K

Kelvin Probe Force Microscopy an electrically conducting tip, which acts as a movable
top electrode that can locally set a potential difference
Bharat Bhushan and Manuel L. B. Palacio between the tip and the sample of interest. As the
Nanoprobe Laboratory for Bio- & Nanotechnology cantilever tip is scanned above the sample, the tip can
and Biomimetics, The Ohio State University, sense the response to the electrical potential. An
Columbus, OH, USA advantage of conducting electrical AFM is that the
topography and electrical properties can be obtained
simultaneously during the experiment [2].
Synonyms The contact, intermittent (tapping) and noncontact
operating modes, along with the use of a conductive
Scanning Kelvin probe force microscopy; Scanning tip, extended the application of AFM from topographic
surface potential microscopy imaging to electrical surface characterization [2–4, 6].
The ability to detect the surface potential difference
between the tip and sample, with the tip lifted a certain
Definition distance above the sample, has led to Kelvin Probe
Force Microscopy (KPFM). The ability for capacitance
Kelvin probe force microscopy (KPFM) is an atomic measurements, such as in n-type silicon, led to Scanning
force microscopy (AFM)-based technique used to Capacitance Microscopy (SCM). Deflection detection
measure the surface potential and work function prop- for electromechanical response of ferroelectric films led
erties of a material surface interacting with a conduc- to Piezoresponse Force Microscopy (PFM). The detec-
tive AFM probe. tion of current through oxide films led to Conductive
AFM (CAFM) and on doped samples as a function of
the load led to Scanning Spreading Resistance Micros-
Overview copy (SSRM). The studies of current flowing into con-
strictions in two-dimensional semiconductor structures
The applications of atomic force microscopy (AFM) led to Scanning Gate Microscopy (SGM). The detection
have continuously evolved to include the measurement of frequency dependence force applying a voltage
of properties aside from surface topography. To date, the across a sample in BiFeO3 and p-doped silicon led to
AFM is widely used for the characterization of surface/ Scanning Impedance Microscopy (SIM). A recently
near-surface properties such as friction, adhesion, reported new technique to map capacitance, impedance,
scratch, wear, fabrication/machining, electrical proper- and dielectric constant variations is called Scanning
ties, local deformation, nanoindentation, local surface Microwave Microscopy (SMM). This technique studies
elasticity, viscoelasticity, and boundary lubrication. the electromagnetic tip–sample interactions using
The mapping of surface or near-surface electrical a specially designed conductive probe to send and
properties using the AFM is possible through the use of receive microwave signals up to GHz frequencies.

B. Bhushan (ed.), Encyclopedia of Nanotechnology, DOI 10.1007/978-90-481-9751-4,

# Springer Science+Business Media B.V. 2012
K 1174 Kelvin Probe Force Microscopy

Kelvin Probe Force a b

Microscopy, φtip>φsample
Fig. 1 Schematic of the Vdc= −ΔΦ +
KPFM principle: (a)
electrostatic potential and AFM tip AFM tip
interaction force between −
a conducting tip and a sample force ΔΦ=(φtip−φsample)/e no force
(for illustration, assume that
+ + + + +
ftip> fsample), (b) external DC Sample Sample
voltage applied to nullify the
force, and (c) external AC
voltage with adjustable DC
offset is applied to the tip c
which leads to its vibration [5]
V=Vdc+Vacsin(ωt)

AFM tip

F =Fdc+Fω+F2ω

Sample

The focus of this entry is on Kelvin probe force where Ftip and Fsample are work functions of the tip and
microscopy. Since the surface potential measurements the sample, respectively, and e is the magnitude of the
are highly sensitive to surface charge variations, charge of one electron. The contact potential differ-
KPFM is widely used in detecting the dopant concen- ence provides material contrast information, because it
tration and the dopant type (either p-type or n-type) on is sensitive to adsorbed molecules on the surface or the
semiconductor materials. Non-semiconductor applica- presence of various phases within a given material.
tions of KPFM include the characterization of adsor- Electrostatic force is created between the tip and sam-
bates on Si substrates, intermetallic grain formation on ple under the influence of this surface potential differ-
metal films, nanotribology studies, and the degradation ence and the separation-dependent local capacitance C
of inorganic materials for batteries, among others. of the tip and sample. This force is given by

1 @C
F ¼ ðDFÞ2 (2)
Kelvin Probe Force Microscopy Principles 2 @z

Kelvin probe force microscopy is used to measure the where z is the distance between the tip and sample.
surface potential with the AFM in noncontact mode, In KPFM, an AC voltage signal (VAC sin(ot)) with
and is based on the Kelvin probe principle [3, 4, 10]. To an adjustable offset DC voltage (VDC) is applied to the
describe the operating principle behind KPFM, con- conducting tip (Fig. 1b and c). As shown in (Eq. 3), the
sider a tip and sample interaction as shown in Fig. 1. electrostatic force is the sum of three components,
When the tip and sample are electrically connected namely, the DC force, a harmonic force with the
(Fig. 1a), electrons flow from the material with the same frequency of the AC tip bias (the first harmonic),
lower work function to the material with the higher and a harmonic force at double the frequency of the AC
work function. Due to the difference in the work func- signal (the second harmonic):
tion of the electrically connected tip and the sample, an
electrostatic contact potential difference (or surface

1 @C VAC 2 @C
F¼ ðDF þ VDC Þ2 þ þ ðDF þ VDC ÞVAC sinðotÞ
potential) is created between the tip and the sample. 2 @z 2 @z
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl} |fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl}
The value of this contact potential difference (DF) is DC term o term

given by the following equation: 1 @C

 VAC 2 cosð2otÞ
4 @z
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl}
Ftip  Fsample
DF ¼ (1) 2o term
e (3)
Kelvin Probe Force Microscopy 1175 K
Kelvin Probe Force Feedback
Microscopy,
Fig. 2 Schematic of the
interleave scan method used in

Laser
KPFM, where the first pass Lock in
records the height profile amplifier

Ph ctor
de
performed in tapping mode,

oto
te
Potential
and the second pass in lift
X-Y-Z signal
mode monitors the variations
piezo Computer
of electrostatic force at
a constant height [12]
Servo controller
DC Voltage (DC single
adjusted to
Sample zero lock-in
signal)

The electrostatic force is used to adjust the exter- using a frequency modulation technique instead of the
nally applied voltage in order to compensate the con- amplitude modulation method. The main difference
tact potential until the force between the tip and the between these two methods is that amplitude modula-
sample is nulled. This is attained when the difference tion is sensitive to the electrostatic force, while fre-
between the work functions of the tip and sample is quency modulation is sensitive to its gradient. Instead
equal to the external voltage. The presence of different of having one feedback control/lock-in amplifier sys-
K
materials on the sample surface leads to the formation tem that sequentially records the topography and
of different material interfaces (with the tip), each surface potential, a separate lock-in amplifier is used
having their own value of the contact potential differ- in the frequency modulation KPFM experiment to
ence. This leads to surface potential contrast [9]. detect surface potential. The frequency modulation
In the initial implementation of KPFM, the topog- method has been shown to measure a potential differ-
raphy and potential are measured simultaneously using ence in close agreement with macroscopic measure-
two feedback circuits [13]. Subsequent modifications ments, and is thus regarded as a more quantitative
to the setup involve the sequential measurement of potential measurement technique compared to ampli-
topography and potential in the so-called lift mode, tude modulation [16].
which is regarded to improve on the lateral resolution As mentioned earlier, KPFM is widely used in
of the topographic image [9]. A schematic of the two- characterizing the doping of semiconductors, and it
pass lift or interleave scan method used in KPFM can also be utilized for non-semiconductor materials.
experiments is shown in Fig. 2. On the first pass, Examples of KPFM applications are described next.
surface height information is recorded in the standard
tapping mode where the oscillating cantilever lightly
taps the surface. On the second pass, the feedback loop Applications of KPFM
controlling the vertical piezo is turned off, and the tip is
lifted from the surface and traced following the surface An example of the applicability of KPFM to semicon-
profile acquired in the first pass and at constant tip– ductor materials is shown in Fig. 3, which is a surface
sample separation distance [7, 8]. An oscillating poten- potential image of a pn junction on a silicon wafer
tial is applied to the DC-biased tip (the tip oscillation created through ion implantation. Below the image is
is electrically driven). The oscillating potential a corresponding cross section of the scan defined by the
frequency is equal to the cantilever’s resonance fre- line A-B on the image [11]. The contrast in the surface
quency. An extra feedback loop is used to adjust the potential image is induced by the charge carriers intro-
DC bias on the tip to nullify the first harmonic compo- duced by the ion implantation process. The brighter
nent of the electrostatic force [5, 9]. features correspond to areas with a higher surface
A more recent implementation of the KPFM exper- potential. In this case, the surface potential in the
iment involves the detection of the surface potential p-type region is higher than in the n-type region.
K 1176 Kelvin Probe Force Microscopy

This is attributed to the bending of energy bands at the

subsurface as a result of the trapping of charge carriers
in surface states [11].
KPFM has been used in the characterization of
perfluoroalkyl alkanes self-assembled on substrates
such as graphite, mica, and silicon. As an example,
Fig. 4 shows images of F14H20 adsorbates on a Si
substrate [1]. Topography and surface potential images
A B
of the adsorbate layers show clusters in Fig. 4a and b.
In the surface potential image, the difference between
the adsorbates and silicon shows a high contrast.
In Fig. 4c and d, the profiles of the white lines marked
in Fig. 4a and b are displayed. The profiles indicate
that there is a large surface potential drop of  0.5 V
1.2
within the adsorbate clusters as a result of possible
Δφ

repulsion of the adsorbate composition. A correlation

0.8 between the surface potential and the topography
1 μm
image suggests that the surface potential corresponds
to the adsorbate clusters. The magnified images in
Kelvin Probe Force Microscopy, Fig. 3 Surface potential Fig. 4e and f also revealed a particular shape in the
image of a silicon pn junction (above), with the bright and dark cluster, spiral structures. In the example KPFM
areas corresponding to the p-type and n-type regions, respec-
tively. The cross section (below) refers to the scan defined by the
images, the capacity of the technique to detect differ-
line A-B on the image, where Df is the surface potential relative ent surface potentials on the sample is shown.
to the probe surface potential [11] Figure 5 shows a KPFM application to recognize
constituent materials in a metal alloy film [14]. In this

a b

Kelvin Probe Force

Microscopy, Fig. 4 KPFM
applied to studies of F14H20
adsorbates on Si: (a) 4 c e
topography, (b) surface 3
potential image of F14H20 2
adsorbates on Si, (c) and (d) 1
the cross-section profiles taken 0
along the directions marked 0 0.5 1 1.5 2 2.5 μm 800 nm
V
with the dashed white lines in 1
the images in (a) and (b), 0.5
d
f
respectively, and (e) and (f) 0
zoomed-in topography and
–0.5
surface potential images on
one of the locations shown in 0 0.5 1 1.5 2 2.5 3 μm
(a) and (b), respectively [1] 800 nm
Kelvin Probe Force Microscopy 1177 K
After wear test
Surface height Surface potential
Uncoated Silicon

Z-TETRAOL (Untreated)
5 μm a

K
Z-TETRAOL (Fully bonded)

5 μm
b

Kelvin Probe Force Microscopy, Fig. 5 KPFM study of var-

ious intermetallics in a metal alloy: (a) surface potential image of
Al/Cu film showing contrast between Al and the Al/Cu interme- 0 10 μm 0 10 μm
tallic (the dark points); the scan field is 40 mm  40 mm, (b) AFM
topography image [14]

0 20 nm 0 1V

example, an alloy consisting of Al and Cu were depos- Kelvin Probe Force Microscopy, Fig. 6 Surface height and
ited as 5 mm lines on a silicon substrate. The lines in the surface potential maps after wear tests for an uncoated silicon
contact potential difference image in Fig. 5a contain wafer and two types of Z-TETRAOL-lubricated silicon wafers
(untreated and fully bonded). The brighter areas correspond to
features that are not present in the topographic image higher values of both the height and surface potential
shown in Fig. 5b. In the potential image, the dark
features contained within the lines correspond to the
Al/Cu intermetallic that forms on the grain boundaries, Fig. 6, surface height and surface potential data (taken
which has a lower contact potential difference relative after nanoscale wear tests were performed) are shown
to the surrounding Al regions. The clear contrast for silicon surfaces coated with the perfluoropolyether
between the Al/Cu and Al shows the capacity of lubricant Z-TETRAOL. The lubricant is applied using
KPFM as a technique to identify the presence of two different techniques, such that the samples are
different materials on alloy surfaces at the nanoscale. either untreated or fully bonded. Data for the uncoated
KPFM can also be applied to nanotribological stud- silicon surface subjected to the same wear test proce-
ies, specifically for studies involving wear initiation. In dure is shown for reference [15].
K 1178 Kelvin Probe Force Microscopy

Unaged Aged
Unaged Aged Unaged Aged, fit
+1.0 V +1.0 V +1.0 V
1.2 20

1.1 15

f (x)
1.0 10

0.9 5

0.8 0
0 5μm 0 5μm 0.7 0.8 0.9 1.0 1.1 1.2
Surface potential (V)

Kelvin Probe Force Microscopy, Fig. 7 Surface potential of the surface potential for the unaged and aged samples, indi-
map of the unaged (left column) and the aged (middle column) cating that the mean value of the surface potential decreases
LiFePO4 cathode samples with external voltage of + 1 V. The after aging
right column shows the normal probability density distribution

The surface height and surface potential images in applied voltage. This indicates that under the external
Fig. 6 show contrast on both sets of images. Compared source the sample tends to charge uniformly even in an
to the uncoated silicon surface, reduction in the debris aged condition. The uniform charging is good for the
is observed in the partially bonded samples. This indi- cell as it avoids large local currents and subsequent
cates that the lubricant is able to provide wear protec- damage to the cathode. The last column of Fig. 7 shows
tion. The pileup of debris observed in the untreated the distribution of the collected data points in unaged
sample could be a combination of removed fragments and aged cathodes. Then a normal probability density
of Si as well as the weakly bound lubricant. The function is used to fit the histogram data. The unaged
untreated sample with its mobile lubricant fraction sample had mean values of the surface potential almost
exhibits a lower surface potential change compared to equivalent to the external applied voltage. The mean
the fully bonded sample, which only has immobile values of surface potential on the aged sample are
lubricant molecules. This is attributed to lubricant lower than that of the unaged sample. Thus, even
replenishment by the mobile lubricant molecules pre- though the externally applied voltage was the same
sent on the untreated sample. In general, a considerable for the unaged and aged samples, the charge sustained
surface potential change is observed on the wear region on the surface of the aged cathode is less than that
when: (1) the lubricant has been fully removed from sustained on the unaged cathode [12].
the substrate; (2) the native SiO2 layer has been The work function of the tip is constant in each
abraded from the surface; (3) wear has caused subsur- surface potential map, since the same kind of tip is
face structural changes; and (4) charges build up as the used in both samples. Therefore, Fig. 7 shows the
charges are unable to dissipate into the surrounding change in the work function of the aged sample relative
material [15]. to the unaged sample. The decreased surface potential
Finally, the application of KPFM in studying mate- in the aged sample is an indication of surface modifi-
rial degradation is illustrated for the aging of lithium- cation as a result of the phase change of the cathode
ion batteries. The surface potential maps of the unaged material upon degradation [12]).
and aged LiFePO4 cathode samples are shown in Fig. 7
[12]. The maps were collected by applying +1.0 V
from an external DC voltage source. On the surface Cross-References
potential maps for both the unaged and aged sample,
a small difference in the contrast is observed, ▶ AFM Probes
suggesting an almost uniform dissipation of charge ▶ AFM, Tapping Mode
over the surface of the samples under the externally ▶ Atomic Force Microscopy
Kinetic Monte Carlo Method 1179 K
References Kelvin prone force microscopy. Appl. Phys. Lett. 66,
3510–3512 (1995)
1. Alexander, J., Magonova, S., Moeller, M.: Topography and 12. Nagpure, S.C., Bhushan, B., Babu, S.S.: Surface potential
surface potential in Kelvin force microscopy of measurement of aged Li-ion batteries using Kelvin probe
perfluoroalkyl alkanes self-assemblies. J. Vac. Sci. Technol. microscopy. J Power Sources 196, 1508–1512 (2011)
B 27, 903–911 (2009) 13. Nonnenmacher, M., O’Boyle, M.P., Wickramasinghe, H.K.:
2. Avila, A., Bhushan, B.: Electrical measurement techniques Kelvin probe force microscopy. Appl. Phys. Lett. 58,
in atomic force microscopy. Crit. Rev. Solid State Mater. 2921–2923 (1991)
Sci. 35, 38–51 (2010) 14. O’Boyle, M.P., Hwang, T.T., Wickramasinghe, H.K.:
3. Bhushan, B.: Springer Handbook of Nanotechology, Atomic force microscopy for work functions on the nano-
3rd edn. Springer, Heidelberg, Germany (2010) meter scale. Appl. Phys. Lett. 74, 2641–2642 (1999)
4. Bhushan, B.: Nanotribology and Nanomechanics I – 15. Palacio, M., Bhushan, B.: Ultrathin wear-resistant ionic
Measurement Techniques and Nanomechanics, II – liquid films for novel MEMS/NEMS applications. Adv.
Nanotribology, Biomimetics, and Industrial Applications, Mater. 20, 1194–1198 (2008)
3rd edn. Springer, Heidelberg (2011) 16. Zerweck, U., Loppacher, C., Otto, T., Grafstroem, S.,
5. Bhushan, B., Goldade, A.V.: Measurements and analysis of Eng, L.M.: Accuracy and resolution limits of Kelvin probe
surface potential change during wear of single-crystal sili- force microscopy. Phys. Rev. B 71, 125424 (2005)
con (100) at ultralow loads using Kelvin probe microscopy.
Appl. Surf. Sci. 157(2000), 373–381 (2000)
6. Bonnell, D.: Scanning Probe Microscopy and Spectroscopy
Theory, Techniques, and Applications, 2nd edn. Wiley –
Key Point Detection
VCH, New York (2001)
7. DeVecchio, D., Bhushan, B.: Use of a nanoscale Kelvin ▶ Waypoint Detection
probe for detecting wear precursors. Rev. Sci. Instrum. 69,
3618–3624 (1998)
8. Elings, V., Gurley, J.: US Patent No. 5,266,801 1993 K
9. Jacobs, H.O., Knapp, H.F., Muller, S., Stemmer, A.: Surface
potential mapping: a qualitative material contrast in SPM. Kinetic Monte Carlo Method
Ultramicroscopy 69, 39–49 (1997)
10. Kelvin, L.: Contact electricity of metals. Phil. Mag. 46,
▶ Computational Study of Nanomaterials: From
82–120 (1898)
11. Kikukawa, A., Hosaka, S., Imura, R.: Silicon pn junction Large-Scale Atomistic Simulations to Mesoscopic
imaging and characterizations using sensitivity enhanced Modeling