Transdermal Microneedle Sensor Arrays based on

Palladium:Polymer Composites

Aaron McConville and James Davis*

School of Engineering, Ulster University, Jordanstown, Northern Ireland, BT37 0QB

Abstract

The solvent based casting of metal particle-polymer mixtures has been investigated as a rapid means

through which to produce a 10 x 10 array of pyramidal (200 x 200 x 350 micron) microneedles for

electroanalytical sensing applications. The incorporation of nano particulate palladium powder

within either a polycarbonate or polystyrene binder is shown to result in mechanically robust

microneedles. The electrochemical properties of the resulting structures have been investigated and

their application for transdermal sensing applications has been demonstrated through the use of

epidermal / skin mimic.

Keywords: palladium, composite, microneedle, transdermal, sensing array

Text Words: 2954

References: -447

Figures 3x200 = 600

Total Words: 3107

* To whom correspondence should be addressed. T:+44 (0) 28 90366407; E:james.davis@ulster.ac.uk

1.0 Introduction

The application of microneedles for drug delivery applications is well established [1-4] but

there has been an increasing interest in the transfer of the technology for sensing purposes [5-10].

The near painless transdermal insertion of the microneedle array offers considerable advantages

over conventional sampling routes [4] but, in most cases, the devices are produced using relatively

sophisticated microfabrication and micromachining processes that can severely compromise the

accessibility of the technology [5-10]. The availability of silicone moulds for the production of

polymer composite microneedles however, has revolutionised research into transdermal drug

delivery [3,11-12] and, it could be envisaged that such systems would be an invaluable tool through

which to develop electrochemical sensors. It can be postulated that the incorporation of metallic

micro and nano particles within the composite formulation could speed the production of

conductive microneedle arrays suitable for a range of biomedical applications. In this

communication, the production of palladium microneedles has been investigated and their potential

applicability for transdermal electroanalysis is demonstrated.

Palladium modified electrodes have been extensively exploited in a wide variety of

electrochemical sensing applications in environmental [13,14], industrial [15,16] and biomedical

contexts [17,18] and are typically employed in the form of micro-nano particulate catalytic clusters

within composite constructions [13-16]. The latter typically involves carbon [13], carbon nanotubes

[14] or graphene/graphene oxide [15-17] but, as yet, there are few reports of the use of Pd in

microneedle architectures. Chandrasekaren et al. produced intricate Pd and Pd-Co alloy

microneedles through electrodeposition onto micromachined substrates but their electroanalytical

performance was not investigated [19]. It was therefore of interest to determine the viability of

constructing microneedle structures based on a more accessible microparticulate Pd composite. The

proposed strategy centred on the encapsulation of Pd powder (<1 micron diameter) within either

polycarbonate or polystyrene binders. While the exploitation of these polymeric systems in

microneedle fabrication is well established [3,20,21], the challenge here was to determine whether

or not the inclusion of the metallic particles, at the ratio necessary to ensure adequate conductivity

for electroanalytical purposes, would compromise the mechanical integrity and cohesiveness of the

needle structure. The proposed approach is based on solvent casting mixtures of powder and

polymeric binder into a silicone mould which, upon release, should produce a 10 by 10 array of

needles with the dimensions indicated in Figure 1A. Ideally, the needles should be of sufficient

length and mechanical rigidity to penetrate the epidermis of the skin (typically 100 m) without

triggering the sensory cells [22]. In this communication, the formulation of a skin mimic (Figure 1B)

based on a calcium alginate hydrogel loaded with ferrocyanide as a redox probe was used to

determine the electroanalytical properties of the composite Pd microarray. A parafilm™ barrier

(~130 m thick) was employed to act as the epidermal layer such that the ferrocyanide probe would

only be accessible upon the needles successfully puncturing the top film.

Figure 1. A) Microneedle dimensions and B) Skin mimic employing a non porous parafilm™ barrier

and alginate hydrogel loaded with a ferrocyanide redox probe.

2.0 Experimental Details

Materials: Palladium powder (<1 μm), polystyrene (avg. MW 192,000), L-cysteine (97%), potassium

ferrocyanide (≥99%), acetone (≥99.8%), dichloromethane (DCM, ≥99.8%), and Parafilm M® were

obtained from Sigma-Aldrich Company Ltd (Dorset, England). Potassium chloride (99%) was obtained

from Alfa Aesar (Lancashire, England) and silicone MPatch™ microneedle templates were purchased

from Micropoint Technologies Pte Ltd (CleanTech Loop, Singapore). All chemicals purchased were of

analytical grade and used without any further purification.

Microneedle Fabrication: Microneedle templates were cleaned prior to each use by way of

sonication in acetone for 600s and allowed to dry at room temperature. Needle casting involved a

three stage process in which an initial 50 μL aliquot of DCM was introduced into a separate beaker,

along with 50 mg of Pd powder. The Pd suspension was pipetted into the template and allowed to

dry at room temperature - the intention being to precoat the needle surface with a greater quantity

of interfacial Pd and enhance the electroanalytical performance. The second stage casting solution

was prepared by dispersing 50 mg of Pd within 50 µL of a 50% w/v solution of polystyrene in DCM to

form the bulk of the needles and initial baseplate. The final stage involved the introduction of a

further 50 µL of 50% PS/DCM containing 50 mg Pd to form the top of the baseplate backing. The

templates were subjected to centrifugal force at 3000 rpm for 300 s to facilitate the removal of air

bubbles and enable efficient packing of the polymer/particle mixture within the needle structure.

Following this, the remaining DCM was allowed to evaporate at room temperature and pressure for

4 hours prior to demoulding of the microneedle array using adhesive tape. Electrical connection to

the MN array was achieved through using silver conductive paint (RS Components Ltd, Northants

England) to bond a length of silver wire (0.5mm diameter, Goodfellow UK) to the front surface of

one corner of the microneedle array to maximise the usable electrochemical surface area.

In order to improve the electron transfer kinetics of the microneedle array, surface modification by

way of a self-assembling monolayer was employed. Due to the presence of thiol functional groups

which can be readily adsorbed onto metallic surfaces, the array was submerged in a solution of 10

mM L-cysteine for 60 minutes. The array was then rinsed with distilled water prior to any

electrochemical analysis.

Alginate Preparation: Alginic acid sodium salt slowly was added to 0.1 M potassium chloride to

produce a 1.5% w/v viscous solution. The mixture (typically 10mL) was stirred for 4 hours at 45°C

after which a 5 mL solution of 0.2M calcium chloride in deionised water was added drop-wise against

the edge of the beaker until the alginate solution was completely covered. The solutions were

covered with parafilm and left overnight to induce complete cross-linking. In the case of gels

containing the redox probe, solid potassium ferrocyanide was added to prior to adding the calcium

cross linker to provide the required concentration (0.5 mM, 1 mM or 2 mM) based on a 15 mL total

volume) and stirred until dissolution was complete.

Characterisation: The fabricated microneedle arrays were characterised by way of digital optical

microscopy (Nexus Aigo GE-5, Brunel Microscopes Ltd, England), focused ion beam scanning

electron microscopy (Quanta 200 3D FIB/SEM, FEI Company, USA), and electrochemically (PG581

Portable Potentiostat, Bio-Logic SAS, France). Electrochemical measurements were carried out at

room temperature 20oC (+/- 2oC). Microneedles were sputtered under vacuum prior to scanning

electronic microscopy using a 80:20 Pd/Au target at 30mA for 3 minutes (Emitech K500X Sputter

Coater, Quorum Technologies Ltd, England). An accelerating voltage of 5kV was used to obtain the

micrographs.

3.0 Results and Discussion

Scanning electron micrographs of the resulting palladium-polymer microneedles are shown

in Figure 2A-C. The needle array is lifted directly from the mould and, as such, mechanical flexing of

the latter is minimised and, it can be seen, the integrity of the needle structure is preserved upon

removal. While the needles exhibit well defined geometry, there is, however, a degree of distortion

where the sides of the pyramid can be seen to curve inwards. It is likely that this artefact is caused

by shrinkage of the polymer resulting from the gradual removal of the solvent. No difference in

morphology was observed when comparing Pd:Polycarbonate or Pd:polystyrene structures with

both displaying a granular surface highlighted in Figures 2A-C and is attributed, primarily, to the

particulate nature of the Pd component. The effect of the latter is emphasized when compared with

the relatively smooth morphology of MNs composed solely of polymer. One such example is

highlighted in Figure 2D where the homogeneity of a polystyrene casting solution ensures more

efficient filling of the needle void and results in greater definition.

It could be anticipated that the surface roughness observed in Figures 2A-C could be

addressed through increasing the proportion of polymer but, it must be recognised that a

compromise is often required whereby structural improvements come at the cost of conductivity

and poor electroanalytical performance. In this case, the 50:50 ratio proved to be the optimal

arrangement although there are differences in the tip structure of the metal composite and

homopolymer needles with the former being notably blunt in comparison. The mean tip diameter of

the composite was found 7.2 micron compared to 4.4 micron for the polystyrene needles.

Nevertheless, the tip width of the metal composite possesses sufficient needle definition and though

it could be anticipated that the use of smaller metal particles would further improve the capacity for

skin puncture. The Pd powder in this case was obtained from commercial sources and is classified as

submicron but will consist of a spectrum of particle sizes and it can be expected that - were greater

control over the particle size available – the granularity exhibited in Figure 2A-C would diminish. It

should also be noted, however, that this same texture provides an enhanced surface area which

could be analytically useful for both sensing purposes and for the electrochemically controlled

release of reagents.

Figure 2. Scanning electron micrographs of the palladium:polycarbonate microneedle (A-C) and

comparison with a polystyrene homopolymer needle produced from the same mould (D).

Cyclic voltammograms detailing the response of the Pd composite MN towards ferrocyanide

(2 mM, 0.1 M KCl, 50 mV/s) are detailed in Figure 3A. It should be noted that the voltammograms

shown in Figure 3A represent the response of the whole assembly (base plate and needle structures)

submerged in aqueous solution and not in the alginate. It can be seen that the unmodified MN

structure initially exhibits poor electron transfer kinetics with the response to ferrocyanide being

noticeably irreversible with no appreciable reduction process observed. Vagaries in the surface oxide

layer are known to influence the electrochemical performance [23,24] but it has been previously

shown that the electrode response can be markedly improved through modification with thiol

monolayers – in an analogous manner to that employed in gold functionalisation [25,26]. Feliciano-

Ramos et al. in particular, have shown that Palladium readily forms the latter upon immersion in

cysteine solutions [25] and, given the potential biocompatibility of the amino acid modifier, their

approach was adopted in this study. The voltammetric response of the cysteine modified MN array

towards ferrocyanide is compared with the untreated assembly in Figure 3A and it can be seen that

there is a significant improvement. The oxidation process (+0.4 V) is sharper and has shifted

markedly to less positive potentials by 220 mV. The reduction process (+0.11 V) is more defined

though there remains a degree of irreversibility with a peak separation of some 300 mV.

The voltammograms highlighted in Figure 3A represent the response from the entire

composite but it would be expected that in a real world application – only the needle tip would be

within reach of any subcutaneous target analyte. The ability of the needle to puncture the epidermal

barrier and the subsequent response of the MN tips were therefore investigated using a skin mimic

based on the schematic detailed in Figure 1B. In this scenario, the electrochemistry of the

ferrocyanide redox probe would only be accessible once the MN had successfully pierced the

impermeable top layer. A preliminary assessment to ensure that this was the case was conducted

using a smooth Pd film with no MN protrusions. The counter and reference electrodes were placed

directly in the hydrogel layer and the Pd electrode pressed firmly onto the top layer of the

Parafilm™. As expected, the latter isolated the working electrode and no response was observed as

indicated by the flat line in Figure 3B. Replacing the smooth Pd film with the cysteine modified MN

array however, resulted in the puncture of the top layer with the effect that the protruding needle

tips could interrogate the hydrogel. Cyclic voltammograms detailing the response of the needle array

towards various concentrations ferrocyanide (0.5mM, 1mM, 2mM) encapsulated within the alginate

hydrogel described in Section 2.0 are shown in Figure 3B. It can be seen that there is a marked

increase in the peak separation in comparison to the response detailed in Figure 3A. This can be

attributed to a degree of uncompensated resistance within the needle component of the composite

assembly. The current response is also markedly smaller which would be expected given that it is

only the tip of the needle structures which are active rather than the base plate.

Figure 3 A) Cyclic voltammograms detailing the response of the Pd:polycarbonate microneedles to ferrocyanide (2 mM, 0.1 M KCl, 50 mV/s) in aqueous solution before and after modification with cysteine. B) Cyclic voltammogram detailing the response of the microneedle tips after puncturing the parafilm layer (130 m) and accessing the ferrocyanide redox probe encapsulated within a calcium alginate gel. Voltammograms recorded for 0.5mM, 1mM and 2mM ferrocyanide loaded alginate gel containing 0.1M KCl. No response was observed with a smooth Pd electrode placed directly on the skin mimic confirming the puncture capability of the microneedle array.

4.0 Conclusion

There are a multitude of advantages to the adoption of microneedle arrays within clinical diagnostics

but it is clear that the development of such technologies requires the provision of a toolset that is

accessible and which can enable rapid prototyping. The ability to form electroanalytically viable

microneedle arrays through the casting of a metallic powder-polymer mixture has been

demonstrated in principle and it is easy to envisage the further adaptation of the approach through

the substitution of the palladium component with other nano particulate species. The increasing

availability of the latter, the generic nature of the manufacturing step outlined and the ability to

rapidly customise the cast formulations provides a foundation tuning the sensing system to a range

of transdermal sensing applications. There is an inevitable compromise between the ratio of metallic

conductor and polymeric binder but, with more rigorous control over particle size, further

improvements in performance could be achieved. At present, only the working electrode is

composed of microneedles but it could be expected that the same fabrication approach could be

used to produce a separate counter/reference system which could be positioned alongside the

former and provide a complete MN sensing system.

Acknowledgements

The authors thank the Department of Employment and Learning (DEL) Northern Ireland for

supporting this work.

References

1. H.L. Quinn, M-C. Kearney, A.J. Courtenay, M.T.C. McCrudden, R.F. Donnelly. Expert Opinion Drug Delivery, 11 (2014) 1769

2. N.S. Rejinold, J-H Shin, H.Y. Seok, Y.C. Kim. Expert Opinion Drug Delivery, 13 (2016) 1093. S.D. Gittard, A. Ovsianikov, B.N. Chichkov, A. Doraiswamy, R.J. Narayan, Expert Opinion Drug Delivery,

7 (2010) 513–5334. R.G. Pawar, S.D. Pawar, M.V. Gadhave, S.L. Jadhav, D.D. Gaikwad. Journal Drug Delivery &

Therapeutics. 2 (2012) 765. S.J. Lee, H.S. Yoon, X. Xuan, J.Y. Park, S.J. Paik, M.G. Allen, Sensors and Actuators B. 22 (2016) 1144-

1151

6. J.R. Windmiller, G. Valdes-Ramerez, N. Zhou, M. Zhou, P.R. Miller, C. Jin, S.M. Brozik,c R. Polsky, E.Katz, R.Narayan, J.Wang. Electroanalysis 23 (2011) 2302 – 2309

7. S.A.Skoog, P.R.Miller, R.D. Boehm, A.V. Sumant, R. Polsky, R.J. Narayan. Diamond & Related Materials 54 (2015) 39–46

8. C. Barrett, K. Dawson, C. O’Mahony, A. O’Riordan, Journal of Solid State Science and Technology, 4 (2015) S3053-S3058

9. N. Vasylieva, S. Marinesco, D. Barbier, A. Sabac. Biosensors and Bioelectronics. 72 (2015) 148-155 10. P.R.Miller, X.Y. Xiao, I. Brener, D.B. Burckel, R. Narayan, R. Polsky, Advanced Healthcare Materials, 3

(2014) 876-88111. Y.A. Gomaa, L.K. El-Khordagui, M.J. Garland, R.F. Donnelly, F. McInnes, V.M. Meidan. Journal

Pharmacy and Pharmacology, 64 (2012) 1592-1602 12. S. Bystrova, R. Luttge, Microelectronic Engineering, 88 (2010) 1681-1684 13. P. Veerakumar, V. Veeramani, S-M.Chen, R. Madhu,S-B. Liu, ACS Applied Materials and Interfaces, 8

(2016) 1319-132614. B.Thirumalraj, S. Palanisamy,S-H.Chen, D-H.Zhao, Journal Colloid Interfacial Science, 478 (2016) 413-

42015. Y. Peng, J. Ye,L.Zheng,K. Zou, RSC Advances, 6 (2016) 24880-2488816. Y. Jing, X. Yuan,Q. Yuan,K. He, Y.Liu, P. Lu, H.Li, B. Li, H. Zhan, G.Li, Scientific Reports, 6 (2016) AR

2923017. J. Wang, H-B. Sun, H-Y. Pan,Y-Y. Ding,J-Q. Wan,G-H. Wang, M.Han, Sensors. Actuators B., 230(2016)

690-69618. D.Li,K. Ao,Q. Wang,P. Lv,Q. Wei, Molecules 21 (2016) 10.3390/molecules2105061819. S. Chandrasekaran, J.D. Brazzle, A.B. Frazier, Journal Microelectromechanical Systems, 12 (2003) 281-

288.20. W. Luangveera, S. Jiruedee, W. Mama, M. Chiaranairungroj, A. Pimpin, T. Palaga, W. Srituravanich,

Journal Mechanical Behaviour Biomedical Materials, 50 (2015) 77-81 21. J.H. Oh, H.H. Park, K.Y. Do, M. Han, D.H. Hyun, C.G. Kim, C.H. Kim, S.S. Lee, S.J. Hwang, S.C. Shin, C.W.

Cho. European Journal Pharmaceutics and Biopharmaceutics, 69 (2008) 1040-1045 22. C.M. Schoellhammer, D. Blankschtein, R. Langer, Expert Opinion Drug Delivery 11 (2014) 393-40723. G.Macfie, A. Cooper, M.F. Cardosi, Electrochimica Acta 56 (2011) 8394– 840224. J.Sun, C. Zhang, T. Kou, J. Xu, Z. Zhang, ACS Applied. Materials and Interfaces 4 (2012), 6038−604525. G.Macfie, L.V. Simpson, D. McColl,M.F. Cardosi, Journal Physical Chemistry C, 116 (2012) 9930−994126. I.Feliciano-Ramos, M. Caban-Acevedo, M. A.Scibioh, C. R. Cabrera, Journal Electroanalytical

Chemistry., 650 (2010) 98–104