Download as pdf or txt
Download as pdf or txt
You are on page 1of 19

The International Journal of Advanced Manufacturing Technology (2024) 132:533–551

https://doi.org/10.1007/s00170-024-13344-x

ORIGINAL ARTICLE

Design, fabrication, and penetration assessment of polymeric hollow


microneedles with different geometries
Pol Vanwersch1,2 · Tim Evens2 · Albert Van Bael2 · Sylvie Castagne1

Received: 13 September 2023 / Accepted: 22 February 2024 / Published online: 12 March 2024
© The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024

Abstract
Hollow microneedles (HMNs) are minimally invasive needle-like microfeatures usually arranged in arrays designed for drug
delivery and body fluid collection in a painless manner. In a recent work, we demonstrated a novel methodology to mass
produce hollow polymer microneedles at a low cost. This methodology combines ultrashort pulse laser ablation to create
inverse needle shapes in moulds and replication through polymer injection moulding. For a HMN to be functional, it should
effectively pierce the skin at a low force and enable fluidic passage through the skin without leakage. This study investigates
the impact of different laser scanning strategies on the cavity morphology and analyses how the various geometrical charac-
teristics of the needle influence the penetration efficacy. To assess the penetration behaviour of the replicated HMNs, a com-
bination of agarose gel and Parafilm® is employed as an in vitro testing platform. Furthermore, a correlation between HMN
geometry, penetration performance, and modification of polymer material and holding pressure during injection moulding
is established. The results indicate that a certain needle length is essential for effective penetration. Moreover, minimising
the tip area, a factor significantly affecting penetration force, can be achieved by increasing the eccentricity of the scan-free
area and expanding the scanning diameter. However, it is important to consider other functional needle features such as the
ridge height or full lumen, which come at a cost to the tip sharpness. This work highlights the multiple interactions between
the scanning strategy, the injection moulding process parameters, the needle geometry, and the penetration force. This study
provides insights into optimization of the HMN design and the fabrication for enhanced penetration efficacy of functional
injection-moulded polymeric HMNs.

Keywords Hollow microneedles · Injection moulding · Ultrashort pulse laser ablation · Penetration force · Artificial skin

1 Introduction techniques. For example, selective etching can be used to


create silicon HMNs [4], while micromilling can be used
Hollow microneedles (HMNs) are minimally invasive nee- to produce stainless steel HMNs [5]. Ceramic HMNs can
dle-like microfeatures usually arranged in arrays designed also be made using ceramic casting [6]. While enabling a
to pierce the external skin layer, for the delivery of drugs wide variety of needle designs, these flexible manufactur-
to the body or the collection of body fluid. They are gain- ing techniques are not suitable for mass manufacturing. In
ing popularity as, among else, vaccine delivery systems [1] a recent work, we demonstrated a methodology for mass
and interstitial fluid extractors [2]. Compared to traditional manufacturing polymeric HMNs using injection moulding
(hypodermic) needles, HMNs can be used for self-medi- with laser-ablated moulds [7]. In this process, the inverse
cation and cause no pain thanks to their short length [3]. of the needle shape is created in the mould with an ultra-
HMNs can be manufactured using a variety of materials and short pulse laser using a cross-hatching scanning strategy,
and the replication of the needles is achieved with a con-
* Pol Vanwersch ventional injection moulding machine. For these HMNs to
pol.vanwersch@kuleuven.be be effective, it is important that they penetrate easily into
the patient’s skin. HMNs can be found in a wide variety of
1
Department of Mechanical Engineering and Flanders shapes and sizes, but previous studies suggest that the tip
Make@KU Leuven ‑ M&A, KU Leuven, Leuven, Belgium
area and wall angle are the most influential geometrical
2
Department of Materials Engineering, Diepenbeek Campus, parameters for the penetration force in skin [8]. This can be
KU Leuven, Diepenbeek, Belgium

Vol.:(0123456789)
534 The International Journal of Advanced Manufacturing Technology (2024) 132:533–551

tested with artificial or ex vivo skin models, or in a digital 2 Materials and methods
environment (i.e. in silico). Penetration tests on artificial
skin models are typically performed with multiple layers of 2.1 Materials
Parafilm® [9, 10], or with a juxtaposition of polyurethane
foil and agarose gel [11]. In vitro tests are highly repeat- Two thermoplastic materials are used in this study: a poly-
able and controllable but can lead to a different penetration carbonate (Lexan™ HP1HFEU, Sabic, Saudi Arabia) and a
behaviour than real-life applications. On the other hand, glass fibre–reinforced polypropylene (Fibremod™ GB311U,
ex vivo models, which can range from porcine skin [12] to Borealis, Austria). The polycarbonate material is a biocom-
rat skin with a supporting PDMS layer [13], can have low patible grade with high stiffness (tensile modulus, 1 mm/
repeatability because of the high variability in skin tissues, min, 2300 MPa). The glass fibre–reinforced polypropyl-
but they can be useful as a final validation tool for penetra- ene is chosen for its combination of high flowability (MFR
tion. The human skin is the most suitable for penetration 230 °C/2.16 kg, 2.0 g/10 min) and stiffness (tensile modu-
testing. However, its use is subject to strict regulatory rules lus, 1 mm/min, 7000 MPa). The mould material is a low
[14]. In silico models are easily tuneable but are highly corrosion tool steel Stavax (grade 1.2083–AISI 420–DIN
dependent on the boundary conditions and can require high X42Cr13), which is widely used in the mould industry
computational power [15]. From the available literature, it because of its high corrosion and wear resistance.
is clear that the shape of a HMN has a large influence on its
piercing behaviour and that the morphology of injection- 2.2 Methodology
moulded HMNs is impacted by both laser scanning strategy
and injection moulding process parameters. However, to 2.2.1 Laser scanning strategy
the best of the authors’ knowledge, there exists no report of
functional hollow microneedle geometry that can be fab- The microneedle cavities in the mould are fabricated with
ricated with injection moulding and laser-ablated moulds. a Lasea LS5 machine equipped with a femtosecond laser
In this work, different HMNs are fabricated with injection source (Satsuma HP, Amplitude Systemes, France). The
moulding and their penetration behaviour is analysed using laser emits pulses of 250 fs with a maximum power on target
a combination of agarose gel and Parafilm®. Firstly, the of 7.7 W. The focal length of the focusing lens is 100 mm.
influence of the laser scanning strategy on the cavity mor- The Gaussian beam has a beam focus radius of 8.2 µm [16].
phology is studied. Secondly, the influence of the polymeric The pulse repetition rate is set to 500 kHz, and the pulse
needle geometry on the penetration behaviour is considered. energy is 15.4 µJ, corresponding to an average fluence of
Finally, a link is made between the replication fidelity and 7.3 J/cm2. A cross-hatching laser strategy with a circular
the penetration behaviour by modifying the holding pres- and oval scanning area is used for the creation of the mould
sure during injection moulding and by modifying the poly- cavities, as shown in Fig. 1. In these strategies, the laser
mer material. covers the scanning area (represented by the dotted line) in

Fig. 1  Cross-hatching laser strategies for a circular and b oval base shapes
The International Journal of Advanced Manufacturing Technology (2024) 132:533–551 535

parallel lines, and the direction of the lines is changed by 90° 2, the injection moulding holding pressure is limited by the
at every layer. The scan-free area (SFA) is the area where no needle deformations occurring for inserts with high numbers
laser pulses are delivered. This creates a pillar in the HMN of needles. Therefore, a selection of 6 HMN geometries is
cavity, which ensures that the replicated microneedles are made for set 3, where the holding pressure can be increased.
hollow. The hatch pitch is the distance between two parallel The selection of the HMN geometries is based on the results
lines, and the layer pitch is the downward movement of the of the penetration tests of set 2. These six selected laser strat-
laser beam focus after every layer. The eccentricity is the egies are repeated on another mould insert, then replicated
centre-to-centre distance between the SFA and the scanning with PC at high holding pressure and glass fibre–reinforced
area. polypropylene (GFR-PP), which has high flowability and
stiffness. Additionally, penetration tests are performed with
2.2.2 Laser machining experiments and µ‑CT scanning these six replicated HMNs in the artificial skin model.
A schematic representation of the laser strategies used
Three sets of experiments are performed in order to gain in this study is shown in Fig. 3. The scanning speed, hatch
insights in different aspects influencing the hollow micronee- pitch, and layer pitch are kept constant for all experiments
dle geometry and to design hollow microneedles suited for at 100 mm/s, 15 µm and 2 µm, respectively. In the first set
penetration and intradermal fluid injection or extraction. of experiments, six variations of the shape and dimensions
A schematic representation of the overall methodology is of the scan-free area are produced, each repeated four times
shown in Fig. 2. The first set of experiments is designed to for repeatability assessment. Set 1 consists exclusively of
study the influence of the shape and size of the scan-free circular base shapes. The position of the different SFAs is
area on the HMN geometry and comprises six laser scanning given in Fig. 3a by the shortest distance between the edge
strategies with various scan-free area geometries. Those of the SFA and the scanning centre. In the second set of
lasered cavities are analysed with micro-computed tomog- experiments, two different base geometries are used, being
raphy (µ-CT), and the outcome is used to design the second circular (Fig. 1a) and oval (Fig. 1b), both including a circular
set of experiments. Set 2 is designed to study the influence scan-free area. The oval shape consists of a half circle with
of various laser scanning parameters on the HMN geom- radius 1, and a half ellipse with radius2 = radius1 × 1.2 on
etry. These HMN cavities are analysed with µ-CT, and the the side of the SFA. The other varied scanning parameters
cavities with full lumen-forming pillars (i.e. the top of the are the diameter, eccentricity, SFA diameter, and number
lumen-forming pillar is in plane with the mould surface) are of layers to create cavities with every possible combination
repeated on a separate mould insert for replication with PC of these parameters, with the exception of the combina-
and penetration analysis with an artificial skin model. In set tion of scanning diameter and SFA diameter of 400 µm and

Fig. 2  Schematic representation


of the methodology
536 The International Journal of Advanced Manufacturing Technology (2024) 132:533–551

Fig. 3  Overview of the laser scanning strategies for set 1 (a), set 2 (b), and set 3 (c)

150 µm, respectively. In total, the second set of experiments the lumen-forming pillar. A significant depth of the lumen-
comprises 90 different HMN cavities. The laser strategies forming pillar may lead to needle clogging, rendering it non-
with the combination of diameter 1 = 400 µm and SFA diam- functional. The ridge-forming trench is the ablated material
eter = 150 µm were omitted due to the large proportion of between the lumen-forming pillar and the outer edge of the
scan-free area to ablation area. In the third set of experi- cavity.
ments, six selected scanning strategies from set 2 (S1–S6)
were lasered in 2 × 2 arrays in a new mould insert, compris- 2.2.3 Replication with injection moulding
ing a total of 24 HMN cavities on the mould insert.
For µ-CT analysis, the lasered inserts are cut with The injection-moulded part selected for this study is a
wire-electrical discharge machining (EDM) into strips of 55 × 60 × 1 mm flat plate. The polymer is injected through a
4 × 4 × 50 mm and analysed with a Phoenix Nanotom micro- hot runner located 17 mm from the edge of the product. The
computed tomography (µ-CT) system. Each scan consists lasered microneedle cavities are located on a 16 × 16 mm
of 2400 images and is taken with the “fast-scan” function insert, which is centred horizontally on the product and posi-
active. The acquisition parameters are as follows: cur- tioned 3 mm from the edge, opposite to the gate, as depicted
rent = 80 μA, voltage = 120 V, and voxel size = 4 µm. GE in Fig. 5.
phoenix datos|x REC is used as reconstruction software to The injection moulding experiments are conducted on a
generate slices, and the slices are analysed with Fiji ImageJ. hydraulic injection machine (ES 200/35 HL, Engel, Austria)
A typical slice through the centre of a HMN cavity is shown with a maximum clamping force of 350 kN and a 25-mm
in Fig. 4a, depicting the cavity depth, the depth of the lumen- horizontal screw. The cooling time is kept at 15 s for all
forming pillar, and the depth of the ridge-forming trench. experiments. The barrel temperature is set to the highest
The orientation of the slice is highlighted in Fig. 4b. The recommended value provided by the polymer manufacturer.
depth of the lumen-forming pillar is defined as the distance The volumetric injection rate is set to the highest achiev-
between the mould surface and the top of the standing pillar able value of the injection moulding machine. The holding
resulting from the SFA. To ensure the functionality of hol- pressure and temperature are varied throughout injection
low microneedles, it is essential to minimise the depth of moulding experiments to find the highest possible value,
The International Journal of Advanced Manufacturing Technology (2024) 132:533–551 537

Fig. 6  Deformed hollow microneedle

the needle base. The injection moulding process parameters


are shown in Table 1.
For the geometrical assessment of the needles, the poly-
mer plates are cut into 2.5 × 2.5 mm squares with nine nee-
dles each. From these nine needles, all but the central one
Fig. 4  a Central µ-CT slice of a hollow microneedle cavity. b Isomet- are scraped off the plate using a surgical knife, resulting in
ric view of a cavity with highlighted orientation of the µ-CT slice a single needle surrounded by around 1 mm of base plate on
either side. In this way, the central needle is not damaged by
the cutting process. The isolated needles are then analysed
with a digital microscope (VH-S30, Keyence, Japan), and
the needle length, tip wall thickness, tip radius, ridge height,
and tip angle are measured with the built-in measurement
software. Figure 7 shows different views of a replicated
HMN.

2.2.4 Penetration tests

The penetration tests are designed to gain insights into the


mechanisms of skin piercing and to compare the different
HMN geometries to each other. As a skin surrogate, a
30 × 30 × 20 mm block of 1 m/v% agarose gel covered by
one layer of Parafilm® is used in this comparative study
because of its high repeatability and similitude to human
skin [18]. In this skin model, which is an adaptation of
the skin model used by Koelmans et al. [11], the Para-
Fig. 5  Schematic representation of a the top and b the side view of film® layer acts as the stratum corneum, and the agarose
the mould cavity. All units are millimetre
gel acts as the dermis. While it is crucial to acknowledge
that the measured penetration forces using the artificial
without causing injection moulding defects. An increased skin model are anticipated to deviate from those in vivo
mould temperature will delay the formation of the poly- in human skin, it is important to clarify that the intention
mer solidification, which improves the replication fidelity is not to replicate human skin. The purpose of employing
[17]. However, an excessive mould temperature can lead to an artificial skin model is solely to facilitate a compara-
defects such as needle deformation (Fig. 6), where the needle tive analysis of the puncture behaviour exhibited by vari-
is stretched, characterised by a decrease in diameter close to ous needles, allowing for the identification of key factors
538 The International Journal of Advanced Manufacturing Technology (2024) 132:533–551

Table 1  Injection moulding Set 1; PC Set 2; PC Set 3; PC Set 3; GFR-PP


parameters for the replication
of HMNs Barrel temperature (°C) 315 315 315 260
Volumetric injection rate ­(cm3/s) 149 147 147 147
Holding pressure (bar) 749 316 633 550
Mould temperature (°C) 115 115 115 110

Fig. 7  Replicated HMN. a 3D-view of the needle on its base plate. b 3D-view for tip wall thickness measurement. c Front view. d Side view

influencing the penetration force of HMNs. For the pen- The penetration testing procedure is as follows:
etration tests, a motorised tension and compression tester
(BZ2.5/TS1S, Zwick, Germany) equipped with a 500-N 1. Place the (artificial) skin model underneath the down-
load cell is used. The agarose gel was laid in a 3D-printed ward facing needle.
case and the Parafilm® layer is maintained in place with 2. Reduce the distance between the needle and the top
two clamps. For every needle, five penetration tests are surface of the artificial skin at 1 mm/min until contact
conducted, each on the same set of agarose gel and Para- (threshold of 0.01 N).
film®, but at different locations. One combination of aga- 3. Once contact is made, lower the head at a speed of
rose gel and Parafilm® is used for approximately four nee- 2 mm/min until 0.4 N is reached while recording the
dles in total, which corresponds to 20 penetration cycles. force and displacement every 1 µs.
To compensate for the variability in the different sets of 4. Retract the head to the start position (i.e. above the arti-
agarose gel and Parafilm® layer due to different tension ficial skin).
in the Parafilm® layer or different humidity levels in the
agarose gel, the force measurements are recorded relative The low insertion speed of 2 mm/min is chosen experi-
to the penetration force of one needle geometry, which is mentally to ensure sufficient resolution of the force–dis-
used on every artificial skin set. The penetration tests are placement measurements. After conducting the penetra-
also performed with commercially available 34G needles, tion tests, the needles were visually inspected using a
which have been reported to penetrate human abdominal digital microscope. None of the needles exhibited any
skin at 0.17 N [19], to assess the similitude between arti- evidence of deformation following the penetration tests.
ficial and human skin.
The International Journal of Advanced Manufacturing Technology (2024) 132:533–551 539

3 Results and discussion

3.1 Set 1: variation of the SFA shape and size

Figure 8 shows the cavity depth, the depth of the ridge-


forming trench, and the depth of the lumen-forming pillar
(see Fig. 4) for the six respective laser strategies. It is
clear that strategies with a smaller distance between the
SFA and the scanning centre result in shallower cavities.
This is a direct consequence of the reduced number of
pulses towards the centre of the cavity, leading to less
ablation. Also, the depth of the lumen-forming pillar,
which should be minimised to prevent needle clogging,
is influenced by the size of the SFA. Figure 9 shows the
depth of the lumen-forming pillar in function of the diam-
eter of the incircle (i.e. the largest circle fitting inside the
Fig. 9  Influence of the incircle diameter on the depth of the lumen-
SFA geometry).
forming pillar. The error bars represent the standard deviation on the
It is clear that smaller incircles, regardless of the shape depth
of the SFA, result in larger depths of the lumen-forming
pillars. The laser ablation of the top of the lumen-form-
ing pillar is due to the fact that the laser focus is lowered The depth of the ridge-forming trench follows a simi-
after every scanned layer. In doing so, the beam is focused lar trend to the cavity depth, increasing with smaller SFAs
towards the bottom of the cavity, but defocused on the top (Fig. 8). This can be explained by the fact that a smaller
surface of the mould. The interaction between the defo- SFA leads to more energy deposition to the material, which
cused beam and the lumen-forming pillar leads to ablation results in a deeper cavity and ridge-forming trench. When
of the latter. By increasing the incircle of the SFA, the only looking at the ridge-forming trench, it can be con-
accumulated fluence on the centre of the lumen-forming cluded that smaller SFAs are beneficial for the HMN geom-
pillar is lowered. This graph also shows a saturation phe- etry because they lead to deeper ridge-forming trenches,
nomenon at 80 µm, where the depth of the lumen-forming which will prevent the replicated needle from leaking when
pillar is close to 0 µm. Above 80 µm, an increase of the inserted into the skin. However, larger SFAs are desired to
incircle diameter does not lead to a decrease of the depth obtain full lumen-forming pillars. Therefore, the optimal
of the lumen-forming pillar. It is expected that this satu- HMN strategy is expected be a compromise between low
ration point is dependent on the number of layers, as this depth of the lumen-forming pillar and large ridge-forming
influences the defocusing of the laser beam. trench.

Fig. 8  Influence of the SFA shape and size on a the cavity depth, b the depth of the ridge-forming trench, and c the depth of the lumen-forming
pillar. No ridge-forming trench could be found for the scanning strategy on the right. The error bars represent the standard deviation
540 The International Journal of Advanced Manufacturing Technology (2024) 132:533–551

3.2 Set 2: variation of different laser scanning of layers. This is caused by the fact that deeper cavities have
strategy parameters and penetration tests steeper walls, resulting in (i) the distribution of the laser
spot over a larger surface, leading to a minimised ablation,
3.2.1 Variation of the base geometry, SFA size and position, and (ii) a larger portion of the scanning area being out of
scanning diameter, and number of layers focus. These two phenomena result in a decrease of the
ablation efficiency for high aspect ratio cavities. A similar
In this set of experiments, a number of laser scanning param- trend has also been reported by Evens et al. [20] for solid
eters are independently varied, and their influence on the microneedles. There is no noticeable difference in cavity
cavity shape and HMN penetration performance is analysed. depth between oval and circular shapes. Indeed, the two laser
Examples of cross-sectional views obtained from the µ-CT strategies are designed in such a way that the central abla-
measurements at the centre of the HMN cavities are shown tion zone (read: the scanning zone on the right of the SFA
in Fig. 10. Cone-shaped hollow microneedle cavities with on Fig. 1) is similar for oval and circular cavities. Figure 11a
very small tip radii (a few micrometres) are observed for all also shows similar cavity depths for different SFA diameters.
sets of laser parameters. However, the shape and dimensions This shows that the influence of the SFA diameter on the
vary between the different microneedle cavities. The influ- cavity depth is negligible in the tested range.
ence of the laser strategy parameters on the geometry of the The depth of the ridge-forming trench is generally
HMN cavities is discussed in the following section. increased by a higher number of layers, with a similar satura-
Figures 11, 12, and 13 show the evolution of the cavity tion effect as for the cavity depth. The ridge-forming trench
depth, the depth of the ridge-forming trench, and the depth is deeper for oval base shapes than for circular ones, which
of the lumen-forming pillar for circular and oval laser strate- is expected since oval base shapes have a larger scanning
gies with small and large SFA for varying number of lay- area between the SFA and the edge of the scanning diameter.
ers, eccentricity, and scanning diameter, respectively. Each Also, the saturation effect is larger for circular base shapes
datapoint represents the average value over the numbers of than for the oval ones. This is due to the smaller width of
layers, eccentricities, or scanning diameters. the ridge-forming trench in the circular strategy, leading to
The cavity depth is generally increased by an increasing trenches with steeper walls and higher aspect ratios. The
number of layers due to the higher accumulated fluence. depth of the ridge-forming trench is larger for the small
However, the increase in depth is lower at higher numbers scan-free areas (100 µm) than for the large SFA (150 µm).

Fig. 10  Influence of the laser scanning parameters on the cavity shape. Example starting with a circular base shape with the independent varia-
tion of each of the scanning strategy parameters. The µ-CT slices are taken at the centre of the HMN cavities
The International Journal of Advanced Manufacturing Technology (2024) 132:533–551 541

Fig. 11  Influence of the number of layers on a the cavity depth, b depth of the ridge-forming trench, and c depth of the lumen-forming pillar

This shows that the SFA diameter has a larger influence on Figure 12a shows that a larger eccentricity results in a
the ridge-forming trench than on the cavity depth. In fact, the deeper cavity. This is expected, since a larger eccentricity
relative increase in SFA diameter over scanning area diam- means that more energy is delivered towards the centre of
eter is larger on the side of the ridge (left of the SFA) than the cavity, which increases the needle depth. Overall, the
on the central side (right of the SFA), leading to a higher circular strategy results in slightly deeper cavities than the
influence on the ridge-forming trench. oval one. On average, the difference in cavity depth between
The depth of the lumen-forming pillar is, as was men- circular and oval strategies equals to 68 µm, which corre-
tioned in Sect. 3.1, mainly influenced by the diameter of sponds to only 4% of the average cavity depth and can thus
the SFA. Figure 11c shows that large scan-free areas result be neglected. In Fig. 12b, it is clear that the eccentricity has
in full lumen-forming pillars, with depths equal or close to a large influence on the depth of the ridge-forming trench.
zero. This means that the top of the lumen-forming pillar is Namely, the larger the eccentricity, the smaller the ridge-
in plane with the top surface of the mould insert. Concerning forming trench. This is a direct result of the reduction of
the small SFA, an increase in number of layers results in a scanning area between the SFA and the edge of the scan-
deeper lumen-forming pillar. This is caused by the fact that ning zone. Figure 12c shows the difference in depth of the
the focus point of the laser is lowered at every layer, which lumen-forming pillar between small and large SFA, as dis-
means that the beam is increasingly defocused on the top cussed for Fig. 11c. Besides this, it does not show a particu-
surface. Hence, its ablation area increases and it can ablate lar trend for the influence of the eccentricity on the depth of
further away from the beam centre, leading to more ablation the lumen-forming pillar. The relatively small differences at
of the lumen-forming pillar. The depth of the lumen-forming SFA 100 µm can be attributed to material impurities, laser
pillar is larger for the oval laser strategy than for the circular scanning variability, and measurement uncertainties.
one. This is a result of the enlarged ablation zone in the oval Figure 13 shows the influence of the scanning diameter on
strategy leading to more pulses around the SFA which, when the cavity depth, the depth of the ridge-forming trench, and
defocused, ablate the lumen-forming pillar. the depth of the lumen-forming pillar for circular and oval
542 The International Journal of Advanced Manufacturing Technology (2024) 132:533–551

Fig. 12  Influence of the eccentricity on a the cavity depth, b depth of the ridge-forming trench, and c depth of the lumen-forming pillar

laser strategies with small and large SFA. No datapoint can This negative dynamic effect can be compensated in the
be found for large scan-free areas and a scanning diameter laser controller software [21]. However, it can be assumed
of 400 µm because this combination of parameters was ruled that this compensation was not seamless.
out of the experiments due to its high SFA to diameter ratio Figure 14 shows the influence of the number of layers
(see Sect. 2.2.2). Each datapoint represents the average value on the cavity diameter for circle and oval base shapes with
over the different numbers of layers and eccentricities. small and large SFA and various scanning diameters. Each
Figure 13a shows that the cavity depth increases with an datapoint represents the average value over the different SFA
increasing scanning diameter. This is a result of the fact that diameters and eccentricities.
wider cavities have less steep walls, resulting in less reflec- It is clear that an increasing number of layers gener-
tion and a lower dispersion of the laser energy. Hence, wider ally leads to a larger cavity diameter. As explained in
cavities absorb more laser energy, which increases their Sect. 2.2.1, every layer is followed by a downward move-
depth. This was also noticed by Evens et al. [20] for solid ment of the laser focus corresponding to the layer pitch.
microneedle cavities. Also, the depth of the ridge-forming This results in an approximately focused beam towards
trench (Fig. 13b) increases with an increasing scanning the centre of the cavity, but defocused beam on the cav-
diameter. This is also a result of the lower energy absorp- ity edges, where the beam interacts with the top surface
tion for steeper slopes. Figure 13c shows that the depth of of the mould insert. If the laser fluence of the defocused
the lumen-forming pillar decreases for increasing scanning beam surpasses the ablation threshold on the top surface,
diameters for the small SFA. This is possibly caused by inac- ablation occurs and the cavity is widened. Figure 14 also
curacies of the laser scanner, which accelerates and deceler- shows that the cavity diameter, even at 150 layers, is larger
ates at the beginning and end of every scanning line, leading than the programmed scanning diameter. Indeed, with the
to discrepancies in the pulse spacing. Since at small scan- focused spot size being around 20 µm, the effective cavity
ning diameters, the scanner quickly accelerates then deceler- diameter is expected to always be at least 20 µm larger
ates, the pulse deposition accuracy is more largely affected. than the scanning diameter, which corresponds to the
The International Journal of Advanced Manufacturing Technology (2024) 132:533–551 543

Fig. 13  Influence of the scanning diameter on a the cavity depth, b depth of the ridge-forming trench, and c depth of the lumen-forming pillar

Fig. 14  Influence of the number


of layers on the cavity diameter.
For oval base shapes, the pro-
grammed short axes are 400,
500, and 600 µm, and their cor-
responding long axes are 440,
550, and 660 µm, respectively

measured values. It is worth highlighting that, in accord- 3.2.2 Replication


ance with the defined parameters of the oval laser strat-
egy, the length of the oval’s long axis consistently exceeds The analysis of the µ-CT slices showed that the top surface
that of its short axis. Finally, the diameter of the circular of the lumen-forming pillar of some HMN cavities is located
cavities and the short axis of the oval cavities are nearly under the top surface of the mould insert (see Fig. 10). In
indistinguishable, which is an obvious result of the design order to ensure optimal functionality of the HMNs, it is
of the laser strategies. crucial that the top of the lumen-forming pillar is in plane
544 The International Journal of Advanced Manufacturing Technology (2024) 132:533–551

with the surface of the insert. Out of the 90 lasered HMN The downward trend of the replication fidelity of the
cavities, 21 cavities with circular scanning shape and 38 ridge shows that the position of the SFA influences the fill-
cavities with oval scanning shape fulfil this requirement. The ing behaviour of the ridge-forming trench. Indeed, a higher
59 (21 circular + 38 oval) corresponding laser strategies are eccentricity leads to a narrower ridge with faster cooling of
selected for the replication phase and subsequent penetration the polymer. The overall replication fidelity, however, is not
tests. The overall replication fidelity, defined as the repli- impacted by the eccentricity, as the displacement of the SFA
cated needle length divided by the cavity depth, is on aver- is minimal compared to the width of the main cavity.
age 82 ± 7%. The replication fidelity of the ridge-forming
trench is on average 48 ± 20%, which is considerably lower 3.2.3 Penetration
than the overall replication fidelity. This difference in rep-
lication fidelity is caused by the fact that the aspect ratio of Throughout the penetration tests, both successful and unsuc-
the ridge-forming trench is higher than that of the overall cessful penetrations were observed. The penetration event is
cavity. This increased surface/volume ratio leads to a faster defined by a sudden decrease in force, as described in [23,
freezing of the polymer and thus a lower replication fidelity. 24]. Typical force–displacement graphs for HMNs with suc-
A similar relationship between aspect ratio and replication cessful and unsuccessful penetration are shown in Fig. 16.
fidelity was noted by Evens et al. [22]. Figure 15 shows the For successful penetrations, the penetration graph can
influence of the eccentricity on the overall replication fidel- be divided into three phases: (I) the force build up at the
ity as well as the replication fidelity of the ridge-forming needle tip, (II) the penetration event, which is characterised
trench. Each datapoint represents the average value over the by a sudden drop in force or a plateau, and (III) the second
different scanning diameters, scanning shapes, numbers of force build up due to the contact between the base plate and
layers, and SFA diameters. the skin model. The penetration force is determined as the
maximum force reached at the junction of phase I and II. For
unsuccessful penetrations, the base plate comes in contact
with the skin model before the penetration event, and only
two phases can be discerned: (I) the force build up at the
needle tip and (II) the second force build up due to the con-
tact between the base plate and the skin model.
Figure 17 shows the penetration success rate in the artifi-
cial skin model for every replicated circular and oval needle
in function of its length. Every needle is tested five times
for penetration on the artificial skin model. A penetration
success rate of 1 corresponds to five successful penetrations
out of five.
It is clear that the length of a needle has a large influence
on its penetration success rate. Figure 17 shows that needles
shorter than 1095 µm do not penetrate the skin model, while
needles above this threshold have varying penetration suc-
cess. This is linked to the fact that for short needles, the force
Fig. 15  Influence of the eccentricity on the replication fidelity of the
main cavity and ridge. The error bars represent the standard devia- at the needle tip cannot reach a sufficiently high level before
tions contact between the base plate and the artificial skin is made.

Fig. 16  Force–displacement
graphs for a successful and b
unsuccessful penetration of
a HMN in the artificial skin
model
The International Journal of Advanced Manufacturing Technology (2024) 132:533–551 545

In the following analysis, only the needles with length higher


than 1095 µm are depicted. The average penetration force
of the oval and circular HMNs in function of the tip angle is
shown in Fig. 18, together with the average penetration force
of the 34G needle with identical skin model.
This figure shows that when successful penetration
occurs, the tested needles require forces between 0.17 and
0.30 N for penetration. Also, a larger tip angle generally
results in a higher penetration force. Indeed, a low tip angle
results in highly localised pressure on the skin, leading to
high stress and faster penetration. In Fig. 19, the side views
of three different HMN geometries are shown. Figure 19a
and b clearly show the differences between HMNs with
low and high tip angles, respectively. Figure 19c depicts a
HMN with a low tip angle, yet without penetration success
Fig. 17  Influence of the needle length on the penetration success rate caused by the particular tip geometry. While the tip angle,
as depicted by the white dotted lines, is relatively small,
the needle surface in contact with the skin model increases
rapidly during the penetration test, leading to a decrease in
localised pressure and thus an unsuccessful penetration.
When comparing the penetration results from the injec-
tion-moulded HMNs with the commercially available 34G
needle (Terumo, Japan), it can be noticed that the injection-
moulded HMNs require 3.5 to six times more force to pen-
etrate the same skin model. However, it is important to note
that the 34G needles used in this study are stainless steel
hypodermic needles. They have a different bevel geometry
than the injection-moulded HMNs and have a length of
4 mm. While their penetration force is low, their relatively
long length makes them unsuitable for intradermal injection.
In Fig. 20, only the datapoints with successful penetration
are analysed, and the influence of the tip radius and tip wall
Fig. 18  Influence of the tip angle on the penetration force. The error
bars and the shaded area represent the standard deviations on the thickness on the penetration force are studied.
penetration force. The dashed line represents the linear regression of Figure 20a shows that a larger tip radius typically leads
the datapoints with successful penetration and R is the linear correla- to an increase in penetration force due to an increase in con-
tion coefficient. Datapoints with a penetration force of 0 N represent tact area between the needle and skin model. Figure 20b,
unsuccessful penetrations

Fig. 19  Side view of three


different HMN geometries. a
Relatively small tip angle (30°)
and low penetration force (0.19
N). b Large tip angle (43°)
and no penetration success. c
Relatively low tip angle (36°)
and no penetration success due
to its particular tip geometry.
The dotted lines represent the
tip angles
546 The International Journal of Advanced Manufacturing Technology (2024) 132:533–551

however, does not show any significant relationship between case of an increase in number of layers, it could be countered
the tip wall thickness and penetration force within the tested by an increase in scanning diameter and SFA diameter.
range. This lack of trend is unexpected, since the tip wall It is important to note that the interactions between the
thickness has a direct impact on the surface in contact with laser scanning parameters, injection moulding process, and
the artificial skin during penetration. However, it could be penetration force are multiple. An example is the two-way
explained by the relatively short range of tip wall thicknesses impact of the eccentricity on the penetration force. Firstly,
measured (34–62 µm). In the following section, it can be a high eccentricity reduces the tip angle, therefore reducing
seen that the tip wall thickness can be influential for the the area of contact between needle and skin and leading
penetration force when the range of tip wall thicknesses is to a lower penetration force. Secondly, higher eccentricity
increased. means more ablation towards the centre of the cavity, which
The penetration results show that a minimum needle eventually gives more room for the injected polymer to flow
length is required for successful penetration of the HMNs, towards the bottom of the cavity and thus form a sharper tip.
and Sect. 2.2.1 shows that the depth of the HMN cavity is Based on the above results, a selection of six HMNs
positively influenced by the number of layers, eccentricity, is made for replication with improved injection moulding
and scanning diameter. Another determining factor for the process (i.e., higher holding pressure) and further penetra-
penetration force is the tip angle. The tip angle can be cor- tion tests. As mentioned in Sect. 2.2.2, the lower number
related to the difference between the needle length and the of cavities on the mould insert of set 3 enables the use of
ridge height. The ridge height is a result of the depth of higher holding pressure without needle deformations. The
the ridge-forming trench and the replication fidelity of the HMN selection includes one circular and one oval needle
ridge. The depth of the ridge-forming trench increases with with low penetration forces (S2, S6), two needles with suc-
increasing number of layers and scanning diameter and with cessful penetration and a high ridge (S1, S5), and two short
decreasing eccentricity. Interestingly, the eccentricity doubly needles (< 1095 µm) with unsuccessful penetration but a
affects the ridge height through its influence on the depth low tip angle, tip wall thickness, and tip radius (S3, S4).
of the ridge-forming trench and on the replication fidelity These needles and their corresponding penetration forces
of the ridge. are depicted in Fig. 21.
Drawing from the findings above, a laser scanning strat-
egy can be designed for a minimised tip angle: the eccen- 3.3 Set 3: variation of the injection moulding
tricity and scanning diameter should be maximised, while process and penetration tests
the number of layers should be chosen so that the minimal
required needle length is reached and the ridge is minimised. 3.3.1 Replication
However, this scanning strategy does not take into account
other functional necessities of HMNs such as the fact that Needles S1–S6 are replicated with PC using both a high and
functional HMNs require a significant ridge in order to pre- low holding pressure and GFR-PP. Their lengths and ridge
vent leakage, or that the distance between the top surface of heights are depicted in Fig. 22. Figure 22a shows an increase
the lumen-forming pillar and the mould surface should be in length for the replicated PC needles with higher holding
minimised. Leakage is a non-negligible issue for intradermal pressure. Indeed, a higher holding pressure forces the poly-
injection and extraction [25], and tests with a more appropri- mer material deeper into the HMN cavities. The replication
ate skin model should be performed to determine the mini- with GFR-PP leads to even longer HMNs. This discrep-
mum required ridge height for leak-free injections. Although ancy in overall replication fidelity across materials can be
the depth of the lumen-forming pillar would increase in the a result of four factors, namely linear shrinkage, interfacial

Fig. 20  Influence of a the tip


radius and b the tip wall thick-
ness on the penetration force.
The error bars and the shaded
area represent the standard devi-
ations on the penetration force.
The dashed line represents the
linear regression and R is the
linear correlation coefficient
The International Journal of Advanced Manufacturing Technology (2024) 132:533–551 547

Fig. 21  Selection of six different needle geometries for improvement of the injection moulding process and replication with GFR-PP and PC

wetting behaviour, creep deformation behaviour, and melt replication settings, they are still shorter than 1095 µm,
viscosity, as discussed by Evens et al. [20]. This increase in which was reported as the lower length limit for penetra-
replication fidelity can also be noticed in Fig. 22b, where tion (Sect. 3.2.3). Therefore, they can be expected to yield
the ridges of the PC needles with high holding pressure are unsuccessful penetration.
higher than those with low holding pressure in all but one Figure 23 shows the evolution of the tip geometry fea-
case, and the ridges of the GFR-PP needles are the high- tures for different holding pressures and polymer materi-
est. A similar reasoning to the needle height can be applied als. Contrary to the evolution of the needle length and ridge
for the ridge height, where a higher holding pressure and a height, no recurring trend can be found for the evolution
higher flowability, better wettability, and slower solidifica- of the tip angle, tip radius, and tip wall thickness. The tip
tion rate increase the filling of the microcavities. For needles angle is generally largest for the GFR-PP needles and can
S3 and S4, it is important to note that, even with the better be linked to the large increase in ridge height compared to

Fig. 22  Evolution of the a needle length and b ridge height for PC with low holding pressure (denoted by ‘low P’ in the legend), PC with high
holding pressure (‘high P’ in the legend), and GFR-PP
548 The International Journal of Advanced Manufacturing Technology (2024) 132:533–551

the relatively small increase in needle length for GFR-PP. radius, and tip angle are decisive features for the surface area
The tip radius varies significantly across the different nee- in contact with the skin during insertion. It can be intuitively
dles and injection moulding parameters, without any signifi- deduced that a low contact area is desired to reduce the pen-
cant trend. The range of the measured tip wall thicknesses etration force. Indeed, a reduction in contact area leads to a
is extended with this set of experiments, with a number of more localised application of the force and higher stress in
needles with tip wall thickness smaller than 30 µm, even as the skin. As a result, the ultimate tensile strength of the skin
low as 15 µm. When looking at the cavity shape, one could can be overcome at a lower penetration force. S3 and S4,
expect the tip wall thickness to be inversely proportional to which do not exhibit successful penetration with low hold-
the overall replication fidelity, due to the conical nature of ing pressure in PC, also do not penetrate in the other tested
the cavity. However, this trend cannot be confirmed from the configurations, even with the increase in needle height noted
data shown in Figs. 22a and 23c. in Sect. 3.3.1, and are therefore not shown in the figure.
The products of tip wall thickness, tip radius, and tip
3.3.2 Penetration angle do not follow any significant trend across all needles in
function of the tested configurations of the injection mould-
Figure 24 shows (a) the products of tip wall thickness, tip ing process. For needles S1 and S2, the products are not
radius, and tip angle and (b) a comparison of the forces changed in any major way by the higher holding pressure or
required to penetrate the artificial skin model for the differ- different polymer materials. For needles S3 and S5, there is a
ent HMNs in PC with low holding pressure, PC with high noticeable increase for both PC with higher holding pressure
holding pressure, and GFR-PP. The tip wall thickness, tip and GFR-PP compared to PC with low holding pressure. For

Fig. 23  Evolution of the a tip angle, b tip radius, and c tip wall thickness for PC with low holding pressure (denoted by ‘low P’ in the legend),
PC with high holding pressure (‘high P’ in the legend), and GFR-PP
The International Journal of Advanced Manufacturing Technology (2024) 132:533–551 549

Fig. 24  Evolution of a the product of the tip wall thickness, tip holding pressure (‘high P’ in the legend), and GFR-PP. The error bars
radius, and tip angle and b the penetration force for PC with low represent the standard deviations
holding pressure (denoted by ‘low P’ in the legend), PC with high

needle S4, the product of geometrical tip features decreases needle in the artificial skin model is measured at 49 ± 3 mN,
for a higher holding pressure and for GFR-PP. Needle S6 while the penetration force of the same needle in human sub-
sees a large decrease for the higher holding pressure and a jects (on the side of the abdomen) was reported between 160
major increase for GFR-PP compared to PC with low hold- and 190 mN [19]. This indicates that the penetration force
ing pressure. obtained using the artificial skin model could be underesti-
From Fig. 24b, it can be seen that for S1 and S2, no mated by up to three times. As a consequence, all needles
significant change in the penetration force is observed tested in this study might not be suited to pierce human skin
between the three configurations. The penetration force due to possible structural constraints leading to mechani-
of S5 increases with high holding pressure in PC and with cal failure during insertion. However, the approach remains
GFR-PP. Meanwhile, S6 penetration force decreases in PC suitable for studying the mechanisms of HMN insertion and
with high holding pressure due to a larger increase in ridge the relationships between fs laser ablation strategy, injec-
height compared to needle height. In contrast, S6 penetration tion moulding, and needle geometry. Also, for closing the
force increases in GFR-PP compared to PC with low holding gap to real-world applications, it is important to consider
pressure. These trends align with observations in Fig. 24a, biocompatibility and sterilisation possibilities of the chosen
where the product of tip wall thickness, tip radius, and tip polymer.
angle indicates that tip wall thickness is a significant factor
influencing penetration force, especially when considering
a broader range of measured tip wall thicknesses. 4 Conclusion
A better overall replication fidelity was expected to
result in a sharper needle and lead to a lower penetration In this work, three sets of experiments were performed with
force. However, this trend cannot be confirmed for injec- injection-moulded polymeric HMNs replicated using fs
tion-moulded hollow microneedles. Indeed, when compar- laser-ablated moulds. A number of laser scanning param-
ing Figs. 22a and 24, it can be noticed that a higher overall eters were varied, being the scanning base shape, the shape
replication fidelity (i.e. an increase in needle length) can and size of the scan-free area, the number of layers, the
even result in a higher penetration force when the increase in eccentricity, the scan-free area diameter, and the scanning
ridge height is significantly larger than the increase in needle diameter. The results showed that the laser scanning strate-
length, leading to a larger tip angle and tip radius. gies produced sharp-tipped cone-like cavities with a stand-
No specific mechanical failure analysis was performed ing pillar and that their morphologies are greatly affected
on the HMNs, but their geometrical features were examined by the laser scanning strategy parameters. The replicated
with a microscope after every penetration experiment (data needles were tested for penetration in an artificial skin model
not shown). These inspections showed that both the PC and consisting of agarose gel and a Parafilm® layer. The penetra-
GFR-PP HMNs could withstand the penetration tests with- tion tests showed that a minimum needle length is required
out any visible deformations. The penetration force of a 34G to achieve successful penetration and that the combination
550 The International Journal of Advanced Manufacturing Technology (2024) 132:533–551

of tip angle, tip radius, and tip wall thickness is a good indi- 2. Miller PR, Taylor RM, Tran BQ et al (2018) Extraction and bio-
cator of the penetration force. The results indicate that the molecular analysis of dermal interstitial fluid collected with hol-
low microneedles. Commun Biol 1:173. https://​doi.​org/​10.​1038/​
needle length and tip angle can be improved with a higher s42003-​018-​0170-z
eccentricity and scanning diameter, while the number of lay- 3. Meng F, Hasan A, Mahdi Nejadi Babadaei M et al (2020) Pol-
ers should be chosen so that the needle length is sufficient ymeric-based microneedle arrays as potential platforms in the
while the ridge height remains low. However, it is impor- development of drugs delivery systems. J Adv Res 26:137–147.
https://​doi.​org/​10.​1016/j.​jare.​2020.​07.​017
tant to note that, while minimising the penetration force, 4. Li Y, Zhang H, Yang R et al (2019) Fabrication of sharp silicon
this strategy does not take into account the potential leak- hollow microneedles by deep-reactive ion etching towards mini-
age resulting from a HMN with low ridge, or the potential mally invasive diagnostics. Microsyst Nanoeng 5:1–11. https://​
needle clogging resulting from an excessive depth of the doi.​org/​10.​1038/​s41378-​019-​0077-y
5. Van Hileghem L, Kushwaha S, Piovesan A et al (2023) Innovative
lumen-forming pillar. It has also been demonstrated that an fabrication of hollow microneedle arrays enabling blood sampling
increase in overall replication fidelity (i.e. only considering with a self-powered microfluidic patch. Micromachines 14:615.
the needle length and cavity depth) could not be directly https://​doi.​org/​10.​3390/​MI140​30615/​S1
related to lower penetration force due to the change in repli- 6. Bystrova S, Luttge R (2011) Micromolding for ceramic micronee-
dle arrays. Microelectron Eng 88:1681–1684. https://​doi.​org/​10.​
cation fidelity, leading to variations of the tip angle. 1016/J.​MEE.​2010.​12.​067
This comparative study demonstrated the importance of 7. Evens T, Van Hileghem L, Dal Dosso F et al (2021) Producing
HMN morphology and highlighted the potential of injection hollow polymer microneedles using laser ablated molds in an
moulding with laser-ablated moulds for the manufacturing injection molding process. J Micro Nanomanuf 9:1–9. https://​
doi.​org/​10.​1115/1.​40514​56
of functional hollow microneedles. It should be noted that, 8. Makvandi P, Kirkby M, Hutton ARJ et al (2021) Engineering
although single needle configurations were studied in this microneedle patches for improved penetration: analysis, skin mod-
work, HMNs are typically arranged in arrays to increase the els and factors affecting needle insertion. Nanomicro Lett 13:93.
throughput and to target a larger area [26]. The configuration https://​doi.​org/​10.​1007/​S40820-​021-​00611-9
9. Permana AD, Mir M, Utomo E, Donnelly RF (2020) Bacterially
of the array and skin model, such as the spacing between the sensitive nanoparticle-based dissolving microneedles of doxycy-
needles, the skin pretension, or the presence of a solid base cline for enhanced treatment of bacterial biofilm skin infection: a
layer, was reported to impact the penetration behaviour of proof of concept study. Int J Pharm X 2:100047. https://​doi.​org/​
the needle arrays and should therefore be taken into account 10.​1016/J.​IJPX.​2020.​100047
10. Permana AD, Paredes AJ, Volpe-Zanutto F et al (2020) Dissolving
for the development of hollow microneedles [27, 28]. microneedle-mediated dermal delivery of itraconazole nanocrys-
tals for improved treatment of cutaneous candidiasis. Eur J Pharm
Acknowledgements The authors would like to thank Olivier Malek Biopharm 154:50–61. https://d​ oi.o​ rg/1​ 0.1​ 016/J.E
​ JPB.2​ 020.​06.​025
from Sirris (department of Precision Manufacturing) for the laser abla- 11. Koelmans WW, Krishnamoorthy G, Heskamp A et al (2013)
tion of the microneedle cavities. The KU Leuven XCT Core facility Microneedle characterization using a double-layer skin simulant.
is acknowledged for the 3D image acquisition and quantitative post- Mech Eng Res 3:51. https://​doi.​org/​10.​5539/​MER.​V3N2P​51
processing tools (https://​xct.​kuleu​ven.​be/). 12. Wang QL, Zhu DD, Chen Y, Guo D (2016) A fabrication method
of microneedle molds with controlled microstructures. Mater Sci
Author contribution Conceptualization, methodology, and investiga- Eng C 65:135–142. https://​doi.​org/​10.​1016/j.​msec.​2016.​03.​097
tion were performed by Pol Vanwersch and Tim Evens. The first draft 13. Singh Kochhar J, Quek TC, Soon WJ et al (2013) Effect of
of the manuscript was written by Pol Vanwersch and all authors com- microneedle geometry and supporting substrate on microneedle
mented on previous versions of the manuscript. array penetration into skin. J Pharm Sci 102:4100–4108. https://​
doi.​org/​10.​1002/​jps.​23724
Funding This work was funded by the KU Leuven Interdisciplinary 14. Flaten GE, Palac Z, Engesland A et al (2015) In vitro skin models
Network project IDN/20/011 MIRACLE: Autonomous microfluidic as a tool in optimization of drug formulation. Eur J Pharm Sci
patch for plasmid-based vaccine, as well as the Fonds Wetenschap- 75:10–24. https://​doi.​org/​10.​1016/J.​EJPS.​2015.​02.​018
pelijk Onderzoek (FWO)–Vlaanderen SBO project S003923N and SB 15. Yadav PR, Han T, Olatunji O et al (2020) Mathematical model-
fellowship 1S31022N. The FWO large infrastructure I013518N project ling, simulation and optimisation of microneedles for transder-
is acknowledged for their financial support of the X-ray infrastructure. mal drug delivery: trends and progress. Pharmaceutics 12:1–31.
https://​doi.​org/​10.​3390/​PHARM​ACEUT​ICS12​080693
Declarations 16. Han J, Malek O, Vleugels J et al (2022) Ultrashort pulsed laser
ablation of zirconia-alumina composites for implant applications.
Competing interests The authors declare no competing interests. J Mater Process Technol 299:117335. https://​doi.​org/​10.​1016/J.​
JMATP​ROTEC.​2021.​117335
17. Evens T, Castagne S, Seveno D, Van Bael A (2022) Compar-
ing the replication fidelity of solid microneedles using injection
compression moulding and conventional injection moulding.
References Micromachines 13:1280. https://​doi.​org/​10.​3390/​MI130​81280
18. Larrañeta E, Moore J, Vicente-Pérez EM et al (2014) A proposed
model membrane and test method for microneedle insertion stud-
1. Menon I, Bagwe P, Gomes KB et al (2021) Microneedles: a new ies. Int J Pharm 472:65–73. https://​doi.​org/​10.​1016/j.​ijpha​r m.​
generation vaccine delivery system. Micromachines 12:4. https://​ 2014.​05.​042
doi.​org/​10.​3390/​MI120​40435
The International Journal of Advanced Manufacturing Technology (2024) 132:533–551 551

19. Præstmark KA, Jensen ML, Madsen NB et al (2016) Pen nee- 25. Kim YC, Park JH, Prausnitz MR (2012) Microneedles for drug
dle design influences ease of insertion, pain, and skin trauma and vaccine delivery. Adv Drug Deliv Rev 64:1547. https://​doi.​
in subjects with type 2 diabetes. BMJ Open Diabetes Res Care org/​10.​1016/J.​ADDR.​2012.​04.​005
4:e000266. https://​doi.​org/​10.​1136/​bmjdrc-​2016-​000266 26. Tuan-Mahmood T-M, McCrudden MTC, Torrisi BM et al (2013)
20. Evens T, Malek O, Castagne S et al (2021) Controlling the geom- Microneedles for intradermal and transdermal drug delivery. Eur
etry of laser ablated microneedle cavities in different mould mate- J Pharm Sci 50:623–637. https://​doi.​org/​10.​1016/j.​ejps.​2013.​05.​
rials and assessing the replication fidelity within polymer injection 005
moulding. J Manuf Process 62:535–545. https://d​ oi.o​ rg/1​ 0.1​ 016/j.​ 27. Xiu X, Gao G, Liu Y, Ma F (2022) Drug delivery with dissolving
jmapro.​2020.​12.​035 microneedles: skin puncture, its influencing factors and improve-
21. Penchev P, Dimov S, Bhaduri D et al (2017) Generic software tool ment strategies. J Drug Deliv Sci Technol 76:103653. https://​doi.​
for counteracting the dynamics effects of optical beam delivery org/​10.​1016/j.​jddst.​2022.​103653
systems. Proc Inst Mech Eng, Part B 231:48–64. https://​doi.​org/​ 28. Shu W, Heimark H, Bertollo N et al (2021) Insights into the
10.​1177/​09544​05414​565379 mechanics of solid conical microneedle array insertion into skin
22. Evens T, Castagne S, Seveno D, Van Bael A (2022) Predicting the using the finite element method. Acta Biomater 135:403–413.
replication fidelity of injection molded solid polymer micronee- https://​doi.​org/​10.​1016/J.​ACTBIO.​2021.​08.​045
dles. Int Polym Process 37:237–254. https://​doi.​org/​10.​1515/​
ipp-​2021-​4207 Publisher's Note Springer Nature remains neutral with regard to
23. Khanna P, Luongo K, Strom JA, Bhansali S (2010) Sharpening jurisdictional claims in published maps and institutional affiliations.
of hollow silicon microneedles to reduce skin penetration force.
J Micromech Microeng 20:4. https://​doi.​org/​10.​1088/​0960-​1317/​ Springer Nature or its licensor (e.g. a society or other partner) holds
20/4/​045011 exclusive rights to this article under a publishing agreement with the
24. Kong XQ, Zhou P, Wu CW (2011) Numerical simulation of author(s) or other rightsholder(s); author self-archiving of the accepted
microneedles’ insertion into skin. Comput Methods Biomech manuscript version of this article is solely governed by the terms of
Biomed Engin 14:827–835. https://​doi.​org/​10.​1080/​10255​842.​ such publishing agreement and applicable law.
2010.​497144

You might also like