IEEE TRANSACTIONS ON MAGNETICS, VOL. 38, NO. 5, SEPTEMBER 20022441
The CantiClever: A Dedicated Probe for Magnetic
Arnout van den Bos, Iwan Heskamp, Martin Siekman, Leon Abelmann, and Cock Lodder
Abstract—We present a new cantilever for magnetic-force
microscopy (MFM), the CantiClever, which is not derived from
atomic-force microscopy (AFM) probes but optimized for MFM.
Our design integrates the cantilever and the magnetic tip in a
single manufacturing process with the use of silicon microma-
chining techniques, which allows for batch fabrication of the
probes. This manufacturing process enables precise control on all
dimensions of the magnetic tip, resulting in a very thin magnetic
element with a very high aspect ratio. Using the CantiClever,
magnetic features down to 30 nm could be observed in a CAMST
Index Terms—Cantilevers, integration, magnetic force mi-
croscopy (MFM), magnetic tips.
be obtained. A typical resolution of 50 nm can be achieved
with very little sample preparation . This makes MFM a very
useful tool for the study and characterization of magnetic ma-
terials for high-density magnetic recording. As the areal den-
sity of magnetic recording systems increases very rapidly, the
current resolution of MFM needs to be improved in order to re-
main useful as a measurement tool . The resolution of MFM
is determined by a combination of tip geometry and measure-
termines the measurement noise levels in modern instruments;
this thermal noise may be reduced by using cantileverswith low
spring constants, high resonant frequencies, and high quality of
resonance. This imposes a great challenge on the design of dy-
namic mode control electronics. However, at the moment it ap-
pears that the resolution of MFM is limited by the geometry of
the magnetic tip. The dimensions as well as the shape of the
magnetic tip have a big influence on the resolution. The ideal
tip shape would be that of an elongated bar with a flat end, as
depicted in Fig. 1 .
An elongated shape induces a strong shape anisotropy, which
increases the stability of the magnetic tip against remagneti-
zation by the sample. A long magnetic element also behaves
as a monopole when imaging small magnetic features, which
increases the signal output. Moreover, for thick films, a long
tip also improves the output at very long wavelengths. The flat
tip-end causes most of the magnetic charges to be confined as
close as possible to the sample. This results in an improvement
ITH magnetic force microscopy (MFM) a high-resolu-
tion image of the stray field of a magnetic sample can
Manuscript received February 14, 2002; revised May 28, 2002.
The authors are with the Systems and Materials for Information Storage
Research Institute, University of Twente, 7500 AE Enschede,
The Netherlands (e-mail: email@example.com).
Digital Object Identifier 10.1109/TMAG.2002.803585.
Fig. 1.Left: ideal tip shape fora MFM tip. Right: cantileverwith the tip plane.
in resolution as compared to point sharp tips, which suffer from
having a large proportion of their magnetic charges at a greater
consist of an atomic-force microscopy (AFM) tip, coated with
a thin layer of magnetic material, usually cobalt alloys . The
pyramidal shape of these tips does not bear much resemblance
to the ideal tip shape mentioned above. A better approximation
to this ideal shape is obtained with electron-beam-induced de-
posited (EBID) carbon needles –, coated on one side with
a thin magnetic layer. However, the EBID carbon needle has a
shape similar to that of a cone. When coated with magnetic ma-
terial, this causes not all the magnetic charges to be confined to
the end of the tip, making the tip sensitive to stray fields from a
larger area. Another major drawback of the EBID is their serial
manufacturing process. Therefore, making one tip takes a con-
for MFM measurement is used by Phillips , who deposits
magnetic material onto an AFM cantilever and uses focused
ion-beam equipment to remove unwanted areas of the mag-
netic coating. With this serial production method 8- m-long
sented in this paper results in an MFM tip with a nearly perfect
shape that is produced in a highly reproducible batch manufac-
A magnetic tip that is suitable for high-resolution MFM
should have lateral dimensions in the nanometer regime. One
would like these dimensions to be controllable and variable to
0018-9464/02$17.00 © 2002 IEEE
2442 IEEE TRANSACTIONS ON MAGNETICS, VOL. 38, NO. 5, SEPTEMBER 2002
conventional approach with perpendicularly vibrating cantilever; right: new
approach with laterally vibrating cantilever.
Tip plane on the cantilever for both cantilever orientations. Left:
ments or samples. The CantiClever design accomplishes this by
defining both lateral dimensions by thin film deposition tech-
niques. Themagnetic tip is made bydeposition ofmagnetic ma-
terial on the side of a free-hanging, very thin layer called the tip
plane. The width and thickness of the magnetic tip are defined
by the thickness of the tip plane and the magnetic layer, respec-
tively. The length of the tip is defined using photolithography.
A schematic drawing of the structure is shown in Fig. 1.
Such a structure is very difficult to realize when using the
conventional fabrication technique in which the cantilevers are
surface of the substrate. Using this approach, the very thin tip
plane needs to be fabricated as a free-standing layer perpendic-
ular to the substrate surface. To avoid this, the free-hanging tip
plane is made with a completely new approach. During fabrica-
tion, the cantilevers are tilted 90 compared to the conventional
cantilevers, creating a laterally oscillating cantilever with its os-
cillation direction parallel to the substrate surface, as shown in
This approach makes the fabrication of the cantilever more
difficult compared to conventional cantilevers, but at the same
time enables precise control over the cantilever resonance fre-
quency which is given by 
fundamental frequency, having a value of 1.875, the thickness
(150 GPa) and
the density of silicon (2330 kg/m ). In con-
this approach allows both the cantilever length as well as the
substrate can carry a large number of cantilevers with different
resonance frequencies, suitable for different applications. Fur-
thermore, standard deposition and etching techniques can be
used to define the tip plane, as it is also oriented parallel to the
substrate surface. The result is a reproducible manufacturing
process that incorporates both the cantilever and the magnetic
tip and allows for batch fabrication of the probes.
is the eigenvalue of the system corresponding to its
Two factors play an important role in realizing a high-resolu-
tion MFM tip with the ideal tip shape, as previously described.
First, the free-standing tip plane should be very thin, as this de-
fines the width of the tip, but also strong enough for contact
imaging. At the same time, the tip plane material should have
low residual stress to prevent bending after release. Second, the
as illustrated in Fig. 3 and the etching process should be direc-
tional to ensure that the tip end remains flat.
KOH wet anisotropic etching. The (110) oriented silicon wafers
plane etch stop also ensures very smooth surfaces, suitable for
interferometric deflection detection.
KOH etching of the cantilevers is done in two steps: the first
step etches the wafer from the backside, defining the width of
the cantilevers. In a second step, the cantilevers are etched from
the frontside of the wafer. This procedure ensures that the width
of the cantilevers can be precisely controlled. The masks used
during KOH etching are thin silicon nitride layers, deposited in
planes. The SiN masking layer used during KOH etching is pat-
terned afterwards to form the tip planes. The sharp cutoff corner
of the tip plane is obtained by removing the rounded end of the
As a final step, a thin Co coating is evaporated on the front side
of the tip plane. To reduce the amount of magnetic material de-
posited on the sides of the tip plane, accurate alignment of the
probe with respect to the evaporation source is needed. A scan-
ning-electron microscope (SEM) photograph of a CantiClever
is shown in Fig. 4. The free-hanging tip plane and the smooth
sides of the cantilever itself can clearly be distinguished.
plane perpendicular to the
A CantiClever with a resonance frequency of 60 kHz,
a 300- m-long cantilever, and a 50-nm Co layer on the
VAN DEN BOS et al.: THE CANTICLEVER: A DEDICATED PROBE FOR MAGNETIC FORCE MICROSCOPY2443 Download full-text
SEM photograph of the CantiClever showing the cantilever and the tip
Fig. 5.MFM measurement on a CAMST reference sample.
50-nm-thick tip plane, was mounted in a DI3100 microscope
for AFM and MFM measurements.
reference sample as used in . For this scan, the CantiClever
was operated in lift mode, detecting the frequency shift at a
15-nm tip to sample distance. In this scan, magnetic features
down to 30 nm could be distinguished. Currently, work is being
done on improvement of the resolution by, among others, re-
ducing the thickness of the tip plane. It is expected that the tip
plane thickness can be reduced below 10 nm.
A new design of a cantilever dedicated for MFM has been
developed. The new design integrates the cantilever and the tip
in one single batch manufacturing process with the use of sil-
icon micromachining techniques. The MFM tip is defined by
deposition of a thin magnetic film on the side of a free-standing
SiN layer. In this way, both thickness and width of the mag-
netic tip can be defined very accurately by layer deposition
techniques. To facilitate the realization of the free-standing SiN
layer, the cantilever itself is rotated 90 with respect to con-
ventional manufacturing techniques. A smooth upper surface,
suitable for interferometry, is obtained by anisotropic etching
of (110) Si wafers in KOH. The manufacturing process is very
reliable and a very precise control over the tip diameter is pos-
sible. Magnetic features down to 30 nm could be observed in
a CAMST reference sample, using a tip with a 50
cross section. The authors believe that this technique will allow
toproduceMFM tipswiththeideal tipshapeanda crosssection
Because the concept of the new probe is centered around the
use of a thin film as a support structure one could imagine other
types of sensors being integrated on the tip plane. For magnetic
imaging, a magnetoresistive element could, for instance, be re-
alized on the tip plane which would make high-resolution mag-
netoresistance microscopy possible.
 L. Abelmann, S. Porthun, M. Haast, C. Lodder, A. Moser, M. E. Best,
P. J. A. Vanschendel, B. Stiefel, H. J. Hug, G. P. Heydon, A. Farley, S.
R. Hoon, T. Pfaffelhuber, R. Proksch, and K. Babcock, “Comparing the
resolution of magnetic force microscopes using the CAMST reference
samples,” J. Magn. Magn. Mat., vol. 190, no. 1–2, pp. 135–147, 1998.
 L. Folks, M. E. Best, P. M. Rice, B. D. Terris, D. Weller, and J. N.
Chapman, “Perforated tips for high-resolution in-plane magnetic force
microscopy,” Appl. Phys. Lett., vol. 76, no. 7, pp. 909–911, 2000.
 S. Porthun, L. Abelmann, and C. Lodder, “Magnetic force microscopy
ofthinfilm mediafor highdensitymagnetic recording,”J.Magn.Magn.
Mat., vol. 182, no. 1–2, pp. 238–273, 1998.
Phys. A—Mater. Sci. Processing, vol. 66, pp. S1185–S1189, 1998.
 T. R.AlbrechtandC.F. Quate,“Atomicresolutionwiththe atomicforce
microscope on conductors and non conductors,” J. Vac. Sci. Technol.
A—Vac. Surf. Films, vol. 6, no. 2, pp. 271–274, 1988.
 P. B. Fischer, M. S. Wei, and S. Y. Chou, “Ultrahigh resolution magnetic
Sci. Technol. B, vol. 11, no. 6, pp. 2570–2573, 1993.
 G. D. Skidmore and E. D. Dahlberg, “Improved spatial resolution in
magnetic force microscopy,” Appl. Phys. Lett., vol. 71, no. 22, pp.
 M. Ruhrig, S. Porthun, J. C. Lodder, S. McVitie, L. J. Heyderman, A. B.
Johnston, and J. N. Chapman, “Electron beam fabrication and charac-
terization of high-resolution magnetic force microscopy tips,” J. Appl.
Phys., vol. 79, no. 6, pp. 2913–2919, 1996.
 G. N. Phillips, L. Abelmann, M. Siekman, and J. C. Lodder, “High res-
olution magnetic force microscopy using focussed ion beam modified
tips,” Appl. Phys. Lett., 2000, to be published.
 L. Meirovitch, Fundamentals of Vibrations, International ed, ser. Me-
chanical Engineering.New York: McGraw-Hill, 2001.
 M. Vangbo and Y. Bäcklund, “Precise mask alignment to the crystallo-
graphic orientation of silicon wafers using wet anisotropic etching,” J.
Micromech. Microeng., vol. 6, pp. 279–284, 1996.