Printing technologies are nowadays used for the design and manufacture of products by deposition of different materials with electrical, optical, chemical, biological, and structural functionalities [1
]. In particular, digital (2D and 3D) printing technologies are appealing for pharmaceutical applications since state-of-the-art technologies enable the deposition of different liquid, semi-solid, and solid materials containing one or several active pharmaceutical ingredients (APIs) according to predefined designs [3
]. Consequently, printing technologies can be used to manufacture personalized doses and drug delivery systems. The possibility of late-stage customization of the medicine according to a patient’s specific needs as well as the flexibility in production volumes make the manufacturing method applicable for decentralized manufacturing of pharmaceuticals at the point of care [7
The drug discovery approach, utilizing combinational chemistry and high-throughput screening, has resulted in vast libraries of poorly water-soluble and poorly permeable lead compounds and drug candidates [8
]. Consequently, different formulation strategies have been developed to overcome the hurdles of poor drug delivery and poor bioavailability [10
]. Formulation approaches utilizing nanoparticles in combination with printing technologies have been conducted to address this issue. Pharmaceutical nanosuspensions have been developed and deposited by means of roll-to-roll [11
] and drop-on-demand (DoD) printing technologies [5
]. Accurate deposition of antibacterial agents onto implant materials, development of transdermal/mucosal and oral, immediate and controlled release drug delivery systems have been demonstrated to be doable using inkjet technology. Furthermore, the possibility of printing different geometries has been shown to be of importance for tailoring the drug release from the printed drug delivery system [16
]. The main reason for the development of the polymeric nanoparticles and cyclodextrin inclusion complexes has been to improve and control the delivery of especially potent and poorly soluble drugs. Apart from the manufacture of pharmaceuticals and modification of implants, printing of nanosuspensions has been used for labeling, diagnostic, and screening applications; Quick Response (QR) codes have been printed using RGB (red, green, blue) nanoparticle inks for anti-counterfeit purposes on capsules [18
] and nanoparticle-based lab-on-a-chip platforms have been developed by combining microfluidics, electronics, and inkjet printing [19
]. Inkjet printing has also been utilized as a screening platform in combination with a separate detection method for rapid, low-cost, high-throughput screening of chemical libraries and novel biomaterials [20
]. Systematic investigations concerning the development of inkjet printable nanosuspensions have been conducted due to interest in inkjet-printed electronics [22
]. The composition of the nanosuspensions with regard to the physicochemical properties of the particles (size, polydispersity and net surface charge), particle concentration, excipient addition (surfactants), and solvent system, has an impact on the performance of the ink formulation [2
]. Furthermore, the interplay between the ink formulation and the substrate must also be considered since it dictates the final deposit morphology of the printed product [25
]. Previously, investigations regarding inkjet-printed drop morphologies of monodisperse silica and silicon dioxide microspheres onto various hydrophilic and hydrophobic materials have been performed [25
Utilization of nanomaterials has proven to be favorable in the fields of drug delivery, in vitro diagnostics, and in vivo imaging, and for the development of biomaterials, implants, and coatings [30
]. These materials have shown to enable improvement or added functionality of previous compounds and could also have potential in the development of super generics [32
]. Such studies showing improved delivery and usability of poorly soluble, poorly permeable, very potent APIs [33
], biomolecules [34
], proteins [35
], peptides [36
], antibodies [37
], and genes [38
] have been published to date.
The FDA-approved hydrophobic colloidal silica material has already been widely used in different pharmaceutical formulations. However, the first proof-of-concept study using ordered mesoporous silica as drug delivery vehicle of a poorly soluble drug was only conducted recently in human [39
]. Mesoporous silica nanoparticles (MSNs) have still not been studied in man, but have been introduced as a versatile and biocompatible drug delivery platform [40
]. Immediate and controlled release MSN-based drug delivery systems have been developed for oral [33
], transdermal [46
], and intravenous (i.v.) [38
] administration of compounds with poor stability or poor solubility. Great potential has been seen in the application of MSNs as targeted drug delivery vehicles for anti-cancer drugs [37
]. Furthermore, uni- and dual-stimuli-responsive MSN drug delivery mechanisms have proven to be possible, having, i.e., a change in pH as a trigger for the cargo deposition [47
]. The versatility in drug delivery described above is enabled by the design and control of key features of the mesoporous material such as particle size, particle surface, and pore size, shape, and volume. Tailoring the surface properties of the particles allows for extensive design opportunities. An essential property of the functionalization is the improved dispersibility of the nanoparticles in media and the benefit seen in studies investigating the cellular uptake of the functionalized particles [38
Biopharmaceutics Classification System (BCS) class IV drugs have poor water solubility and poor permeability. Thus, formulation strategies including MSNs could possibly address these issues by facilitating drug delivery. Mesoporous silica nanoparticles have previously been printed using soft-lithography and inkjet printing as a top-down method to create micro- and nanostructures [49
]. However, deposition of the MSNs has not been studied for pharmaceutical applications. The purpose of this study was to develop a pharmaceutical MSN-based ink formulation, including a BCS class IV drug, to be deposited onto polyester transparency and hydroxypropylmethyl cellulose (HPMC) films using inkjet printing technology. The impact of drug-loaded compared to drug-free MSN–PEI was investigated with regard to (1) ink formulation development; (2) printability, as well as the (3) ink–substrate interactions. Attention was drawn to the print morphology regarding the drop and nanoparticle deposition patterns. Furthermore, a comparison of different drop deposition characterization techniques was also made. Here, we show that inkjet printing could be used for accurate deposition of pharmaceutical MSN ink suspensions according to predefined patterns, suggesting that the method could be utilized flexibly for screening purposes.
4. Materials and Methods
4.1. Synthesis of MSNs
The MSNs were synthesized in an aqueous basic solution with the addition of absolute methanol (HPLC gradient grade, J.T. Baker, Philipsburg, NJ, USA) as co-solvent. Cetyltrimethylammonium chloride (CTAC solution, 25 wt % in H2
O, Sigma-Aldrich, Steinheim, Germany) was dissolved in the basic reaction solution and served as a structure-directing agent (SDA). Tetramethylorthosilicate (TMOS, 99%, Sigma-Aldrich) was added to the reaction solution as silica source. The reaction was conducted in a conical flask overnight under stirring at room temperature. The molar composition in the synthesis solution was TMOS:CTAC:NaOH:MeOH:H2
O (1:1.4:0.3:1433.7:3188.7). Sodium hydroxide (NaOH) was purchased from Merck KGaA, Damstadt, Germany. Fluorophore labeling of MSNs was carried out according to the protocol of our previous work [48
]. Briefly, a mixture of fluorescein isothiocyanate (FITC, Sigma-Aldrich) and aminopropyl trimethoxysilane (APTMS, Sigma-Aldrich) (FITC:APTMS, 1:3) was pre-reacted in methanol for 30 min and added to the synthesis solution. Finally, the silica source TMOS was added (TMOS:APTMS, 100:1). The formation of the particles took place overnight under continuous stirring at room temperature.
After the particle synthesis was accomplished, the structure-directing agent was removed by washing the particles with an extraction solvent. The extraction solvent was a 1:8 mixture of HCl (37–38%, J.T. Baker, Philipsburg, NJ, USA) and ethanol (99.5%, Altia Oy, Helsinki, Finland). The dispersion was centrifuged and the supernatant was removed and replaced with fresh extraction solvent. The solution system was sonicated for 30 min after which the washing cycle, starting with centrifuging, was repeated three times. After the three cycles d, pure ethanol was added to wash away the extraction solvent. Half of the particles were kept for drying in vacuum whereas the other half was stored as an ethanol dispersion for further polyethyeleneimine (PEI, Sigma-Aldrich) surface functionalization. The surface of the MSNs was modified with PEI by the surface growing method according to our previously described protocols and the samples were named as MSN-PEI in the study [56
]. The prepared samples were kept in the fridge as an ethanol suspension. MSN-PEI particles with and without FITC label was used for drug loading in the study.
4.2. Characterization of MSN and MSN-PEI
The dispersibility of the unloaded and drug-loaded MSN-PEI in the solvent mixture, the hydrodynamic particle size (Z-average, intensity) and ζ-potential were analyzed with a DLS/ζ-potential instrument (Malvern ZetaSizer NanoZS, Malvern Instruments Ltd., Malvern, UK). The efficiency of the functionalization was evaluated based on the ζ-potential data. The weight percentage of accommodated PEI in respect to MSN was determined using thermogravimetric analysis (TGA, STA 449 F1 Jupiter, NETZSCH, Selb, Germany). The analysis was performed over a temperature range of 25 °C to 770 °C. Nitrogen adsorption (Autosorb-1, Quantachrome Instruments, Boynton Beach, FL, USA) was performed to evaluate the surface area, pore size and pore volume or in general the porosity of the particle. The fine structure of MSN and MSN-PEI particles were analyzed by transmission electron microscopy (JEOL JEM-1400 Plus, JEOL Ltd., Tokyo, Japan). The particle size of the printed MSN-PEI and MSN-PEI-F15 (n = 30) on transparency was quantified using the SEM pictures and the image analysis program ImageJ (v. 1.50i 2011, National Institutes of Health, Bethesda, MD, USA).
4.3. Drug Loading
The BCS class IV drug, furosemide, F (Ph.Eur., Fagron Nordic, Copenhagen, Denmark) was chosen as a model drug to be loaded into the MSN-PEI samples in this study. The drug loading was carried out using the solvent immersion method. In this process 5 wt % and 15 wt % of F with respect to the MSN-PEI mass was soaked into a cyclohexane solution that contained MSN-PEI to obtain particles with 5 and 15 wt % loading degrees (abbreviated as MSN-PEI-F5 and MSN-PEI-F15, respectively). The drug-loaded suspensions were ultrasonicated and kept in a rotating wheel mixer overnight at room temperature. After adsorption of F, the F loaded MSN-PEI samples were centrifuged and the obtained precipitates were vacuum-dried. Afterwards, F elution was carried out in order to determine the amount of drug-loaded on MSN-PEI. The loading degrees were investigated by preparation of 1 mg/mL suspensions in EtOH (Etax Aa 99.5%, Altia Oy). The suspensions were held for 30 min in a sonication bath and an additional 90 min in a rotating wheel mixer (50 rpm) protected from light. The suspensions were centrifuged (8000 rpm, 10 min) and the drug content was measured from the supernatant after dilution in EtOH.
4.4. Ink Preparation and Characterization
The 1 and 5 mg/mL nanoparticle suspensions were prepared by dispersing the MSN and MSN-PEI reverse-osmosis purified water (distilled water, MQ) using a Covaris Acoustic Ultrasonicator (Covaris, Brighton, UK). Propylene glycol (PG, ≥99.5%, Sigma-Aldrich), was added as a humectant and stabilizing agent to the suspension during sonication, after which proper dispersion of the particles in the aqueous phase was obtained. The solvent mixture consisted of equal volumes of MQ and PG. The solvent mixture MQ/PG and the MSN-PEI 1 and 5 mg/mL nanosuspensions were characterized with regard to their physical fluid properties as described below.
4.4.1. Dynamic Viscosity
The dynamic viscosity was measured using a stress controlled rheometer (Physica MCR 300, Anton Paar, Graz, Austria, Software: Rheoplus), connected with a refrigerator bath and a temperature control unit (Techne RB-12A & TU-16D, Vernon Hills, IL, USA) and equipped with a double gap measurement geometry (DG26.7/T200/SS, internal ø: 24.655 mm, external ø: 26.669 mm, concentricity: ±8 µm). The dynamic viscosity of the inks was monitored after sample conditioning @ 22 ± 0.5 °C by application of a shear stress ramp at rates of 10, 100, and 1000 s−1.
4.4.2. Surface Tension and Density
A contact angle goniometer CAM 200 (KSV Instruments Ltd., Espoo, Finland, later Biolin Scientific) was used to measure the surface tension of the inks at room temperature (23 ± 0.5 °C). The pendant drop method was applied to measure the surface tension of the inks. A 5-µL drop was dispensed from a disposable plastic tip (Fintip 200 µL, Thermo Scientific, Vantaa, Finland) and imaged for 10 s. The recorded drop shape was fitted to the Young–Laplace equation using the OneAttension software (Theta1.4) to calculate the surface tension of the ink. The density of the suspensions was measured by weighing 1 mL of 23 ± 0.5 °C suspensions and calculating the density according to the obtained weight.
4.4.3. Colloidal Stability of MSN-PEI Suspensions
The 1 and 5 mg/mL MSN-PEI suspensions were irradiated in the near infrared region (air = 850 nm) with an electroluminescent diode using multiple light scattering (MLS, Turbiscan MA2000, FormulAction, Tolouse, France) to evaluate the colloidal suspension stability. Sampling was performed every minute during a total sampling time of 180 min from the 7-mL sample (25 °C, n = 1). The results were analyzed using the Turbisoft software (v 1.2.1, FormulAction, CIRTEM, Tolouse, France), generating mean transmission profiles.
4.4.4. Drug Release in Ink
The drug release from the MSN-PEI-F5 and F15 particles (c = 1 mg/mL, n = 3) into the ink was studied for 5 h. The particles were dispersed in MQ/PG and mixed in a rotating wheel mixer (50 rpm) protected from light. An aliquot of 2 µL was withdrawn every hour from the ink supernatant, gained by centrifuging the samples at 5000 rpm for 5 min. The ink suspensions were re-dispersed by vortexing and sonicating them after each sample withdrawal. Drug quantification was performed in the pendant mode of a UV-Vis spectrophotometer (Nanodrop 2000c spectrophotometer, Thermo Scientific) at λmax 273 nm.
4.5. Inkjet Printing
A piezoelectric inkjet printer, (PixDro LP50, Meyer Burger, Eindhoven, The Netherlands) equipped with a Spectra SL 128 AA, print head (nozzle Ø 50 µm, Fujifilm Dimatix Inc, Santa Clara, CA, USA), was used for sample preparation. The suspension was introduced to the ink container with a syringe without any filtration. Monitoring of working nozzles and droplet ejection was performed using a high-speed time-of-flight camera that was attached to the printer. Droplet formation was studied using the advanced drop analysis software (ADA, v.2.3, PixDro) by monitoring the droplet at 50–600 µs after ejection at a frequency of 1600 Hz. The print settings (ink pressure, voltage, and waveform) were optimized. Printing was performed at a speed of 200 mm/s, using one nozzle in the print head with the resolution set at 150 and 500 dpi. The droplets selected for the printing task were checked and analyzed with regard to their volume (n = 10) before every printed layer with the help of a high-speed camera. The “print view” was calibrated according to a three-image calibration procedure given by the manufacturer of the printer. Drop volume calculations were performed using the PixDro software based on a snapshot of the ink droplet. Substrate plate heating was set at 30 °C to facilitate drying of the deposits during the printing. The samples were stored at room temperature and protected from light. The printed doses (500 dpi, 10 layers) were left to dry at room temperature.
The inks were printed onto orodispersible hydroxypropylmethyl cellulose (HPMC) and polyester transparency films (Folex Imaging, Clear transparent X-10.0 films). The orodispersible films consisted of HPMC (15 wt %, Pharmacoat 606, Shin Etsu, Tokyo, Japan), glycerol (3 wt %, Sigma-Aldrich), and purified water and were cast using a film applicator (Multicator 411, Erichsen GmbH & Co. KG, Hemer, Germany), with the wet thickness set at 500 µm.
4.7. Quantification of the Prints
The BCS class IV model drug furosemide incorporated in the MSN-PEI was quantified from the MSN-PEI-F15 10 layer prints using a UV-Vis spectrophotometer (Nanodrop 2000c spectrophotometer, Thermo Scientific, Wilmington, MA, USA) at λmax 273 nm. The 1 × 1 inch samples (n = 3) were cut into four parts and placed in 1 mL of EtOH (99.9%, Etax Aa, Altia, Helsinki, Finland) in an Eppendorf vial. The samples were sonicated in a 25 °C water bath for 30 min and kept in a rotating wheel mixer for an additional 1.5 h. The MSNs and the HPMC film were centrifuged for 10 min in EtOH at 8000 rpm. The drug-free MSN-PEI prints were treated in the same manner and served as blank for the UV-Vis measurements.
4.8. Contact Angle
Contact angle measurements of the solvent mixture and the 1 and 5 mg/mL nanosuspensions (23 ± 0.5 °C) were performed on the transparency and HPMC films according to the sessile drop method by applying a 5-µL drop of ink onto the films in triplicate and monitoring the contact angle for 60 s. The measurements were performed using the same instrument as for the surface tension measurements described in Section 4.4.2
4.9. Visual Characterization of the Prints
The 1 and 5 mg/mL MSN-PEI and MSN-PEI-F deposits were characterized by confocal laser scanning microscopy (CLSM), scanning electron microscopy (SEM), optical microscopy, and scanning white light interferometry (SWLI).
4.9.1. Confocal Laser Scanning Microscopy
The prints were imaged using a Leica TCS SP 5 confocal scanning laser microscope (CLSM, Leica Microsystems GmbH, Wetzlar, Germany, Lenses: HCX PL APO 40×/1.15 and 63×/1.32 oil objectives) with an excitation wavelength of 488 nm.
4.9.2. Scanning Electron Microscopy
Scanning electron microscopy (SEM, LEO Gemini 1530, Carl Zeiss Microscopy GmbH, Oberkochen, Germany) was used to image the prints. The samples were pretreated with a carbon layer. Images were recorded at an acceleration voltage of 5 kV using the secondary electron detector. Images of 50, 100, 250, and 1000 time magnifications of the drop deposits, corresponding to a Polaroid 545 print with the image size of 8.9 × 11.4 cm, were captured.
4.9.3. Optical Microscopy
An optical microscopy imaging system (Evos XL Core Imaging System, Fisher Scientific GmbH, Schwerte, Germany) was used to image the prints using ×4, ×20, and ×40 magnification lenses (LPlan PH2).
4.9.4. Scanning White Light Interferometry
Scanning white light interferometry (SWLI) was used to obtain 3D images of the prints. All samples were imaged using a custom-made scanning white light interferometer instrument. The instrument featured a NIKON reflective microscope frame, equipped with a NIKON IC Epi Plan DI 10× MIRAU interferometry objective (Edmund Optics Ltd., Nether Poppleton, York, UK), a 100 μm piezoelectric z-scanner (Physik Instrumente P-721 PIFOC®, Karlsruhe, Germany), a high-resolution CCD camera (Hamamatsu Orka Flash 2.8 CMOS, Hamamatsu City, Japan), and two motorized translation stages (STANDA 8MTF-102LS05, Vilnius, Lithuania). As a white light source, a standard halogen lamp was used. The total magnification of this setup was ×6.3.
Scanning and data acquisition was controlled with a C++ based software built in-house. 3D image construction and 3D data analysis (e.g., determination of deposition diameter and deposition thickness) were performed using the commercial MountainsMap® Imaging Topography 7.4 software (Digital Surf, Besançon, France).
Three different 1 × 1 inch areas of each of the printed samples of interest were imaged with the SWLI instrument without any further sample pretreatment. If necessary, the samples were flattened using small weights to minimize the waviness originating from the film substrates. The 1 × 1 inch printed samples were imaged both in the center parts and in the peripheral parts of the printed area.