinc oxide thin films were also deposited on masked silicon wafers for profilometry studies. A Hitachi S-3200 scanning electron microscope (Hitachi, Tokyo, Japan) equipped with a Robinson backscattered electron detector was used to image the microneedle devices. The pulsed laser deposition-coated microneedle devices were imaged without further modification; the uncoated microneedle devices were sputter coated 60% gold-40% palladium using a Technics Hummer II instrument (Anatech, Battle Creek, MI) in advance of imaging. A Dimension 3000 AFM with NanoScope analysis software (Veeco, Santa Barbara, CA) and an AC160 tip (Olympus, Melville, NY) operating in tapping mode was used to examine the surface topographies of pulsed laser deposition-grown silver and zinc oxide coatings on silicon substrates. A Tencor Alpha Step 200 stylus profilometer (PLA-Tencor, Milpitas, CA) was used to measure the thicknesses of the pulsed laser deposition-grown silver and zinc oxide coatings on silicon substrates. For each coated substrate, scans were performed in triplicate over a 400 µm range, which contained both coated and uncoated surfaces. A 20 mg stylus scanning at 10 µm/second was used for profilometry measurements. The antimicrobial activity of the pulsed laser deposition-coated microneedle devices was assessed using an agar diffusion assay. Escherichia coli ATCC 12435, Staphylococcus aureus ATCC 6548, and Staphylococcus epidermidis ATCC 35984 (American Type Culture Collection, Manassas, VA) were cultured overnight in tryptic soy broth (VWR International, West Chester, PA). The cell suspensions were then centrifuged at 4,500 rpm for 10 minutes. A cell suspension with a density of ~108 CFU/ml was then made by resuspending the pellet in 1x phosphate-buffered saline (VWR International, West Chester, PA). Mueller Hinton agar plates (VWR International, West Chester, PA) were inoculated with a lawn of bacteria using a sterile swab. The microneedle devices were placed on the inoculated plates with the coated side facing the agar; in this arrangement, the microneedle devices projected into the agar. After incubating at 37ºC for 24 hours, the surfaces of the plates were imaged with optical microscopy to evaluate microorganismmaterial interactions.

Table A. Input Measurements and Scanning Electron Microscopy Measurements of Microneedle Dimensions Input Height (mm)

Actual Height (mm)

Input Width (mm)

Actual Width (mm)

Input Depth (mm)

Actual Depth (mm)

A

1,000

828 ± 19

750

805 ± 24

250

311 ± 17

B C D

1,000 1,250 1,250

665 ± 40 831 ± 9 748 ± 31

500 750 500

510 ± 9 817 ± 6 571 ± 25

250 250 250

285 ± 15 332 ± 21 347 ± 11

Geometry

Vol. 63 No. 6 • JOM


have also been developed for treatment of eczema.11 In addition to antimicrobial activity, Jun et al. demonstrated that silver nanoparticles are associated with altered fibrogenic cytokine (e.g., VEGF, IL-10, and IFN-γ) levels as well as reduced wound inflammation.12 Liu et al. examined topical application of silver nanoparticles in a murine full-thickness excisional wound model in mice; they showed that silver nanoparticles increase the wound closure rate by promoting migration and proliferation of keratinocytes into the wound bed.13 In addition, silver nanoparticles promote differentiation of fibroblasts into myofibroblasts. Wright et al. examined the use of nanocrystalline silver-coated dressings in a porcine full-thickness wound model. Use of nanocrystalline silver-coated dressings was associated with rapid wound healing and reduced local matrix metalloproteinase levels; matrix metalloproteinases are believed to retard healing of chronic ulcers and other wounds.14 Elston noted that nanocrystalline silver is associated with little evidence of toxicity.15 Lansdown et al. noted that zinccontaining materials may be useful in wound care due to their activity against microorganisms that are commonly found in wounds.16 They stated that zinc inhibits growth of Gram-positive microorganisms and Gram-negative microorganisms; however, Gram-positive bacteria were noted to exhibit greater sensitivity to zinc than Gram-negative bacteria. Atmaca et al. examined zinc interaction with Staphylococcus aureus, Staphylococcus epidermidis and Pseudomonas aeruginosa membranes; their work indicated that zinc prolongs the lag phase of the growth cycle.17 Sawai et al. noted that intracellular damage may be associated with production of hydrogen peroxide by zinc oxide.18 Zinc oxide may also directly damage the bacterial cell membrane, leading to intracellular content leakage and cell death.19 Disruption of Grampositive bacterial membranes and Gram-negative bacterial membranes has been noted after zinc oxide exposure.20 Agren et al. and Akiyama et al. described zinc oxide-mediated inhibition of Staphylococcus aureus growth and 61

attachment, respectively.21,22 Soderberg et al. obtained minimum inhibitory concentration data for human wound infection-derived bacteria that were exposed to Zn+2.23 Different susceptibility rates were noted for Enterococcus sp., Proteus sp., Pseudomonas aeruginosa, (MICs of 8–32 mmol/L); Enterobacter sp., Escherichia coli, Klebsiella sp. (MICs of 4–8 mmol/L); Staphylococcus aureus, Streptococcus group B (MICs of 2–4 mmol/L); and Streptococcus groups A, C, and G (MICs of 0.5–2 mmol/L). Lansdown et al. also noted that topical application of zinc may increase migration of keratinocytes, breakdown of collagen, and removal of necrotic tissue.16 Although topical application of zinc oxide may lead to high (1,000–3,000 mmol/L) local zinc levels, these zinc levels may be associated with nontoxic and antioxidant properties.16,24–26 For example, Agren et al. showed that topical administration of zinc oxide was associated with increased levels of zinc (1,540 µM) in wound fluid as well as lower rates of Staphylococcus aureus within wounds.26 Zinc oxide-containing paste bandages, known as Unna boots, are used to protect inflamed peri-ulcer skin in leg ulcers and reduce inflammation in lower extremity eczema with venous stasis; in these bandages, cotton gauze is saturated with zinc oxide paste. Elston suggested topical use of zinc compounds for acne treatment due to their anti-inflammatory and bacteriostatic properties.15 Microneedles are miniaturized lancet-, thorn-, or hypodermic needle-shaped devices that are used to penetrate the keratinized stratum corneum layer of the epidermis; this layer blocks transport of pharmacologic agents through the skin. Microneedles are sharp projections exhibiting at least one lateral dimension that is less than 500 mm.27,28 Microneedles are associated with low levels of patient pain since these devices do not penetrate deep into the dermis, where Meissner’s corpuscles, Pacinian corpuscles, and large nerve endings are present.29–32 These devices facilitate transdermal delivery of pharmacologic agents, including protein-containing and nucleic acid-containing agents, by providing conduits through the stratum corne62

um layer of the skin.28 Donnelly et al. used an in vitro model to examine the movement of microorganisms through microneedle-fabricated pores.33 They showed that movement of Candida albicans, Pseudomonas aeruginosa and Staphylococcus epidermidis across Silescol® membranes was an order of magnitude lower for microneedle-fabricated pores than for 21G hypodermic needle-fabricated pores. On the other hand, adherence of microorganisms to microneedles was one order of magnitude higher than adherence of microorganisms to hypodermic needles. No microorganisms traversed viable porcine epidermis that had been punctured with microneedles. Several investigators have recently considered the use of microneedles for treatment of skin conditions. For example, Doddaballapur described the use of microneedles for treatment of acne scars, stretch marks, scars, and wrinkles.34 Majid described a clinical study involving use of microneedles for treatment of atrophic facial scars. 94% of microneedle-treated patients obtained a reduction in scarring severity of one or two grades.35 More recently, Chandrashekar and Nandini described the combined use of microneedles and subcision for depressed acne scars.36 Lane described transdermal delivery of antimicrobial agents, including antibacterial and antifungal agents, by means of microneedles.37 He discussed systemic delivery of antimicrobial agents, including agents that cannot be orally administered (e.g., vancomycin), in a manner that exceeds the minimum inhibitory concentration or minimum microbicidal concentration. Furthermore, the emergence of drug resistant microorganisms would be reduced and the course of therapy would be shortened. In recent work, fabrication of microneedles containing antimicrobial agents has been demonstrated. In one study, a master structure of a microneedle array was produced by two photon polymerization.38 Microneedles were subsequently prepared out of an organically modified ceramic hybrid sol–gel (Ormocer®) material using a polydimethylsiloxane mold. These two photon polymerization/micromoldingfabricated microneedle arrays were www.tms.org/jom.html

coated with silver using pulsed laser deposition. The viability of human epidermal keratinocytes on silver-coated Ormocer surfaces was similar to that on uncoated Ormocer surfaces. The silver-coated Ormocer microneedles were shown to be effective at inhibiting growth of Staphylococcus aureus in an agar diffusion assay. In another study, two photon polymerizationmicromolding was used to produce microneedle arrays out of a photosensitive material that contained polyethylene glycol 600 diacrylate and 2 mg/ mL gentamicin sulfate.39 The polyethylene glycol 600 diacrylate- gentamicin sulfate microneedles were shown to inhibit Staphylococcus aureus growth in an agar plating assay. In this study, microneedles were fabricated by means of a rapid prototyping process known as visible light dynamic mask microstereolithography. In rapid prototyping, structures are built up in a layer-by-layer manner from computer models. In this study, a commercial rapid prototyping system containing a Digital Micromirror Device (DMD™) was utilized for dynamic mask microstereolithography. In visible light dynamic mask microstereolithography, a dynamic mask is used for selective polymerization of a photosensitive material. Since an entire layer of material is polymerized at once, processing times associated with visible light dynamic mask microstereolithography are significantly less than those associated with other rapid prototyping processes. Through use of a system with a large build envelope, multiple structures can be built in parallel. Visible light dynamic mask microstereolithography has been used to prepare several types of microscale biomedical devices. Sun et al. demonstrated use of Digital Micromirror Device-based microstereolithography to create a micro coil array with a wire diameter of 25 mm and a coil diameter of 100 mm.40 In addition, they fabricated a high aspectratio micro rod array. In this study, a curing depth of the resin of 45 mm was obtained. Covington et al. used microstereolithography to create a microfluidic structure for a biomimetic microsensor array-based olfactory device.41 Snowden et al. used microstereolithography to create 192 or 250 mm high JOM • June 2011

Table I. Hole Fabrication Data Input (µm)

Square Actual (µm)

% Holes Open

Input (µm)

Circle Actual (µm)

% Holes Open

800

656.25 ± 15.83

100%

800

No Open Holes

0%

700

429.2 ± 42.5

100%

700

No Open Holes

0%

600

245.9 ± 94.3

40%

600

No Open Holes

0%

550

245.5

20%

550

No Open Holes

0%

500

219.4

20%

500

No Open Holes

0%

450

199.6

20%

450

No Open Holes

0%

400

No Open Holes

0%

400

No Open Holes

0%

of microneedle arrays, microfans, and other high aspect ratio structures.45 Park et al. discussed fabrication of a 3×3 array of microcone cylinders by means of microstereolithography; a 5 mm wide tip and a maximum error of ~2 mm at the tip were obtained.46 More recently, Miller et al. used Digital Micromirror Device-based microstereolithography to create a hollow microneedle array out of an acrylate-based polymer.47 Carbon fiber electrodes were incorporated within the microneedles; these electrodes were modified to facilitate monitoring of ascorbic acid and hydrogen peroxide. Detection of hydrogen peroxide and ascorbic acid by the carbon fibers within the integrated electrode-hollow microneedle devices was demonstrated. In addition, Park et al.

microfluidic components for electrochemical flow detection; laminar flow conditions with volume flow rates of up to 64 mL min−1 were obtained with these structures.42 Stampfl et al. used microstereolithography to create structures, including a free spinning turbine wheel and a fixed axe, out of hybrid sol–gels, hydrogels, and elastomers.43 They described processing of materials with a wide range of elastic moduli (0.1 MPa–8,000 MPa) by means of microstereolithography. Neumeister et al. discussed fabrication of complex small-scale structures (e.g., small-scale micromechanical components) out of hybrid sol–gel materials.44 Choi et al. described use of a custom-built Digital Micromirror Device-based microstereolithography system for fabrication 600

Input File Actual Line

Line Width (mm)

500 400 300 200 100 0

0 a

2

4

6 Line Number

Figure 1. (a) Input line width and measured line width dimensions for the line width resolution study. Error bars indicate standard deviation of mean values. (b) Image of diffusion-driven polymerization in a line width resolution test structure.

Vol. 63 No. 6 • JOM

8

10

b

12

1 mm


used a segment curing method to create micro-lens arrays; they also discussed the use of gray-scaled cross-sectional images to improve the surface profiles of microstereolithography-fabricated structures.48 Digital Micromirror Device-based microstereolithography has also been used to create prostheses and scaffolds for tissue engineering. For example, Gittard et al. utilized visible light dynamic mask microstereolithography to create carpal bone prostheses from patient computed tomography data.49 The stereolithography-fabricated scaphoid and lunate prostheses exhibited ~50-mm thick layers. The minimum compressive forces necessary for fracture of the scaphoid and lunate prostheses were 1360 N and 1248 N, respectively. Han et al. used a Digital Micromirror Device-based system to create three-dimensional scaffolds out of poly(ethylene glycol) diacrylate.50 These multilayered scaffolds were covalently conjugated with fibronectin to facilitate attachment of murine marrow-derived progenitor cells. Choi et al. described processing of porous biodegradable poly(propylene fumarate) scaffolds with ~100 mm interconnected pores by means of Digital Micromirror Device-based microstereolithography.51 More recently, Yasar et al. created branched structures for tissue engineering using Digital Micromirror Device-based microstereolithography.52 Antimicrobial coatings were deposited on the visible light dynamic mask microstereolithography-fabricated microneedles by means of pulsed laser deposition.53 In this physical vapor deposition process, material from a solid target is ablated using a high energy laser (e.g., an excimer laser). The atomic and molecular species generated by laser ablation exhibit an average kinetic energy of species of 100–1,000 kT.54,55 On the other hand, thermal deposition techniques provide species with energies on the order of kT, which is 0.0259 eV at 300 K. Chemical reactions between the substrate and the high energy species in the growing film serve to enhance film adhesion.56 In addition, thin films deposited by pulsed laser deposition exhibit high densities and low porosities. Pulsed laser deposition is well suited for film growth on polymer 63

substrates since deposition of many films can be performed at room temperature. In previous work, zinc oxide films were grown on silicon wafers using pulsed laser deposition; CDC Biofilm Reactor and disk diffusion studies confirmed the antimicrobial properties of these materials.57 In this paper, the use of visible light dynamic mask microstereolithography combined with pulsed laser deposition to prepare antimicrobial microneedles was investigated. Studies were initially performed to determine the resolution of the microstereolithography system. Solid microneedles with four different geometries were produced; one of these geometries was subsequently selected for further evaluation. Needles in a cross shape, which may be used for wound closure, were also created using visible light dynamic mask microstereolithography. Pulsed laser deposition was then used to coat microneedle arrays with silver or zinc oxide thin films. Scanning electron microscopy, atomic force microscopy, and profilometry were used to examine microneedle surface morphology, pulsed laser deposition-grown film structure, and film thickness, respectively. Penetration of the visible light dynamic mask microstereolithography-produced microneedles into porcine skin was confirmed by trypan blue staining. Agar diffusion assays were used to examine the interactions between coated microneedles and Escherichia coli, Staphylococcus epidermidis, and Staphylococcus aureus bacteria. Our findings show that visible light dynamic mask microstereolithography combined with pulsed laser deposition is an appropriate approach for creating solid microneedle arrays with antimicrobial properties. rEsults and discussion Optical microscopy images of the test structures produced by means of visible light dynamic mask microstereolithography are provided in Figure Bd–f. In these images, the largest input hole (800 µm) is on the left for the hole resolution images; the widest input line width (140 µm) is at the bottom for the line width images. As can be seen in these images, the lateral dimensions of the visible light dynamic mask microstereolithography-fabricated structures 64

a

200 mm

200 mm

200 mm

200 mm

200 mm

200 mm

200 mm

200 mm

200 mm

b

200 mm c

200 mm d

200 mm

Figure 2. SEM images of eShell 200 microneedle array structures, which were produced using visible light dynamic mask microstereolithography. The geometries of the input STL file are (a) 750 mm x 250 mm base and 1,000 mm height for the “A” microneedle array structure, (b) 500 mm x 250 mm base and 1,000 mm height for the “B” microneedle array structure, (c) 750 mm x 250 mm base and 1,250 mm height for the “C” microneedle array structure, and (d) 500 mm x 250 mm base and 1,250 mm height for the “D” microneedle array structure. The images show the shorter axis of the microneedle, the longer axis of the microneedle, and a perspective view. The images were obtained at a 60º tilt.

are significantly larger than the lateral dimensions of the corresponding input files. The system was better at producing holes with square inputs than those with circular inputs. It should be noted that the holes produced using square input file exhibit rounded corners. The square 800 µm and 700 µm input files were able to consistently produce holes; on the other hand, the corresponding circular input file produced no open holes. Although some holes were able to be produced for smaller input sizes, these holes were not consistently produced. The input values and measured values for the square and circular holes are provided in Table I. All of the holes were more than 140 µm smaller than the corresponding input dimensions; the hole size error ranged from 144–354 µm. As can be seen from the line width www.tms.org/jom.html

resolution test image, only input structures containing trenches with widths of 500 µm or greater produced trenches in the corresponding test structures. A plot of input line width values and measured line width values across the largest gap (500 µm) for the line width resolution structure is provided in Figure 1a. The measured line width values were determined by averaging the line widths of four different structures; the error bars indicate the standard deviation of the mean. The measured line width values were noted to be more than two times greater than the input line width values; the size discrepancy increased with the input line width. It is interesting to note that the 60 µm wide input line resulted in production of a line; this result indicates that a pixel does not need to be completely filled for the pixel to be illuminated. JOM • June 2011

a

100 mm a

1 mm

Radius of curvature = 18 mm

200 mm b Figure 3. SEM images of the “D” microneedle array structure; the input dimensions for this structure are (a) 250 × 500 µm base and (b) 1,250 µm height. Plan view and 90º tilt images are presented; the radius of curvature of the tip is shown on the 90º tilt image.

However, lines less than 60 µm wide did not result in illumination of pixels. This study indicates that visible light dynamic mask microstereolithography system software may completely polymerize material in locations where incomplete structures or surfaces exist in the corresponding input file. As an example, a feature with a size of one pixel, which lies between four pixels, may result in the illumination of four pixels; in this case, the measured dimensions of the visible light dynamic mask microstereolithography systemfabricated feature would be much larger than the input dimensions. Scanning electron micrographs of the uncoated four different geometries are shown in Figure 2. The images show the shorter axis of the microneedle, the longer axis of the microneedle, and a perspective view; all of these images were obtained at a 60º tilt. The input and measured dimensions of the microneedles are provided in Table A. For all of the microneedle geometries, the measured lateral dimensions were larger than the input lateral dimensions. In contrast, the measured vertical dimensions were shorter than the input vertical dimensions. Although the input heights were identical, the 750 µm Vol. 63 No. 6 • JOM

b Figure 4. (a) SEM image of microneedles on a cross-shaped substrate that was produced by visible light dynamic mask micro-stereolithography. (b) CAD drawing of the input file that was used to produce the structure using visible light dynamic mask micro-stereolithography.

wide structures were taller than the 500 µm wide structures. As shown in Table A, the lateral dimensions of the microneedle were larger than the corresponding input data. The heights of the microneedles were shorter than the corresponding input data. In visible light dynamic mask microstereolithography, the surface of the STL model is converted in a process that is known as tessellation to a series of polygons.65 This series of polygons is sliced into cross-sectional layers, which are subsequently used for fabrication of the structure (e.g., the microneedle device) in a layer-bylayer manner. Variations between the input dimensions and measured dimensions are ascribed to translation of the STL model to the physical structure. The discrepancy in lateral dimensions, particularly the fact that the measured dimensions are larger than the input dimensions, can be attributed to the fact that the lateral voxel size (155 µm) is larger than the pixel size (70 µm). There are also discrepancies in the vertical dimensions between the test structures and the visible light dynamic mask microstereolithography-fabricated structures. The pixel size in the projection mask corresponds to approximately 70 www.tms.org/jom.html

µm; features below a certain size may not result in production of features by the visible light dynamic mask microstereolithography system. Higher magnification top-down and side views were used to obtain z-step size, tip radius of curvature, and aspect ratio measurements (Figure 3). The build layers can clearly be seen in the scanning electron micrographs of the visible light dynamic mask microstereolithography-produced microneedles. The z-direction step size was measured to be 29.89 ± 0.61 µm. The radius of curvature was measured to be 18 µm. The aspect ratio of wide face of the D geometry microneedle was obtained by dividing the average z-step size by twice the x and y step distance from layer to layer (via the top-down image) over the entire microneedle length. The average aspect ratio measured in a layer-by-layer manner along the entire length of the needle was 2.4 ± 0.3 to 1; the net aspect ratio (total height divided by width) was determined to be 2.2 to 1. Other factors that contribute to discrepancies between input and measured dimensions in visible light dynamic mask microstereolithography are diffraction, refraction, and photoinitiator diffusion.61–63 Quantitative analysis of voxel size spreading due to diffraction has been performed by Sun et al.61 In the visible light dynamic mask microstereolithography system, the halogen bulb used as a light source emits a spectrum that contains a range of wavelengths; in addition, the photoinitiator can be excited by a range of wavelengths. Light of various wavelengths will refract in different amounts as it passes through the material. In addition, already polymerized material will refract differently than unpolymerized material.62 Diffusion-driven polymerization can result in fusing of nearby features. In diffusion driven polymerization, electronically-excited photoinitiator molecules diffuse out of the voxel and produce membrane-like features between nearby features. The membrane-like features of diffusion driven polymerization may be clearly observed in some of the line resolution structures between the larger lines (Figure 1b). This phenomenon has been observed in other photopolymer65

ization processes, including two photon polymerization.63 The appearance of layers in the polymerized structures indicates that the voxel height is larger than the step size and that there is significant overlap between the voxels during processing. The visible light dynamic mask microstereolithography system was able to produce 32 microneedle arrays in a single, four-hour batch when using a 30µm layer spacing. Both the low input material cost and the high production rate (approximately 7.5 minutes per microneedle device) make the visible light dynamic mask microstereolithography system an appealing approach for scalable production of microneedles. A microneedle device with a crossshaped geometry was also able to be produced with visible light dynamic mask microstereolithography; the STL file used to make this device is shown in Figure 4a. A scanning electron micrograph containing two arms of the cross-shaped microneedle device is shown in Figure 4b. A device with this cross-shaped geometry may be used for facilitating wound healing and holding the skin together. Porcine skin and trypan blue have been used to assess the functionality of microneedles for transdermal drug delivery.47,64,65 Figure 5a and b shows optical micrographs of microneedles before insertion into skin and after insertion into skin, respectively, which show that the microneedles remain intact after entering skin. Optical micrographs showing delivery of trypan blue into microneedle-fabricated pores within skin are shown in Figure 5c and d. The blue colored spots (identified by arrows) indicate penetration of trypan blue dye through the stratum corneum layer by the microneedle as well as localization of trypan blue inside the microneedlegenerated pore. The pores are larger in one direction than the other, which is in accordance with the microneedle dimensions. Optical microscopy images of the pulsed laser deposition-coated microneedle devices are shown in Figure 6a and b. The silver-coated microneedle device (Figure 6c), zinc oxidecoated microneedle device (Figure 6b), and uncoated microneedle device (Fig66

ure 6a) are shown. Differences in the surface roughness of the zinc oxidecoated surface and the silver-coated surface are noted in Figure 6e and f. Atomic force micrographs obtained in tapping mode of the zinc oxide and silver thin films on silicon substrates are shown in Figure 7. Root mean squared surface roughness values for silver thin films and zinc oxide thin films were 1.4 nm and 1.3 nm, respectively. The low roughness values of the thin films indicate that splashing was not significant.51 The thicknesses of the coatings were measured by profilometry to be 53.5 ± 8.8 nm for Ag and 287.2 ± 30.3 nm for ZnO. Deposition rates were calculated from profilometry data; rates of 0.018 nm/pulse and 0.10 nm/pulse were determined for silver and zinc oxide, respectively. In comparison, Choi et al. observed a zinc oxide growth rate of 0.017–0.033 nm/pulse for pulsed laser deposition involving an ArF ex-

cimer laser.66 Warrender et al. observed a silver growth rate of 0.0018–0.006 nm/pulse for pulsed laser deposition involving a 248 nm KrF excimer.67 Variations in deposition rate can be attributed to laser fluence, target temperature, pressure, and other parameters.54,66,67 Images of the agar diffusion assay results for silver-coated microneedle devices, zinc oxide-coated microneedle devices, and uncoated microneedle devices are presented in Figure 8. The agar diffusion assay results indicate that silver thin films and zinc oxide thin films provided the microneedle devices with antibacterial properties. Both thin films had significant effects against Staphylococcus epidermidis and Staphylococcus aureus. On the other hand, significant effects against Escherichia coli were not noted. The silver-coated microneedle devices produced zone voids of growth; in

1 mm

Figure 5. (a) Optical micrograph showing two microneedles before insertion into porcine skin. (b) Optical micrograph showing two microneedles after insertion into porcine skin. (c,d) Optical micrographs showing delivery of trypan blue into microneedle-fabricated pores within porcine skin.

a

a

1 mm

b 1 mm

1 mm

c

d

b

c

2 mm

50 mm 50 mm 50 mm d e f Figure 6. (a–c) Optical microscopy images and (d–f) high magnification SEM images of (a,d) uncoated, (b,e) zinc oxide-coated, and (c,f) silver-coated microneedle array structures.


JOM • June 2011

15.0 nm

7.5 nm

0.0 nm

0

2.5

5.0

7.5

10.0

15.0 nm

7.5

7.5 nm

10.0

7.5

5.0

0.0 nm 5.0

2.5

2.5

0 10.0 mm 0

2.5

5.0

7.5

0 10.0 mm

Figure 7. AFM topographical image of ZnO coating on silicon (left) and Ag coating on silicon (right); these coatings were deposited by means of pulsed laser deposition. These images were obtained from 10 µm × 10 µm areas with identical height (z-) scale.

Control

Ag

ZnO

E. coil

S. epidermids

S. aureus

contrast, the zinc oxide-coated microneedle devices produced zones of inhibited or reduced growth. Some of the thin films remained on both the silver-coated microneedle devices and the zinc oxide-coated microneedle devices after 24 hours of contact with the agar. In comparison, the uncoated arrays had no antimicrobial effect. As mentioned in the introduction, zinc oxide exhibits weaker bacteriostatic inhibition of Gram-negative bacteria than of Grampositive bacteria.68 The agar diffusion assay confirmed previous findings that pulsed laser deposition-grown thin films of both silver and zinc oxide provide antibacterial properties.38,57 conclusions In this study, we have demonstrated that visible light dynamic mask microstereolithography is an appropriate technique for scalable production of microneedle devices. Although the visible light dynamic mask microsteVol. 63 No. 6 • JOM

Figure 8. Agar diffusion assay results for uncoated (left), Ag-coated (center), and ZnO-coated (right) “D” geometry microneedle array structures. The microneedle array structures were evaluated using Escherichia coli (top), Staphylococcus epidermidis (center), and Staphylococcus aureus (bottom). The microneedle array structures examined in this study were distributed over 18 mm ×18 mm areas.

reolithography system reproduces lateral dimensions with good fidelity, the lengths of the tapering structures were significantly reduced. The successful fabrication of microneedles on crossshaped substrates, which may be for holding the skin together and facilitating wound healing, as well as square substrates demonstrates that visible light dynamic mask microstereolithography provides flexibility for creating medical devices with complex shapes. The combination of dynamic mask stereolithography with pulsed laser deposition was used to prepare acrylatebased polymer microneedle arrays with antimicrobial properties; these devices may be useful for wound treatment. Agar diffusion assays confirmed that pulsed laser deposition-grown silver and zinc oxide coatings exhibit antimicrobial activity. The ability to produce microneedle arrays for treatment of skin infections by means of visible light dynamic mask microstereolithogwww.tms.org/jom.html

raphy and pulsed laser deposition is a promising step towards bringing microneedle technology into wider use. acknowlEdgEmEnts Portions of the research in this paper were funded from Office of Naval Research grant N00014-07-1-1099. references 1. P.B. Cornia et al., Expert Opinion on Pharmacotherapy, 9 (2008), pp. 717–730. 2. E.V. Ki and C. Rotstein, Canadian Journal of Infectious Disease and Medical Microbiology, 19 (2008), pp. 173–184. 3. M.E. Jones et al., International Journal of Antimicrobial Agents, 22 (2003), pp. 406–419. 4. R. Durai et al., AORN Journal, 91 (2010), pp. 599– 606. 5. K.K.Y. Wong and X. Liu, Med. Chem. Comm., 1 (2010), pp. 125–131. 6. M.E. Samberg et al., Environmental Health Perspectives, 118 (2010), pp. 407–413. 7. M.E. Samberg et al., Nanotoxicology, in press (doi: 10.3109/17435390.2010.525669). 8. P.L. Nadworny and R.E. Burrell, J. Wound Technology, 2 (2008), pp. 6–12. 9. M. Schaller et al., Skin Pharmacology & Physiology, 17 (2004), pp. 31–36. 10. M. Bhattacharyya and H. Bradley, Int. J. Lower Extremity Wounds, 7 (2008), pp. 45–48. 11. C.E. Schnopp et al., Expert Opinion on Pharmacotherapy, 11 (2010), pp. 929–936. 12. T. Jun et al., Chemmedchem, 2 (2007), pp. 129– 136. 13. X.L. Liu et al., Chemmedchem, 5 (2010), pp. 468–475. 14. J.B. Wright et al., Wound Repair and Regeneration, 10 (2002), pp. 141–151. 15. D.M. Elston, Dermatologic Clinics, 27 (2009), pp. 25–31. 16. A.B.G. Lansdown et al., Wound Repair and Regeneration, 15 (2007), pp. 2–16. 17. S. Atmaca et al., Turkish Journal of Medical Sciences, 28 (1998), pp. 595–597. 18. J. Sawai, J. Fermentation and Bioengineering, 86 (1998), pp. 521–522. 19. Y. Liu et al., J. Applied Microbiology, 107 (2009), pp. 1193–1201. 20. Z. Huang et al., Langmuir, 24 (2008), pp. 4140– 4144. 21. H. Akiyama et al., J. Dermatological Science, 17 (1998), pp. 67–74. 22. M. S. Agren et al., European Journal of Surgery 157 (1991), pp. 97–101. 23. T. Soderberg et al., Infection, 17 (1989), pp. 81–85. 24. E.F. Rostan et al., Int. J. Dermatology, 41 (2002), pp. 606–611. 25. M.S. Agren et al., J. Wound Care, 13 (2004), pp. 367–369. 26. M.S. Agren et al., Wound Repair and Regeneration, 14 (2006), pp. 526–535. 27. S.D. Gittard and R.J. Narayan, in Toxicology of the Skin, ed. N.A. Monteiro-Riviere (New York: Informa Healthcare, 2010), pp. 301–316. 28. S.D. Gittard et al., Expert Opinion on Drug Delivery (2010), pp. 513–533. 29. M.I. Haq et al., Biomedical Microdevices, 11 (2009), pp. 35–47. 30. S.M. Bal et al., European Journal of Pharmaceutical Sciences, 35 (2008), pp. 193–202. 31. R.K. Sivamani et al., Skin Research and Technology, 11 (2005), pp. 152–156. 32. H.S. Gill et al., Clinical Journal of Pain, 24 (2008),

67

pp. 585–594. 33. R.F. Donnelly et al., Pharmaceutical Research, 26 (2009), pp. 2513–2522. 34. S. Doddaballapur, J. Cutaneous and Aesthetic Surgery, 2 (2009), pp. 110–111. 35. I. Majid, J. Cutaneous and Aesthetic Surgery, 2 (2009), pp. 26–30. 36. B.S. Chandrashekar and A.S. Nandini, J. Cutaneous and Aesthetic Surgery, 3 (2010), pp. 125–126. 37. E.M. Lane, U.S. patent application 20080085301A1 (2008). 38. S.D. Gittard et al., Biofabrication, 1 (2009), pp. 041001. 39. S.D. Gittard et al., Advanced Engineering Materials, 12 (2010), pp. 77–82. 40. C. Sun et al., Sensors and Actuators, A121 (2005), pp. 113–120. 41. J.A. Covington et al., IET Nanobiotechnology, 1 (2007), pp. 115–121. 42. M.E. Snowden et al., Analytical Chemistry, 82 (2010), pp. 3124–3131. 43. J. Stampfl et al., J. Micromechanics and Microengineering, 18 (2008), pp. 125014. 44. A. Neumeister et al., J. Laser Micro/Nanoengineering, 3 (2008), pp. 67–72. 45. J.W. Choi et al., J. Mechanical Science and Technology, 20 (2006), pp. 2094–2104. 46. I. Park et al., Int. J. Advanced Manufacturing Technology, 46 (2010), pp. 151–161. 47. P.R. Miller et al., Biomicrofluidics, doi:10.1063/1.3569945.

48. I. Park et al., Int. J. Precision Engineering and Manufacturing, 11 (2010), pp. 483–490. 49. S.D. Gittard et al., Biotechnology Journal, 4 (2009), pp. 129–134. 50. L.H. Han et al., J. Manufacturing Science and Engineering, 130 (2008), pp. 021005. 51. J.W. Choi et al., J. Materials Processing Technology, 209 (2009), pp. 5494–5503. 52. O. Yasar et al., Biofabrication, 1 (2009), pp. 045004. 53. M.L. Morrison et al., Diamond and Related Materials, 15 (2005), pp. 138–146. 54. J.M. Warrender and M. Aziz, Physical Review B, 75 (2007), pp. 085433. 55. P.R. Willmott, Progress in Surface Science, 76 (2004), pp. 163–217. 56. J.M. Lackner et al., Surface and Coating Technology, 188-189 (2004), pp. 519–524. 57. S.D. Gittard et al., Applied Surface Science, 255 (2009), pp. 5806–5811. 58. Technical Data: envisionTEC e-Shell 200 Series. http://www.envisiontec.de/fileadmin/pdf/MatSheet_ eShell200_en_s.pdf (Retrieved 27 January 2011). 59. S.D. Gittard et al., J. Diabetes Science and Technology, 3 (2009), pp. 304–311. 60. General Discussion, Faraday Discussions, 14 (2011), pp. 227–245 (DOI:10.1039/C0FD90010A). 61. C. Sun et al., Sensors and Actuators A, 121 (2005), pp. 113–120. 62. M. Miwa et al., Applied Physics A, 73 (2001), pp. 561–566. 63. A. Ovsianikov et al., Acta Biomaterialia, 7 (2011),

pp. 967–974. 64. J.H. Park et al., IEEE Transactions in Biomedical Engineering, 54 (2007), pp. 903–913. 65. H.S. Gill and M.R. Prausnitz, J. Controlled Release, 117 (2007), pp. 227–237. 66. J.H. Choi et al., J. Crystal Growth, 226 (2001), pp. 493–500. 67. M.J. Aziz, Applied Physics A, 93 (2008), pp. 579– 587. 68. J. Sawai et al., J. Chemical Engineering Japan, 28 (1995), pp. 288–293. Shaun D. Gittard, Philip R. Miller, Chunming Jin, Timothy N. Martin, Ryan D. Boehm, Nancy A. Monteiro-Riviere, and Roger J. Narayan are with the Joint Department of Biomedical Engineering, University of North Carolina and North Carolina State University, Chapel Hill, NC 27599; Monteiro-Riviere is also with the Center for Chemical Toxicology Research and Pharmacokinetics, Department of Clinical Sciences, North Carolina State University, Raleigh, NC 27695, USA; Bret J. Chisholm, Shane J. Stafslien, Justin W. Daniels, and Nicholas Cilz are with the Center for Nanoscale Science and Engineering, North Dakota State University, 1805 Research Park Drive, Fargo, ND, 58102; and Adnan Nasir is with the Department of Dermatology, University of North Carolina, Chapel Hill, NC, and Wake Research Associates, 3100 Duraleigh Rd. Ste. 304, Raleigh. Dr. Narayan can be reached at (919) 696-8488, e-mail [email protected].

Roger Narayan is a TMS Member! To read more about him, turn to page 9. To join TMS, visit www.tms.org/Society/Membership.aspx.

68


JOM • June 2011