Mechanical Strengthening of Fiberoptic Microneedles Using an Elastomeric Support

Abstract

Background and Objectives:

Microneedles made from silica fiberoptics permit transmission and collection of light, which is an important functional advantage over metal or silicon microneedles. This added functionality may enhance or even enable new percutaneous light-based clinical diagnostic and therapeutic procedures. Micron-diameter fiberoptic microneedles, created from solid fibers capable of light emission and detection, are designed to penetrate several millimeters into tissue while minimizing tissue invasion and disruption. The mechanical strength (critical buckling force) of high aspect ratio (length to diameter) microneedles is a potential problem, which has motivated our invention of an elastomeric support device. In this study, we have tested our hypothesis that embedding the microneedles in an elastomeric support medium may increase microneedle critical buckling force.

Materials and Methods:

The critical buckling force of silica microneedles with 55, 70, and 110 μm diameters and 3 mm lengths were measured with and without a surrounding elastomeric support (PDMS, polydimethylsiloxane). These experimental results were compared to theoretical calculations generated by the Rayleigh–Ritz buckling model. The insertion force required to penetrate ex vivo porcine skin was measured for microneedles with 55 and 70 μm diameters.

Results:

Use of the PDMS support increased critical buckling force for microneedles of 55, 70, and 110 μm diameters by an average of 610%, 290%, and 33%, respectively. Theoretical calculations by the Rayleigh–Ritz model consistently overestimated the experimentally determined strengthening, but correlated highly with the greater enhancement offered to thinner microneedles. Aided by mechanical strengthening, microneedles 55 μm in diameter were able to repeatedly penetrate.

Conclusions:

The critical buckling force of microneedles can be increased substantially to allow extremely high-aspect ratio microneedles, 55–110 μm in diameter and 3 mm in length, to penetrate ex vivo porcine skin. By this strengthening method, the safety and reliability of microneedles in potential clinical applications can be considerably enhanced.

INTRODUCTION

The therapeutic potential presented by microneedles for bypassing the skin’s barrier function has inspired a great amount of research interest. To date, microneedles have been investigated for several skin related clinical applications, such as drug delivery [1–4], blood sampling [5], biopsy extraction [6], and light delivery [7]. Microneedles avoid excessive mechanical damage to the healthy tissue while also mitigating blood loss, tissue disruption, and patient discomfort [8–10]. Despite ongoing research efforts, lack of mechanical strength of microneedles presents an obstacle against widespread clinical adoption. If a microneedle is damaged during insertion, fragments severed from the microneedle can mix into to the blood stream of a patient, which can cause serious complications such as embolism [11]. Milder potential side effects include pain, discomfort, bleeding, and infection.

As a microneedle is being pushed into skin, the axial force on the microneedle, F, starts to increase. At the instant F becomes equal to the threshold force for insertion of skin, F_INS (also called puncture force), the microneedle penetrates. However, if the F_INS is greater than the ultimate mechanical strength of the microneedle, the microneedle fractures before penetrating. Buckling is the common mode of mechanical failure for slender microneedles. Thus, the ultimate mechanical strength of microneedles with slender geometries is determined by the critical buckling force, F_CR. To prevent failure, F_CR of that microneedle should be sufficiently greater than F_INS (F_CR > F = F_INS). Sharper tips on microneedles have been demonstrated to lower F_INS and help facilitate microneedle penetration [1,12,13]. A mosquito’s fascicle (needle), a highly advanced microneedle, was shown to possess an ultra-sharp tip with a diameter <1 μm [14]. This tip diameter was demonstrated to lower F_INS down to an average of 18 μN [15]. Techniques to increase F_CR above F_INS also enable microneedle penetration. By modeling the fascicle as a column and labium as an elastic foundation, Ramasubramanian [14] estimated that the labium sheath increased the critical buckling force, F_CR, of a mosquito’s fascicle by a factor of 5. To facilitate insertion, Khumpuang et al. [11] manufactured an insertion guide to house an array of microneedles.

In a previous paper, we introduced silica fiberoptic microneedles that penetrate ex vivo porcine skin [7]. These microneedles were very slender with 3 mm length and 73–125 μm diameter. The hypothesis of this manuscript is that by providing microneedles with an elastomeric (an elastic polymer) lateral support, an increase in mechanical strength and a reduction of the threshold microneedle diameter necessary to penetrate skin can be attained. To test this theory, we developed a method of embedding silica microneedles in polydimethylsiloxane (PDMS, Sylgard® 184, Dow Corning, Midland, MI), and measured the amount of increase in F_CR obtained. These experimental findings were compared to theoretical calculations. Finally, we inserted elastically supported microneedles 55 and 70 μm in diameter into ex vivo porcine skin to demonstrate the potential of this design for enabling penetration with thinner, less invasive microneedles.

MATERIALS AND METHODS

Manufacturing of Fiberoptic Microneedles

Fiberoptic microneedles were manufactured from multimode silica optical fibers with core/cladding diameters of: 50/55, 50/70, and 100/110 μm (FVP050055065, FIP050070085, and FIP100110125 Polymicro, Phoenix, AZ). To manufacture the microneedles used in the buckling experiments, one end of an optical fiber was flat-polished. The polished fiber was adhesively bonded inside a metal tube (127 μm inner diameter, 236 μm outer diameter) with 3 ± 0.1 mm length extending beyond the tube ending with the flat-polished surface. Microneedles used in the skin penetration experiments were manufactured using the same method, but the fibers were angle-polished at a 25–30° angle using an Ultrapol fiber lensing machine (Ultratec, Santa Ana, CA).

Measurement of Critical Buckling Force of Microneedles

A BOSE™ Electroforce™ 3100 mechanical testing stage was utilized to record the axial force and displacement during failure testing of fiberoptic microneedles. This instrument has a 1.5 μm displacement resolution over its 5 mm range. The accuracy of the force cell was ±55 mN for measurements conducted with 110 μm diameter microneedles. A different force cell with a higher accuracy of ±6 mN was used for measurements with 55 and 70 μm diameter microneedles. Microneedles of each diameter were tested with both unsupported and supported conditions (3 mm length). Each experiment was repeated for N = 5. For supported experiments, the microneedles were embedded inside a 2.5 mm deep layer of PDMS. To manufacture embedded microneedles, liquid PDMS composed of a monomer and a hardener with 10:1 weight ratio was poured into a 10 mm diameter mold around the microneedles. PDMS was hardened at 150°C for 15 minutes. The experimental setup for buckling experiments with support is shown in Figure 1a. For experiments without support, the setup was identical but without PDMS. A bright field microscopy image of a representative microneedle used in buckling experiments is shown in Figure 1b (Leica DM IL LED, Leica Microsystems, Buffalo Grove, IL). The metal tube was guided by a zirconia ferrule with 250 μm inner diameter. For the supported experiments, the metal tube was pushed through the PDMS by the linear actuator of the testing stage with a velocity of 0.1 mm/s. Thus, the microneedles, adhesively bonded to the tube, were pressed against the hard surface of the load cell. Double-sided tape was placed on the load cell surface to prevent the microneedles from sliding. As the actuator moved the tube, the microneedles shortened, deflected laterally (demonstrated by the dashed lines), and finally buckled. Force and displacement data was recorded throughout microneedle contact with the load cell, during bending, and following buckling failure of the microneedle.

Fig. 1.

a: Schematic representation of the setup for buckling experiments. b: A flat-polished microneedle with 3 mm length (1 mm scale bar). Dashed lines demonstrate a deflected microneedle.

Predicting Critical Buckling Force of Microneedles

Microneedles without support.

The critical buckling force, F_CR, of unsupported microneedles can be estimated by Euler’s buckling formula for a cylindrical column:

$$F_{\text{CR}}=\frac{E{\pi }^3{d}^4}{64{\left(KL\right)}^2}$$ (1)

In Equation (1), E = 73 GPa is the elastic modulus of silica, a material property determined by its stiffness [16]. L = 3 mm is the unsupported length of the microneedle, which is the entire microneedle length for this case, d is the diameter of the microneedle, and K is the effective length factor, which is determined by the boundary conditions of the microneedle interface with the substrate (load cell) and the metal tube. Boundary conditions are determined by how rigidly the microneedle is restrained at its ends. Possible boundary conditions for our microneedles include pinned (end cannot deflect but can rotate) or fixed (end can neither deflect nor rotate). K = 1 if both ends are pinned, K = 0.699 if one end is fixed (possibly the base, due to the application of the adhesive) and the other end is pinned (possibly the tip) [17].

Microneedles with elastomeric medium (PDMS) support.

We used the Rayleigh–Ritz approximation method to obtain a theoretical estimation for the F_CR of elastically supported microneedles. A detailed description of the Rayleigh–Ritz method is given in [17,18]. To obtain a Rayleigh–Ritz approximation, a lateral displacement function, $\overline{v}$, for the first mode of buckling of the microneedles with fixed-pinned boundary conditions can be approximated by:

$$\overline{v}=a\left[\frac{x}{L}+1.025\sin \left(\frac{4.493x}{L}\right)\right]$$ (2)

where a is an undetermined coefficient. $\overline{v}$ satisfies the fixed-pinned boundary conditions. Schematic representation of the microneedle with fixed-pinned boundary conditions and the elastic support along 5/6th of its length is given in Figure 2.

Fig. 2.

Schematic representation of an elastically supported microneedle with fixed-pinned boundary conditions.

F_CR of a fixed-pinned microneedle was calculated as:

$$F_{\text{CR}}=\frac{E{\pi }^3{d}^4}{64{\left(0.699L\right)}^2}+0.0782E_{\text{S}}{L}^2$$ (3)

In Equation (3), the first term equals the F_CR of an unsupported fixed-pinned microneedle. The second term estimates the strengthening provided by the elastic support (E_S = 1.8 MPa is elastic modulus of the support medium, PDMS). By following the same procedure, F_CR for a pinned-pinned microneedle was found by assuming the deflected shape function:

$$\overline{v}=a\sin \left(\frac{\pi x}{L}\right)$$ (4)

$\overline{v}$ in Equation (4) satisfies pinned boundary conditions at both ends. F_CR for a pinned-pinned microneedle was found as:

$$F_{\text{CR}}=\frac{E{\pi }^3{d}^4}{64L^2}+0.0984E_S{L}^2$$ (5)

Similar to Equation (3), the first term in Equation (5) equals the F_CR of a pinned-pinned microneedle. The second term estimates the strengthening provided by the elastic support.

Penetrating Ex Vivo Porcine Skin With Microneedles

Ex vivo abdominal pig skin was acquired from a local abattoir. Specimens were cut to 1–2 mm thickness, including both the epidermal and dermal layers. In order to maintain hydration until the experiments were conducted, dissected skin was placed inside a plastic bag between paper cloths saturated with isotonic saline and maintained at 4°C. The high accuracy load cell was used for force (load) measurements with 55 and 70 μm diameter microneedles with and without the elastic support. Skin penetration experiments were not conducted with 110 μm diameter microneedles as microneedles with diameters equal or larger than 73 μm have been previously demonstrated to penetrate skin [7]. A schematic representation of the skin penetration experiments is given in Figure 3a. A bright field microscopy image of a representative microneedle used in the skin penetration experiments is shown in Figure 3b.

Fig. 3.

a: Schematic representation of the setup for skin penetration experiments. b: Angle-polished microneedle with 3 mm length (1 mm scale bar). Dashed lines demonstrate a microneedle that penetrated through the skin.

Prior to experiments, the microneedles were positioned less than 1 mm away from the surface of the ex vivo porcine skin stretched over a U-channel. During the experiments, microneedles were displaced toward the skin’s surface with a velocity of 0.1 mm/s. The penetration experiment was finalized when the microneedle either buckled or penetrated through the skin. Buckling was observed by the instantaneous drop in the force on the microneedle and an audible report. Successful penetration was observed by seeing the tip of the microneedle penetrate through the entire 1–2 mm thickness of the skin (demonstrated by the dashed lines). Following the sequence of penetration experiments, each microneedle was removed from the skin, and examined for damage using both a surgical microscope (Seiler Revelation, St. Louis, MO) and the bright field microscope.

RESULTS

Critical Buckling Force of Microneedles

Representative force versus displacement curves of two buckling experiments conducted with a supported and an unsupported 55 μm diameter microneedle are demonstrated in Figure 4a. F_CR was the peak force measured during the experiments as shown in Figure 4a. Theoretical calculations and experimental measurements for F_CR for unsupported microneedles are given in Figure 4b and Table 1. The ranges of measurements for F_CR were between the fixed-pinned and pinned-pinned calculations (closer to the fixed-pinned calculation). A representative photograph of an unsupported microneedle at the moment of buckling is given in Figure 4b. The deflection shape in the photograph closely resembled the deflection shape for buckling of a fixed-pinned column (inset in Figure 4b).

Fig. 4.

a: Representative force versus displacement curves for a supported and an unsupported 55 μm diameter microneedle. b: Theoretical calculations for F_CR of a pinned-pinned and a fixed-pinned microneedle, and experimentally measured values for 55, 70, and 110 μm diameter microneedles (N = 5). Image shows a photograph of a deflected unsupported microneedle and the theoretical deflection shape. c: Theoretical calculations of F_CR for elastically supported microneedles for pinned-pinned and fixed-pinned boundary conditions, and experimentally measured values for 55, 70, and 110 μm diameter microneedles (N = 5). [Color figure can be seen in the online version of this article, available at http://wileyonlinelibrary.com/journal/lsm]

TABLE 1.

Tabulated Values From Figure 4b,c

	Experimentally measured FCR (mN)
	Minimum	Average ± standard deviation	Maximum	Theory (P-P)	Theory (F-P)
d = 55 μm, no support	59	67 ± 5	72	36	73
d = 70 μm, no support	147	161 ± 9	169	94	193
d = 110 μm, no support	1,053	1,137 ± 51	1,141	575	1,174
d = 55 μm, with support	361	474 ± 112	657	1,630	1,340
d = 70 μm, with support	576	627 ± 84	775	1,688	1,459
d = 110 μm, with support	1,118	1,506 ± 390	2,023	2,443	2,169

Theoretical calculations and experimental measurements for F_CR of supported microneedles are given in Figure 4c and Table 1. The variation in the experimental results was considerable. The experimental results were always less than the theoretical calculations for both pinned-pinned and fixed-pinned boundary conditions. Theoretical calculations and experiment results for the percentage increase in F_CR owing to the support are given in Figure 5 and Table 2. The percentage increase in F_CR was calculated as:

$$\%\text{increase in}{F}_{\text{CR}}=\text{100}\times \frac{F_{\text{CR with support}}-F_{\text{CR no support}}}{F_{\text{CR no support}}}$$ (6)

Fig. 5.

Theoretical calculation and experimentally measured values for percentage increase in F_CR provided by the support.

TABLE 2.

Tabulated Values From Figure 5

	% Increase in FCR
	Minimum	Average ± standard deviation	Maximum
d = 55 μm, with support	439	609 ± 167	881
d = 70 μm, with support	259	291 ± 52	383
d = 110 μm, with support	0	33 ± 34	78

The amount of mechanical strengthening increased as microneedle diameter decreased similar to the theoretical calculation. However, strengthening effect was significantly less than the theoretical calculation as mentioned before. We defined an efficiency coefficient of strengthening, η_STR, as shown in Equation (7). This term is the ratio of the experimentally obtained increase in F_CR to the theoretical calculation.

$${\eta }_{\text{STR}}=\frac{\%\text{increasein}{F}_{\text{CR}}\left(\text{experimental}\right)}{\%\text{increasein}{F}_{\text{CR}}\left(\text{theoretical}\right)}$$ (7)

where F_CR (Theoretical) is the fixed-pinned calculation. η_STR values are listed in Table 3.

TABLE 3.

Efficiency Coefficients of Strengthening

	ηSTR
	Minimum	Average ± standard deviation	Maximum
d = 55 μm, with support	0.25	0.35 ± 0.097	0.51
d = 70 μm, with support	0.39	0.44 ± 0.080	0.58
d = 110 μm, with support	0	0.30 ± 0.31	0.72

Skin Penetration Experiments

A representative force versus displacement curve of a skin penetration experiment is demonstrated in Figure 6a. F_INS was taken as the peak force that occurred during the experiments as shown in Figure 6a. A photograph of a 55 μm diameter microneedle penetrating ex vivo porcine skin is given in Figure 6b. Three 55 and three 70 μm diameter microneedles failed to penetrate ex vivo porcine skin without the mechanical support. These microneedles buckled at the F_CR values given in Figure 7 and Table 4. F_CR values on skin were slightly less than on a hard surface (Fig. 4b and Table 1).

Fig. 6.

a: A representative force measurement curve of a skin penetration experiment showing the maximum force observed by the microneedle during insertion, F_INS (d = 70 μm). b: A successful skin penetration experiment showing a 55 μm diameter microneedle penetrating ex vivo porcine skin.

Fig. 7.

F_CR of unsupported microneedles that buckled during penetration experiments, and F_INS observed during insertion of supported microneedles that successfully penetrated ex vivo porcine skin.

TABLE 4.

Tabulated Values From Figure 7

	Force (mN)
	Minimum	Average ± standard deviation	Maximum
FCR on skin for d = 55 μm, no support	61	66 ± 4	68
FINS for d = 55 μm, with support	111	128 ± 39	176
FCR on skin for d = 70 μm, no support	107	143 ± 32	154
FINS for d = 70 μm, with support	136	179 ± 27	204

Elastically supported 55 and 70 μm diameter microneedles were able to penetrate the ex vivo porcine skin. A 70 μm diameter microneedle penetrated the skin five times without buckling. Three 55 μm diameter microneedles were tested. The first two 55 μm diameter microneedles penetrated twice each, before buckling during the third attempt. The experiment was finalized after obtaining another successful penetration with the third microneedle (total of five successful penetrations out of seven attempts with three 55 μm diameter microneedles). Measurements for F_INS are given in Figure 7 and Table 4.

Tip Damage Determination

Bright field microscopy images of two 55 μm diameter microneedles and a 70 μm diameter microneedle, taken before and after skin insertion testing, are shown in Figure 8. Careful observation of the microscope images did not indicate any damage on these microneedles that successfully penetrated skin.

Fig. 8.

Microneedles before and after insertion (100 μm scale bar).

DISCUSSION

In our earlier studies, we manufactured fiberoptic microneedles with sharp, tapered tips using a melt-drawing technique [7]. We also produced microneedle diffusers that delivered laser energy over millimeter lengths by removing the cladding and some of the core from the end of an optical fiber [19]. In this current study, we have tested our hypothesis that microneedle F_CR may be increased if microneedles are embedded in an elastomeric support medium. To test this hypothesis it was not necessary to create tapered microneedles or optical diffusers. Instead, we manufactured fiberoptic microneedles by simply polishing (flat or at a 25–30° angle) commercially available optical fibers with 55, 70, and 110 μm diameters. This range of diameters matched the microneedles that were manufactured with our earlier methods.

Safety and reliability of microneedle insertion into tissue must be demonstrated before this technology can be translated to clinical applications. In addition to their application specific design requirements, microneedles should possess the required critical buckling force to be inserted into tissue. In this study, we have experimentally demonstrated that slender microneedles can be made stronger by being embedded within an elastomer in a configuration that allows the forward movement of the microneedles during insertion. Buckling experiments with unsupported and supported microneedles were conducted to show the effect of mechanical strengthening on F_CR. The amount of strengthening that was achieved increased significantly as the microneedle diameter decreased (Tables 1 and 2). However, the experimentally measured strengthening effect ranged from 30% to 44% of the theoretically calculated strengthening effect (Table 3).

Experimental results for F_CR were compared to theoretical calculations given by Equation (1) for unsupported microneedles and Equations (3) and (5) for supported microneedles. The buckling shape and the measured F_CR results of the unsupported microneedles agreed closely with the fixed-pinned calculation (Fig. 4b, Table 1). For supported microneedles, the mechanical strengthening effect was experimentally measured to be lower than the theoretical calculations for both pinned-pinned and fixed-pinned conditions (Fig. 4c, Table 1). We attribute this result to the following possibilities (or combinations thereof): (i) The PDMS possibly did not surround the microneedle tightly enough due to air gaps that might have formed between the microneedle and the PDMS. Thus, the reaction force of the PDMS was less than the theoretical support medium assumed in the model. (ii) The assumed value of E_S overpredicted the actual value. (iii) There was uncertainty in the supported length 2.25 ± 0.25 mm due to the manual pouring of the liquid PDMS. The F_CR of supported microneedles varied over a wider range than the unsupported microneedles (Table 1). This variation in F_CR might have also been caused by the uncertainty in the supported length mentioned earlier.

In our earlier studies, we were unable to penetrate ex vivo porcine skin with silica fiberoptic microneedles less than 73 μm diameter (3 mm unsupported length) [7]. Reducing the threshold diameter for penetrating skin with fiberoptic microneedles is challenging, as F_CR is directly proportional to the fourth power of the microneedle diameter. For example, F_CR of a 73 μm diameter microneedle is 210% and 18% greater than F_CR of 55 and 70 μm diameter microneedles respectively. In this study, these thinner microneedles could penetrate into the skin, because mechanical strengthening by the PDMS compensated for the loss in mechanical strength. Average F_INS observed during insertion of 55 and 70 μm diameter microneedles was 128 and 179 mN, respectively. Microneedles that were not supported with PDMS did not possess the required buckling strength to penetrate skin, as the F_CR of unsupported 55 and 70 μm diameter microneedles was 66 and 143 mN, respectively. This lack of buckling strength resulted in the mechanical failure of all unsupported microneedles that were tested for penetrating skin.

Thinner microneedles are not only preferable for decreasing invasiveness, but also may provide beneficial light delivery for some applications. In our earlier work, we observed that fiberoptic microneedles of 33–72 μm diameter (3 mm length) produced a more uniform intensity and temperature distribution compared to microneedles with 97–99 μm diameter and standard multimode optical fiber with 125 μm diameter [19]. Through mechanical strengthening, as outlined in this study, these thinner, diffusing microneedles may become practical to penetrate skin and deliver light directly into sub-dermal regions.

In our previous work, we have observed that sharper tips (d_TIP = 2–6 μm) enhance the penetration characteristics of the fiberoptic microneedles. However, these ultra-sharp tips were more susceptible to tip damage, which can cause medical complications. In this study, mechanical strengthening allowed angle-polished microneedles to penetrate ex vivo porcine skin. No tip damage was observed for the three microneedles shown in Figure 8.

The mechanical strengthening method outlined in this study can be used for other types of microneedles such as hollow microneedles for drug delivery. Microneedles with thinner walls can be used to lower viscous losses and increase drug delivery flow rate without increasing outer diameters and corresponding invasiveness [20]. Microneedles with smaller diameters can reduce blood vessel trauma and pain due to the stimulation of peripheral nerves. With the effect of mechanical strengthening, longer microneedles can be used to possibly reach deeper tissue regions in the skin or in other soft tissues. Mechanical strengthening can increase the safety and reliability of microneedles, thus enabling future clinical applications.

CONCLUSIONS

We have shown that F_CR of microneedles can be increased substantially by embedding them inside an elastomeric medium during insertion. Experimentally obtained increase in F_CR was less than what was theoretically predicted by the Rayleigh–Ritz mathematical model for buckling inside an elastic medium. Increased buckling strength obtained from the elastomeric support allowed extremely slender microneedles of 55 μm diameter and 3 mm length to penetrate ex vivo porcine skin.

NOMENCLATURE

η _STR

efficiency of mechanical strengthening

$\overline{v}$

displacement function

a

undetermined coefficient

d

diameter

d _TIP

tip diameter

E

elastic modulus of silica

E _S

elastic modulus of PDMS

F

force on the microneedle

F _CR

critical buckling force of a microneedle

F _INS

skin’s threshold force for insertion

K

effective length factor

L

unsupported length

N

experiment count