(030) 8062 - 43199
UHV-KPFM Setup and Results
Content
Setup of the UHV-KPFM
Working Principle
When two materials with different work function are connected, electrons will flow from the material with the lower work function to the one with the higher work function, until their Fermi-levels are equilibrated. The application of a voltage between the two materials will shift the Fermi-levels apart again. The voltage necessary to bring the local vacuum levels of the two materials to the same energy corresponds to the contact potential difference (CPD). This principle can be used macroscopically as well as microscopically to determine the CPD of two materials.
The KPFM is based on a commercial AFM setup in UHV. The microscope is operated in the regular non-contact detection mode. The cantilever is oscillated at its resonance frequency and scanned across the sample at a constant distance by maintaining a certain frequency shift to the free resonance frequency (frequency modulation mode, FM mode). To detect the oscillation, a laser beam is reflected off the back of the cantilever onto a position sensitive photo detector. A voltage is applied between the sample and tip, consisting of a dc part Udc and an ac-voltage Uac at the second resonance frequency f2 of the cantilever to allow a resonant and independent detection of electrostatic forces. The application of this voltage results in an additional electrostatic force, which can be split up into three contributions. The dc part contributes to the topography (red color will be used to represent topography). The term at f2 is used to measure the CPD (green will be used to represent CPD images). A third term at 2f2 can be used to do capacitance microscopy. During the scan Udc will be adjusted so that the electrostatic forces between the tip and the sample become zero and thus the oscillation amplitude of the cantilever at the frequency f2 becomes zero. Since the electrostatic force at f2 depends on (Udc - Δ Φ/e), Udc corresponds to the CPD. If the work function of the cantilever is known, the work function of the sample can be deduced. Thus, in one scan, the topography and the CPD of the sample are determined simultaneously.
Figure 2: Schematic drawing of the working principle of the Kelvin Probe Force Microscope. (a / see left) Representation of the tip and sample with the oscillation detection system and the sample bias. (b) Block diagram of the measurement system.
Resolution of the KPFM
To determine the energetic resolution of the system a dc-voltage source is connected in series with the active Kelvin controller, so that both, the actual CPD to the sample and the applied voltage have to be compensated. The output of the controller is a straight line with the slope 1, as expected. The peak to peak noise level is in the range of 5 meV. This noise level is obtained with an ac-amplitude as low as 100 mV, which allows the application of the measurement method to semiconductor materials. Even if different cantilevers of the same type are used, a reproducibility of the measured CPD values within 200 mV is found.
To determine the lateral resolution a test sample was prepared, i.e. C59N was evaporated on HOPG (highly oriented pyrolytic graphite). The C59N grows in large islands as seen from the topography picture. From the line profile a lateral resolution for the topography of about 3 nm can be determined. In the CPD image we see the chemical contrast of the two materials with the higher CPD of the C59N. In the line profile the width of the CPD-transition between the two materials is about 20 nm. The lower resolution in the work function measurement is due to the long range electrostatic forces on which the measurement of the CPD is based.
Figure 3: Measurement of the contact potential difference with an additional dc bias varied from - 1 to 1 V. The inset shows that the noise level in the measurement is about 5 meV.
Figure 4: Measurement on a C59N film on a graphite substrate. The topography (a) shows the C59N island on the graphite substrate and the CPD image the chemical contrast between the two materials. The resolution for topography and CPD measurements can be seen in the line scans.
Some results
In the first experiments semiconductor surfaces were studied after cleavage in UHV, to ensure clean surfaces. The UHV cleaved (110) surface of n-doped GaAs is free of surface states and thus is in flat band condition after the cleavage. Figure 5 (a) shows the topography with a two mono layer deep groove running across the sample surface. In the CPD image a higher CPD at the steps can be observed (in all pictures dark contrast corresponds to a lower value and bright contrast to a higher value). The higher work function at the steps can be explained as follows: there are surface states at the steps which result in an upward band bending for the n-doped material. This results in a larger work functions
Another example of an application is the imaging of single charged dopant sites in p-type WSe2 UHV cleaved at the (0001) surface. The topography image shows an almost flat surface but the CPD image exhibits some bright and dark contrast, which is caused by charged dopant sites in the material. The occurrence of both, acceptors and donors is in accordance with the degree of compensation and the doping concentration as determined by Hall measurements.
Figure 5: Work function variation at a 2-monolayer step on the n-GaAs (110) surface (a) topography (400 ´ 300 nm², z = 0.45 nm); (b) corresponding CPD-image (Δ Φ = 45 meV).
Figure 6: Dopant sites on p-WSe2(0001): (a) topography (300 ´ 200 nm², z = 0.22 nm); (b) corresponding CPD-image (CPD = 0.928 - 0.940 eV).
