Biomedical Microdevices 6:3, 177±182, 2004
# 2004 Kluwer Academic Publishers. Manufactured in The Netherlands.
Communications & Therapeutic Micro and Nanotechnology Section:
Microneedle Insertion Force Reduction Using Vibratory
Actuation
Abstract. The effect of vibratory actuation on microneedle insertion
force was investigated. Hollow micro hypodermic injection needles
were fabricated by a two-wafer polysilicon micromolding process.
A vibratory actuator operating in the kHz range was coupled with
the hypodermic microneedles. The force to insert microneedles into
excized animal tissue was measured with a load cell. Results showed
a greater than 70% reduction in microneedle insertion force by
using vibratory actuation. The application of vibratory actuation
provides a promising method to precisely control the microneedle
insertion forces to overcome microneedle structural material
limitations, minimize insertion pain, and enhance the ef®ciency of
drug delivery.
Key Words. microneedle, insertion force, vibratory actuation, drug
delivery
Introduction
Microneedles are promising microfabricated devices for
minimally invasive drug delivery applications.
Microneedles are designed to be high performance
minimally invasive conduits, through which drug
solutions may pass into the body. In order to be
minimally invasive, the needles are designed to be as
small as possible. Needles are also designed to be
extremely sharp, with sub-micron tip radii. This allows
the needles to be effectively inserted into the skin.
Microneedles offer an attractive way for advanced drug
delivery systems by mechanically penetrating the skin
and injecting drug just under the stratum corneum where
it is rapidly absorbed by the capillary bed into the
bloodstream.
The stratum corneum is the outermost layer of the
skin, which serves as an effective material permeation
barrier. Beneath this lies the viable epidermal layer that is
devoid of blood vessels and contains very few nerve
endings. The next layer is the highly vascularized dermis
Section Editor Dr Tejal Desai
Ming Yang1 and Jeffrey D. Zahn1,2*
1
Department of Bioengineering, The Pennsylvania State University
2
Materials Research Institute, The Pennsylvania State University,
224 Hallowell Building, University Park, PA 16802
E-mail: jdzbio@engr.psu.edu
that consists of blood vessels, nerves, lymph vessels,
dendritic cells, hair follicles, collagen, and sweat glands.
Underlying the dermal layer is the fatty subcutaneous
layer, with fat cells, blood vessels, and connective tissue.
Using optical lithographic techniques, microneedles
can be fabricated to be almost any size and geometry.
Many different microneedle designs have been fabri-
cated (McAllister et al., 2000), the majority being either
micro-hypodermic injection needles (Zahn et al.,
2000a,b; Talbot and Pisano, 1998; Lin et al., 1993;
Lebouitz and Pisano, 1998; Chen and Wise, 1994;
Papautsky et al., 1997; Brazzle et al., 1999, 2000;
Chandrasekaran and Frazier, 2002; Stupar and Pisano,
2001) (several mm long) or high density arrays of shorter
needles (Henry et al., 1998; McAllister et al., 1999;
Stoeber and Liepmann, 2000a,b, 2002) (200±300 mm
long) normal to the plane of a wafer designed to penetrate
the stratum corneum. Since the technology is based on
silicon processing techniques, the majority of micro-
needles use silicon as a structural material. However,
there has been some recent work using electroplated
metals (Papautsky et al., 1997; Brazzle et al., 1999, 2000;
Chandrasekaran and Frazier, 2002; Henry et al., 1998)
and polymer (Stupar and Pisano, 2001) microneedle
designs.
Since microneedles allow such compact, potentially
portable and precise delivery of therapeutics, they are
well suited for any drug delivery regime in which a
continuous infusion is preferred to a bolus injection. A
microfabricated delivery module (Liepmann et al., 1999)
with integrated microneedles also allows more freedom
in a patient's drug regiment by freeing the patient from
bulky intravenous lines, or portable belt worn pumps
with a large stainless steel needle protruding into the
abdomen. Such a device could be worn as a patch style
*Corresponding author.
177
178 Yang and Zahn

device on an arm or the abdomen, so that the patient vibratory actuation on the microneedle insertion force
hardly realizes they are wearing it. was studied.
Most hypodermic needle injections are given sub-
cutaneously or intramuscularly, causing compression of
nerve endings in the dermal layer and causing pain. The Fabrication
development of smaller needles allows the delivery of
therapeutic compounds under the stratum corneum into Molded microneedles use patterned mold wafers, similar
the viable epidermis avoiding nerve compression and in concept to Keller and Howe (1995), that are not
hence pain (Kaushik et al., 2001). The compounds are sacri®ced during the fabrication process. Therefore,
then transported to the dermal layer where they are many different designs of needles may be made, and
absorbed into the bloodstream. This delivery avenue is the mold may be reused numerous times at a cost saving
almost as fast as intravascular delivery but because of the compared to sacri®cial processing techniques.
small size of the device it can be done less invasively. Microneedles have been fabricated as previously
Microneedles have potential advantages over other reported (Zahn et al., 2000a). A double polished silicon
approaches to transdermal drug delivery such as wafer ( p-type h100i single crystal) is used as a starting
electroporation (Prausnitz, 1995), ultrasonic delivery material as shown in Figure 1. The wafer is wet oxidized
(Kost, 1993), or chemical modi®ers/enhancers (Smith at 1,000  C to grow 1 mm of thermal oxide. The needle
and Malbach, 1995) all of which rely on decreasing the shapes are patterned using standard lithographic tech-
permeation barrier of the stratum corneum. Oral drug niques and the patterned oxide is etched with reactive ion
delivery routes are subject to ®rst pass clearance in the etching (RIE). The wafer is then coated with 0.3 mm of
liver, and pulmonary delivery is more suited to short term low pressure chemical vapor deposition (LPCVD) low
bolus drug delivery rather than long term maintenance
regulation.
Microneedles mechanically penetrate the skin barrier
and allow the injection of any volume of ¯uid over time.
This allows precise localization of a drug solution in
order to obtain effective absorption into the bloodstream.
The drug delivery does not depend on transient delivery
of therapeutics through the skin. The delivery is
independent of the drug composition and merely relies
on the subsequent drug absorption into the bloodstream,
which occurs at a much faster rate than permeation of a
solution across the skin. This also allows complex drug
delivery pro®les. Since drug is actively injected into a
patient the dosage may be varied with time. In addition,
by employing multiple needles or effective ¯uid control
with mixing of solutions, multiple drugs may be injected
simultaneously speci®c to a patient's personal needs.
A fundamental problem in the design and fabrication
of microneedles, however, is to achieve a satisfactory
balance between structural rigidity and miniaturization.
Needle deformation must be minimized during insertion.
The needles must also be able to tolerate forces
associated with intact removal and normal human
movements. One of the barriers to the advancement of
microneedle technology for drug delivery is being able to
precisely control the forces the needle experiences for
effective insertion. Mosquitoes use vibratory cutting at a
frequency of 200±400 Hz to pierce the skin. A precise
mechanical actuator may be used for controlling the
forces on the needle during insertion to minimize
insertion pain. A way to take advantage of material
limitations is to precisely control the stresses and forces Fig. 1. (Top) SEM of a microneedle mold. (Bottom) SEM of a
the needle experiences. In this research, the effect of microneedle with a 6 mm shaft and a micro®lter in its base.
 Microneedle Insertion Force Reduction Using Vibratory Actuation
stress silicon nitride. The backside of the wafer is aligned
and patterned to the needle design using a Karl Suss
contact mask aligner with backside alignment capabil-
ities. The silicon nitride and oxide ®lms are then etched
away using RIE leaving a through-hole mask. Afterwards
the through holes are etched with potassium hydroxide
(KOH). The nitride is then removed in phosphoric acid at
175  C. The needle mold is then etched using deep
reactive ion etching (DRIE) in a surface technology
systems (STS) etcher. The resulting trench is typically
between 100 and 125 mm deep. The wafer is then wet
oxidized at 1,000  C to decrease surface roughness.
Afterwards 2 mm of phosphosilicate glass (PSG) is
deposited onto the mold wafer. A second bare silicon
wafer is coated with 2 mm of PSG. These two wafers are
pressure bonded at 1,000  C in a nitrogen atmosphere.
After bonding, 4 mm of polysilicon is deposited onto the
mold wafers at 580  C. The molds are then annealed at
1,000  C in a N2 atmosphere. The polysilicon is annealed
between depositions to alleviate ®lm stress. The
deposition and annealing is repeated until the desired
thickness is reached, typically 15±20 mm. The poly-
silicon deposits conformally onto the outside of the wafer
and inside the mold through the KOH etched through
holes. After deposition, the external polysilicon is
removed by RIE. The needles are then released in
concentrated HF overnight. Figure 2 shows the SEM
micrographs of the microneedle mold (top) and a
microneedle with a 6 mm shaft and a micro®lter in its
base (bottom).
Fig. 2. Fabrication scheme for making microneedles using a single
wafer patterned on both sides.
179
Experimental Apparatus
In order to characterize microneedle insertion force, a
microneedle was coupled to a vibratory actuator and
placed on a motorized stage (Figure 3). Microneedles
were mounted on a glass slide using cyanoacrylate ester
adhesive. The base was attached to the slide with the tip
and shank cantilevered off the edge of the slide. Then the
slide was coupled to a vibratory actuator operating in the
kHz frequency range. The vibratory actuator was ®nally
mounted onto a motorized stage (Nanomotion Ltd.). A
piece of excized animal tissue was coupled to a LCL-
816G force transducer (Omega Engineering Inc.) for
microneedle insertion test. The microneedle insertion
force was determined by displacing the stage towards the
tissue sample at a speed of 1 mms. The control of motor
displacement and data acquisition were achieved through
a LabVIEW program. Figure 4 shows the static and
vibrating status of a microneedle.
Results and Discussion
Needles must be able to tolerate forces associated with
insertion, intact removal and normal human movements.
Current microneedle designs do not meet these require-
ments and have not effectively entered the biomedical
market because they are either too fragile or too ductile
to ef®ciently penetrate the skin. Compared to silicon
needle, metal and polymer needles are less stiff so they
can absorb larger stresses by plastically deforming.
However, this ability also makes piercing the skin more
dif®cult. Metal deposition also uses thin ®lm processing
techniques, and therefore the metals are mechanically
weaker than bulk hardened metals such as stainless steel.
This makes the metal and polymer microneedles more
dif®cult to insert into the skin. Using vibratory actuation,
the stresses and forces the needle experiences can be
precisely controled, therefore improved needle insertion
can be achieved. This will make it feasible to use silicon,
metal and polymer microneedles for effective drug
delivery.
Arrays of microneedles with multiple ¯ow channels
may enhance the ef®ciency of drug delivery since desired
liquid ¯ow rates may be more easily achieved by
adjusting the needle density. However, arrays of
microneedles (Henry et al., 1998; McAllister et al.,
1999; Stoeber and Liepmann, 2000a,b, 2002) make skin
penetration more dif®cult because of the so-called ``bed
of nails'' effect. This phenomenon occurs because the
force per unit area of the needle array against the skin
(i.e., pressure) determines if the needles will pierce the
skin. The area over which an imposed load acts is
180 Yang and Zahn
Fig. 3. Experimental setup of microneedle insertion test.
determined by the effective area of the microneedle tips
in contact with the skin. If there were only one needle,
the entire load would be distributed over the very small
area presented by the tip of the one needle. In this case,
the force per unit area, would be very large (because the
total area is small) and would likely result in skin
penetration. However, when an array of needles is used,
the same imposed load is distributed over hundreds of
needles. Therefore, the force applied to any one needle is
correspondingly reduced, with the result that the force
per unit area at the tip of any one needle will be well
below the level required to pierce the skin. Therefore, a
proportionally higher imposed load is required for
effective needle penetration. The load for microneedle
arrays inserted into the skin has been measured and found
to be about 0.3 N and was found to increase linearly with
increasing needle surface area (Davis et al., 2002).
As the needle impacts the skin the resulting force is
measured on the transducer. The force increases with
further needle displacement until the skin is penetrated
by the needle where a dramatic reduction in force is seen.
The force again rises as the needle penetrates into the
underlying tissue (Figure 5). The maximum insertion
force is regarded as the force required for skin
penetration. The force determined is comparable to the
force determined by other researchers (Chandrasekaran
and Frazier, 2002; Stupar and Pisano, 2001; Davis et al.,
2002). As seen in Figure 5, there is a greater than 70%
decrease in insertion force of a silicon hypodermic style
microneedle 100 mm in diameter and 6 mm long when a
 Microneedle Insertion Force Reduction Using Vibratory Actuation
Fig. 4. (Top) Micrograph of static microneedle. (Bottom) Micrograph of vibrating microneedle.
vibratory actuator operating in the kHz range with a
lateral vibration amplitude of 0.6 mm is used during
microneedle insertion. Also, the penetration distance into
the same sample is greatly reduced when using vibratory
actuation. This may be because the skin is thinner during
this insertion. However, since both experiments are done
in a serial fashion in the same piece of tissue it is more
likely that the vibratory motion caused the skin to tear at
a smaller displacement.
This vibratory motion dramatically lowers micro-
needle insertion forces by three mechanisms: (1) direct
Fig. 5. Graph of skin penetration force from excized animal tissue
showing a greater than 70% reduction in insertion force using the
vibratory actuator.
181
mechanical impacts of the microneedle with the skin at
high frequencies leads to an increase in the skin's
dynamic stiffness, (2) the interaction of vibratory
pressure waves with the tissue ¯uid causing cavitation
which destroys tissue, and (3) localized thermal damage
due to frictional interactions between the microneedle
and skin. These three mechanisms occur simultaneously
leading to lower microneedle insertion forces. Silicon
based ultrasonic surgical devices have been explored for
opthamological surgery (Lal, 1998; Lal and White, 1995)
and have dramatically reduced cutting forces in test
materials. This shows a reduction in cutting force from
70 g with no power to 5 g when power is supplied to the
device continuously, a greater than 90% force reduction
(Lal and White, 1995). The applying of vibratory
actuation during microneedle insertion is expected to
substantially reduce the amount of force required for
microneedle penetration.
Conclusion
A fundamental problem in the design and fabrication of
microneedles is to achieve a satisfactory balance
between structural rigidity and miniaturization. Needle
deformation must be minimized during insertion. The
182 Yang and Zahn
needles must also be able to tolerate forces associated
with intact removal and normal human movements. In
this research, the effect of vibratory actuation on the
microneedle insertion force was studied. Results showed
signi®cant decrease in insertion force of silicon
hypodermic style microneedles when a vibratory
actuator operating in the kHz range is used during
microneedle insertion. The applying of vibratory actuator
may solve the problem mentioned above and the forces
that the needle experiences may be precisely controled
for effective insertion. Future work will involve
redesigning the vibratory actuators to oscillate axially
to the needle rather than laterally to better couple the
vibratory energy into microneedle insertion as well as
testing less brittle microneedles made from metals and
polymer.
Acknowledgments
The authors would like to thank Dorian Liepmann and
members of the Zahn laboratory for technical discussions
and assistance.
References
J. Brazzle, I. Papautsky, and A.B. Frazier, IEEE Engineering in
Medicine and Biology Magazine 18(6), 53±58 (1999).
J. Brazzle, D. Bartholomeusz, R. Davies, J. Andrade, R.A. Van
Wageman, and A.B. Frazier, Proceedings 2000 Solid State Sensor
and Actuator Workshop Hilton Head, S.C., 199±202 (2000).
S. Chandrasekaran and A.B. Frazier, IEEE EMBS Special Topic
Conference on Microtechnologies in Medicine and Biology,
Madison, WI, 94±98 (May, 2002).
J. Chen and K.D. Wise, Proceedings 1994 Solid State Sensor and
Actuator Workshop, Hilton Head, S.C., 256±259 (1994).
S.P. Davis, M.G. Allen, and M.R. Prausnitz, Proceedings of the 2nd
Joint EMBS/BMES Conference Houston, TX, 498±499 (2002).
S. Henry, D.V. McAllister, M. Allen, and M. Prausnitz, Proceedings of
the IEEE Eleventh Annual International Workshop on MEMS,
Heidelberg, Germany, 494±498 (1998).
S. Kaushik, A.H. Hord, D.D. Denson, D.V. McAllister, S. Smitra, M.G.
Allen, and M.R. Prausnitz, Anesth. Analg. 92, 502±504 (2001).
C.G. Keller and R.T. Howe, International Solid State Sensors and
Actuator Conference. Transducers `95, Stockholm, Sweden, 376±379
(1995).
J. Kost, J. Controlled Release 24, 247±255 (1993).
A. Lal, Proc. of the 20th Annual International Conference of the IEEE
Engineering in Medicine and Biology Society 20(6), 2785±2790
(1998).
A. Lal and R.M. White, Proc. IEEE Ultrasonic Symposium 1995 2,
1593±1596 (1995).
K.S. Lebouitz and A.P. Pisano, Proceedings Microstructures and
Microfabrication Systems IV, 194th meeting of the Electrochemical
Society, Boston, MA (Nov. 1±6, 1998).
D. Liepmann, A. Pisano, and B. Sage, Diabetes Technology &
Therapeutics 1(4), 469±476 (1999).
L. Lin, A.P. Pisano, and R.S. Muller, Proceedings 7th International
Conference on Solid State Sensors and Actuators (Transducers'93),
Yokohama, Japan, 237±240 (1993).
D.V. McAllister, M.G. Allen, and M.R. Prausnitz, Annu. Rev. Biomed.
Eng. 2, 289±313 (2000).
D.V. McAllister, F. Cros, S.P. Davis, L.M. Matta, M.R. Prausnitz, and
M.G. Allen, Proceedings 10th International Conference on Solid
State Sensors and Actuators (Transducers'99), 1098±1101 (1999).
I.E. Papautsky, J.D. Brazzle, H. Swerdlow, and A.B. Frazier, IEEE
International Conference on Engineering in Medicine and Biology
Conference Chicago, IL (October, 1997).
M.R. Prausnitz, In: Electronically Controlled Drug Delivery, edited by
B. Berner and S.M. Dinh (CRC Press, Boca Raton, FL, 1995).
E.W. Smith and H.L. Malbach, Percutaneous Penetration Enhancers
(CRC Press, Boca Raton, FL, 1995).
B. Stoeber and D. Liepmann, Proceedings of the ®rst IEEE-EMBS
Special Topic Conference on Microtechnology in Medicine &
Biology, 224±228 (2000a).
B. Stoeber and D. Liepmann, Proceedings of the ASME MEMS
Division, IMECE, 1, 355±359 (2000b).
B. Stoeber and D. Liepmann, Proceedings 2002 Solid State Sensor and
Actuator Workshop Hilton Head, S.C., 77±80 (2002).
P. Stupar and A.P. Pisano, Proceedings 11th International Conference on
Solid State Sensors and Actuators (Transducers'01), Munich,
Germany, 1356±1359 (2001).
N. Talbot and A.P. Pisano, Proceedings 1998 Solid State Sensor and
Actuator Workshop, Hilton Head S.C., 265±268 (1998).
J.D. Zahn, N.H. Talbot, D. Liepmann, and A.P. Pisano, Biomedical
Microdevices 2(4), 295±303 (2000a).
J.D. Zahn, D. Trebotich, and D. Liepmann, Proceedings of the First
IEEE EMBS Joint Conference on Microtechnology in Biology and
Medicine, 375±380 (2000b).