Patent Application
20170307655
The present invention relates to an scanning probe microscope. The present invention further relates to a method of operating an scanning probe microscope.
Scanning probe microscopes (SPM), such as Atomic Force Microscopes (AFM) are widely used for the physical characterization of materials and devices when high spatial resolution and small feature sizes are of interest. AFMs are primarily used in imaging modes to provide topographic information, but they can also record the force between a probe tip and a sample. In AFM, the probe tip is typically positioned on a cantilever spring, which serves as a flexible carrier that translates force in bending or displacement. Other flexible carriers may also be used for this purpose instead of a cantilever spring, such as diaphragms or double clamped beams. Movement of the flexible carrier is typically measured using an optical beam deflection setup, although different read-out schemes also exist. In the following description the wording ‘cantilever’ is used as a particular example of a flexible carrier. The disclosure is however equally applicable to embodiments using another type of flexible carrier, as suggested above.
In the common imaging mode ‘Amplitude Modulation’ AFM (also called ‘tapping mode’, ‘AC mode’ or ‘intermittent contact mode’), the cantilever is oscillated at or near its resonance frequency. The amplitude of this oscillation is reduced by the force between tip and surface and is measured and used as input for a feedback system which adjusts the height of the tip such that the amplitude remains constant. When the tip is scanned over the surface, the topographical variations between positions continuously change the amplitude which the feedback then corrects for. The speed of correction is limited by a number of factors, including the cantilever response time and the speed at which the cantilever can be mechanically repositioned.
Cantilever response time is determined by the cantilever resonance frequency and damping or Q factor. In this connection people have worked both at increasing the resonance frequency and decreasing the Q factor to speed up AFM imaging. The first mention in literature of reducing Q for enhanced speed is Mertz et al. in “Regulation of a microcantilever response by force feedback”, Applied Physics Letters, 1993, 62, 2344, which introduces the idea to feed back the measured cantilever motion in an external drive signal to reduce Q. Other schemes have been developed to reduce Q too, for example, Fairbairn et al. describes the use of an active impedance in a piezoelectric shunt control framework in “Sensorless Implementation of a PPF Controller for Active Q Control of an AFM Microcantilever”, IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY. Fairbairn compares results achieved with a quality factor Q of 226 and with a reduced cantilever Q factor of 17 and concludes that reducing the cantilever Q factor with active piezoelectric shunt control allows for improved image quality at high scan speeds.
It is noted that US20070024295 discloses a probe for an atomic force microscope is adapted such that, as a sample is scanned, it experiences a biasing force urging the probe towards the sample. This improves probe tracking of the sample surface and faster scans are possible. This is achieved by either including a biasing element which is responsive to an externally applied force, on the probe and/or reducing the quality factor of a supporting beam. The quality factor may be reduced by coating the beam with a mechanical-energy dissipating material. It is a disadvantage of this known probe that the coating causes a relatively substantial mechanical stress and therewith a distortion of the beam. Moreover, this mechanical stress and resulting distortion depend on external circumstances, such as temperature, which render the probe less accurate. Also non-uniformities and defects in the coating can easily result in deviations of the mechanical properties of the probe.
It is an object of the present invention to provide for an improved image quality at high scan speeds or to enable an increase of the scan speed while maintaining image quality by reducing the Q factor of the probe, while avoiding that this results in deviations of its other mechanical properties.
According to a first aspect of the present invention a scanning probe microscope is provided that comprises:
a scanning probe comprising a flexible carrier, the flexible carrier having a movable part provided with a tip;
a holder to hold a sample in an environment free from liquid;
a scanning arrangement to induce a relative motion of said scanning probe with respect to said sample along a surface of said sample;
a driver to generate a drive signal to induce an oscillating motion of said scanning probe relative to said surface,
a measuring unit to measure a deflection of the scanning probe, and to provide a deflection signal indicative for said deflection;
an amplitude detector to detect an amplitude of said oscillating motion as indicated by the deflection signal and to provide an amplitude signal indicative for said amplitude.
The scanning probe is accommodated in a casing comprising a liquid to dampen motion of said scanning probe, the uncontrolled scanning probe therewith having a quality factor Q which is less than or equal than 5, the tip of said scanning probe extending through an opening in said casing.
According to a second aspect of the present invention, a method of operating a scanning probe microscope is provided that comprising:
Accommodating a scanning probe in a casing comprising a liquid to dampen motion of said scanning probe, the uncontrolled scanning probe therewith having a quality factor Q which is less than or equal than 5, the scanning probe comprising a flexible carrier, the flexible carrier having a movable part provided with a tip, which tip extends through an opening in said casing;
holding a sample in an environment free from liquid;
inducing a relative motion of a scanning probe with respect to a sample along a surface of said sample;
inducing an oscillating motion of said scanning probe relative to said surface;
measuring a deflection of the scanning probe;
detecting an amplitude of said oscillating motion as indicated by the deflection signal.
In the method and the apparatus according to the present invention, the liquid in which the scanning probe is arranged provides for a substantial damping of the scanning probe. I.e. the liquid damped scanning probe has quality factor Q which is less than or equal than 5. This enables a fast response. Due to the presence of the liquid in which the probe is arranged a additional dampening coating of the probe is superfluous, therewith avoiding an increase in mechanical stress and unpredictable variations in other mechanical properties. It is noted that application of scanning probe microscopy is known wherein the sample is immersed in a liquid for protection. For an important class of applications, e.g. inspection of nanostructured products such as wafers or masks in the semiconductor industry these known devices are however unsuitable. Contact between a liquid, possibly including contaminations dissolved therein, is typically harmful to such products. Accordingly, in the apparatus and method as claimed the sample is maintained in an environment free from liquid. The flexible carrier is for example a cantilever having the tip provided on a free end. Alternatively, the flexible carrier may be diaphragm having the tip provided in a central position of its surface, or a double clamped beam having the tip arranged in a central position between its ends. The casing around the cantilever further enhances damping and therewith makes it possible to reduces the Q factor more than would be possible when the cantilever were not encased.
It is noted that use of encased cantilevers is known wherein the casing is filled with a gas. These serve to avoid contact of the cantilever with a liquid wherein the sample is arranged and not to dampen the cantilever. In order to sufficiently dampen the cantilever by the gas in such an arrangement it would be necessary to dimension the cantilever with small distances between the edges of the cantilever and the casing. Minor deviations in these distances would already result in substantial deviations in the desired damping.
If desired an additional damping by other means may be applied, e.g. by active damping.
For relatively small scanning area of the sample, it may suffice to keep the scanning probe positioned at a fixed absolute position. It may further be contemplated to adapt the tip oscillation amplitude in response to an observed relative position between the surface and the probe. Note that the present invention also increases the speed with which this adjustment can be made.
An embodiment of the scanning probe microscope according to the first or second aspect further comprises a positioning unit for positioning said scanning probe in a direction transverse to said surface. This facilitates scanning of larger areas and/or tolerates larger misalignments of the sample.
In an embodiment of this embodiment a feedback unit is included to provide a control signal for controlling said positioning unit in order to minimize deviations in an amplitude of oscillating motion based on said amplitude signal and a reference signal.
Typically the amplitude of this oscillation is reduced by the force between tip and surface and is measured and used as input for the feedback unit, which adjusts the height of the tip such that the amplitude remains constant.
Various signals involved may be used to render an image of the sample surface.
For example, in an embodiment, the control signal determining the height of the probe, can be used to render the topography image. The input for the feedback, the amplitude of the tip, denoted as the amplitude signal or the error signal can be used to render a difference image.
The oscillating motion of the scanning probe may be imposed in various ways. Most common is to shake its support, for example by dither piezo. This will work but might also shake the casing and therefore make the cantilever response more complex as the motions couple through the liquid.
Alternatively, the cantilever can be driven directly, e.g. magnetically (after integrating a magnet or magnetic layer) or piezoelectrically (after integrating a piezoelectric layer) or optically (heating and using the bilayer [bimetal] effect) or by other methods that directly actuate the cantilever. These methods to drive the cantilever are preferred, as they act directly on the cantilever and do not excite other structures.
Water may be used as the liquid inside the casing. This is advantageous in that it is chemically inert, has relatively high surface tension causing it to be confined in the casing as well as a relatively low evaporation rate. Also other liquids may be contemplated, such as oils, aceton, IPA. In some cases a surface of the opening for the tip facing the liquid may be prepared to have a relatively high surface tension for said liquid, to enhance confinement of the liquid in the casing.
In an embodiment of the scanning probe microscope having the encased cantilever, the casing includes a second opening for insertion of the liquid, for example to compensate for evaporation of the liquid.
It could still further be contemplated to dimension the cantilever in a manner that further enhances damping by the liquid in which it is arranged. This may be achieved for example by increasing its width. A reduction of its resonance frequency would not be a problem, as long it remains high enough to avoid a limitation on the resolution with which the surface can be scanned. The upper limit in this respect is determined by the scanning speed divided by the resonance frequency.
These and other aspects are described in more detail with reference to the drawing. Therein:
Like reference symbols in the various drawings indicate like elements unless otherwise indicated.
The scanning probe microscope further includes a driver 30 to generate a drive signal Sd to induce an oscillating motion of the scanning probe 10 relative to the surface of the sample SMP. Typically, this oscillating motion is performed in a direction transverse to the surface of the sample SMP, however alternative embodiments are possible wherein the oscillating motion is performed in a direction parallel to the surface instead.
A measuring unit 40 is provided to measure a deflection of the scanning probe 10, and to provide a deflection signal S67 indicative for said deflection. Various techniques are suitable for this purpose, including capacitive, self-sensing techniques based on piezoelectric or piezoresistive effects, interferometry and optical beam deflection.
An amplitude detector 50 is provided to detect an amplitude of the oscillating motion as indicated by the deflection signal Sδ. In response the amplitude detector 50 provides an amplitude signal Sa indicative for this amplitude.
In the embodiment shown the scanning probe microscope further comprises a positioning unit 60 that positions the scanning probe 10 in a direction transverse to the surface of the sample SMP. To that end it issues a control signal Sz to a positioning actuator (not shown). More specifically the scanning probe microscope as shown also comprises a feedback unit 70 to provide a control signal Se for controlling the positioning unit 60. The feedback unit 70 is arranged to minimize deviations in an amplitude of the oscillating motion of the scanning probe 10 based on the amplitude signal Sa and a reference signal Sa, ref that is representative for a desired amplitude of the oscillating motion. As shown in
The liquid L in the casing 90 in which the scanning probe 10 is arranged dampens its motion. Therewith a quality factor Q of the scanning probe in an uncontrolled state is less than or equal than 5. The wording “uncontrolled state” is used to indicate that the presence of the liquid L alone already provides for this damping. Additionally the feedback system may provide for a further damping by electronic control means, as schematically shown in
Various signals involved may be used to render an image of the sample surface, as illustrated by
The signal S2 shows the intrinsic response of the probe, i.e. the amplitude of its tip in case the probe is maintained at a fixed height with respect to the scanning plane.
Signal S3 shows the height of the probe in case a feedback control is provided.
Signal S4 shows the amplitude in case a feedback control is provided.
The control signal S3 (Sz in
As further shown in
As shown in
As shown in
The casing 90 confines the liquid L, so that the holder 5 can hold the sample SMP in an environment free from liquid. A liquid L can be confined in the casing 90 if the opening 91 is sufficiently small. In practice capillary action (determined by the surface tension of the liquid L (e.g. water) will be sufficient to confine the liquid L inside the casing. This is determined by a combination of a negative surface tension of the liquid with respect to the casing and a positive surface tension of the liquid with respect to air in the environment. Some evaporation could occur dependent on the size of the opening 91 for the tip 12 and on the distance of the liquid L surface to the opening. However due to the small cross-sections involved it is expected that loss of liquid by diffusion and/or evaporation is negligible in the normal lifetime of a probe. Nevertheless, in an embodiment an inner surface of the casing 90 near the opening 91 that faces the liquid L may have a relatively large positive surface tension for said liquid in comparison to a remainder of said inner surface. Therewith the liquid is repelled near the opening 91 and attracted towards the remainder of the inner surface.
In an experimental setting the influence of damping on cantilever dynamics and maximum sensible pixel rate in Atomic Force Microscopy was characterized. The cantilever used was a BudgetSensors Multi75-G with manufacturers specifications: f0=75 kHz, k=3 N/m, l×w×h=225×28×3 μm fabricated from silicon with an aluminum reflex coating. Measurements were performed in water with a fully submerged cantilever and in air to obtain the Q factor, as a measure of the cantilever damping, in both conditions. To that end thermal motion spectra were obtained, as shown in
The table indicates apart from the damped resonance frequency (f0) and the quality factor Q, also the response time τ.
The experiment reveals that the response time τ can be substantially reduced by reduction of the quality factor Q of the cantilever caused by damping in the liquid L.
The implications for the achievable pixel rate were investigated in more detail. To that end the transfer function of the measurement system was determined by application of a noise signal modulating the cantilever amplitude, in the relevant frequency range up to the cantilever resonance frequency. The elements which contribute to the transfer function are the cantilever dynamics and the lock-in amplifier used to measure the cantilever amplitude. Thus, limiting factors that may show up in the final spectrum are:
limited frequency content of input
limited cantilever dynamics
limited bandwidth of lock-in
Noise input on the amplitude was generated by scanning a rough surface. The small features on the sample modulate the cantilever amplitude when the tip is scanned over them and thus they provide a noise input with a flat spectrum, up to some frequency which is determined by the smallest feature size on the sample and the scanning speed.
Results of these measurements are shown in
These measurements are summarized in the following table
The second column of the table shows the speed with which the surface of the sample is scanned. The third column shows the sample frequency in kS/s. The fourth column shows the highest frequency of the spectrum of the moving sample that can still be meaningfully observed. The fifth column shows the value of this frequency that is expected on the basis of the relation
The sixth column shows the value of the maximum possible observable frequency, being the resonance frequency of the cantilever. It can be seen that an increase of the scanning speed from 0.06 to 0.54 mm/s is expected to result in an increase of the highest observed frequency fd from 1 kHz to 9 kHz on the basis of a linear relation between those two. However as becomes apparent from the observations, the actually observed maximum frequency is substantially lower, i.e. implying that detailed information of the sample is lost when increasing the scanning speed to 0.54 mm/s. At a further increase of the scanning speed to 1.11 mm/s the actually observed maximum frequency does not increase further at all, implying an even further loss of information. The table also shows in a seventh column the approximate size (Size) of details that is still detectable. This is calculated as follows.
Therein dp is the pixel size, here 15 nm.
Similar measurements were performed for the cantilever damped by water instead of by air. The results are shown in
This second table shows that the observed maximum frequency is substantially increased, even in the first measurement in this series. I.e. with the scanning speed of 0.075 mm/s the observed maximum frequency fd equals 3 hKz, which is more than would be expected on the basis of the results obtained with air as the damping medium. When doubling the scanning speed from 0.075 mm/s to 0.15 mm/s the observed maximum frequency fd also approximately doubles. Only when further increasing the scanning speed above 0.15 mm/s the observed maximum frequency fd does not further increase significantly. Further when comparing the measurement results N=1 (air) and N=4 (water) it is concluded that observed maximum frequency fd for measurement N=1 is substantially lower as compared to the measurement N=4 than what would be expected on the basis of the difference in scanning speed. Based on this difference a maximum
would be expected
As can be seen, it was possible in these circumstances to increase the scanning speed v, without losing spatial information. The response of the apparatus is now only limited by the bandwidth of the amplitude detector, at frequencies above 10 kHz. Faster measurements of the amplitude may be performed by using a different class of amplitude detector, e.g. a peak detector, making it possible to approach the cantilever resonance frequency, here 30 kHz. It is noted that for this cantilever arrangement the damping provided by air is insufficient to achieve a quality factor Q that does not exceed 5, unless the dimensions of the cantilever very closely match the dimensions of the casing. This would however have the disadvantage that minor deviations in these dimensions already result in substantial deviations in the properties of the cantilever, in particular its quality factor Q.
| Number | Date | Country | Kind |
|---|---|---|---|
| 14187831.4 | Oct 2014 | EP | regional |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/NL2015/050699 | 10/5/2015 | WO | 00 |
