AFM and u-XRD capabilities on ID01

Layout and capabilities

The in-situ AFM recently developed at the ESRF was designed having the particular issues and geometries of synchrotron endstation in mind.

A schematic overview of the in-situ AFM is given above. The main frame consists of two side plates hosting the connectors and two big Kapton windows enabling almost unlimited optical access. The sample is placed on a magnetic holder on a piezo scanner with a scan range of approximately 40×40×24μm (XYZ). Besides the automatic scanning and feedback modes, the position of the sample underneath the tip can be controlled ''manually'' (i.e. the translations stages are not controlled by SPEC) in steps of around 2nm.

The scanner table (red in figure 1) is placed atop the Attotower (black, grey and blue in figure 1), a stack of 3 piezo stick-slip stepper motors for the coarse positioning in x, y and z-direction. The total travel range of these motors is given to 4mm with a minimum step size of ca. 40nm.

As a tip, an electrochemically etched tungsten wire attached to a quartz tuningfork (TF), which serves as the force sensor, is used. The TF is excited mechanically by a piezo element attached to the lower side of the tip-holder. The use of a free standing TF enables the measurement of frequency changes of 5 orders of magnitude: for standard AFM-imaging, changes of less than 1Hz occur, while shifts of more than 10kHz are measured when a preasure is applied to a stiff object (such as SiGe islands or other compact nanostructures).

The tip-holder is mounted on one of the side plates which provides the connectors for the tuning fork excitation and read-out. The very weak signal from the tuning fork is lead through a pre-amplifier and then sent to a Lock-In amplifier triggered by the excitation frequency. The second side plate holds connectors for the scanning table, the Attotower and additionally a bias signal which can be applied to the sample directly.

Additionally, the casing features two Swage-Lock connectors: a slow gas flow of He during the experiment yields temperature stability. Furthermore, the incident X-ray beam would produce ozone by ionizing oxygen in the ambient atmosphere which would attack the sample surface. This effect is reduced in a He-atmosphere.

 
Integration

The AFM is mounted on the vertical rotation stage (Hphi) by using an L-shaped holder (figure 3) with an integrated vertical translation stage (8mm travel, stepsize 1μm). The Hphi stage is directly mounted on the Huber-tower giving access to all degrees of freedom provided by its translation and tilting stages.

The AFM is controlled by a PC connected via USB to the controller hardware. The PC has to reside close to the AFM. Therefore, it is controlled from outside the experimental hutch by a network remote control. Drivers which allow the direct control of the AFM mtors from a SPEC session are being developed by attocube at the moment but are not ready to use yet. This solution is epxected to be available by the middle of 2009.

The operation of the AFM requires an optical microscope for coarse positioning the sample underneath the tip. The beamline is permanently equipped with an inline telescope pointing at the diffractometers center of rotation and including a small angle with the X-ray beam. A second sample microscope can be mounted directly above the AFM and fixed to the eta-rotation stage allowing it to be rotated to a position with sufficient optical access.

Above, a photograph of the big vacuum vessel, the diffractometer with the AFM integrated. Below, scheme of AFM integration in the beamline.

Photocurrent Imaging

The tungsten-probe of the in-situ AFM gives access to a novel alignment and imaging technique called photocurrent imaging (PI). The principle of this method is the measurement of photon induced currents in the AFM-tip. Photons of the X-ray beam penetrate the AFM-tip and ionize the tungsten atoms. An induced current of several pA overlays the signal from the piezoelectric effect in the TF which is used to derive the amplitude of oscillation. After the pre-amplifier of the AFM, the signal is read out by a separate Lock-In amplifier. A mechanical chopper is placed in the X-ray beam path before the sample (see figure 5). Its frequency is controlled by the Lock-In amplifier and serves as the reference frequency to filter the photo current signal. The readout of this signal is integrated into SPEC which allows an easy readout of the photocurrent.

The combination of PI, topographic AFM imaging and Scanning X-ray Diffraction Microscopy (SXDM) allows an alignment of the sample with respect to the tip and the incident X-ray beam to a very high accuracy. The figure above shows a comparison between a combined PI-SXDM map of a SiGe-island sample (left image) and an SEM image of the same area (right image). The PI-signal (red) shows the exact position of the tip with respect to the sample (SiGe islands, blue). This in turn allows to approach specific nanostructures chosen from prior ex-situ measurements (e.g. SEM).

Nanoindentation

The main goal of using an AFM for in-situ measurements is the ability to interact with the sample while probing it with the X-ray beam. One possible interaction is the application of a pressure onto the sample. This can be done employing the AFM tip as an indentation device. The stiffness of the TF thereby provides the load constant. However, controlling the exact amount of pressure in this environment is extremely difficult and up to now, the accuracy is inferior to lab-based specialized nanoindentation devices.

A thin AFM-tip would be too fragile for indentation experiments. Therefore, blunt W-tips are used (see left figure). The apex radius of curvature of these blunt tips ranges from several hundred nm up to several μm, depending on the actual experiment.

Future developments involve the integration of a diamond indenter-tip into the AFM-setup allowing a better control over the applied pressure and the indentation of harder materials.