Research Topics

We have developed Frequency Modulation Atomic Force Microscopy (FM-AFM) and Scanning Tunneling Microscopy (STM). FM-AFM is also called non-contact Atomic Force Microscopy (nc-AFM) because it detects the weak attractive force between tip and sample, and they are not in contact when it is used in an ultra-high vacuum. In our research papers, it is referred to as UHV nc-AFM and FM-AFM in ambient condition. In this website, it is referred to as FM-AFM.

Atomic Force Spectroscopy

• Phys. Rev. Lett., 93, 256101-1-256101-4 (2004).

Toward atom-resolution nanomechanics we often measure the shift of resonance frequency (∆f) of a force sensor used in an atomic force microscope (AFM) using a frequency-modulation (FM) technique; we call this method FM-AFM. The value of ∆f negatively increases with decreasing AFM tip–sample separation when the attractive forces act between the tip and the sample. When the bias voltage (V) between them is changed while the tip–sample separation is kept constant, –∆f behaves proportionally to V² in the region of dominant electrostatic interactions. In general, at the wide separations where only van der Waals force and the electrostatic force act as long-range forces, –∆f changes along with the separation and the bias voltage; the surface plot in the figure below shows such changes of –∆f. Now then, how does the surface plot emerge at the closer separations in the figure where chemical bonding forces act as the short-range forces?

At the dawn of establishing quantum mechanics, Linus Pauling described in his great book of “The Nature of the Chemical Bond (the first edition published in 1939)” that the chemical bond is explained as the quantum mechanical resonance between the outermost electronic states of two atoms in proximity, the energy levels of which are so close. This means that if the energies of the two isolated atoms are not so matched, the chemical bond between the two atoms is not produced.

We discovered that strong attractive forces acted between the specific atoms on a sample surface and the AFM tip in close proximity, when a certain specific voltage is applied between the tip and the sample. This phenomenon seemed to correspond to the quantum mechanical resonance between the surface atom and the tip atom under the conditions that the energy levels of the atoms are in agreement; this could be the chemical bond. Then, using our custom-made UHV FM-AFM/STM_3, in which we were able to measure the tunneling current passing between them, we conducted the force spectroscopy while the tip was being scanned over the surface and instantly the force–bias voltage curves were measured at a constant tip–sample separation. We measured the ∆f–bias voltage curves while the tip–sample separation was decreased stepwise over the surface Si atom. As shown in the figure below, we found the sharp peaks at the specific bias voltages, and the peaks increased with decreasing separation. This indicated that the chemical forces could be produced by tuning the bias voltage between the AFM tip apex atom and the sample surface atom, corresponding to the quantum mechanical resonance, which is the concept introduced by Pauling.

Our experimental finding implied that we could extend the concept of the chemical bonding between two isolated atoms to the surface chemical bonding of two pieces of condensed matter by tuning the bias voltage between them. Our experimental sample was a Si surface and a Si tip with covalently bonding. We are studying the resonating forces between heterogenous atoms in condensed matter by use of our developed force spectroscopic method, and further extending this method of bias force spectroscopy to fabricate nanoscale structures.

Evaluation of surface resistances with Joule heat dissipation energy

• Jpn. J. Appl. Phys., 57, 08NB04 (2018).

In the FM-AFM to measure the forces between the tip and the sample, we measure the shift of resonance frequency ∆f of the force sensor of a tiny cantilever having an atomically sharpened tip while the cantilever is being oscillated at a constant amplitude; the ∆f roughly corresponds to the distance derivative of the force acting between the tip and the sample. Meanwhile, when the mechanical energy stored in the oscillating cantilever dissipates through nonconservative interactions, the oscillating amplitude becomes smaller. Then, the sinusoidal signal to excite the cantilever oscillation at a constant amplitude becomes larger to compensate the decrease in the amplitude; we use the feedback control to do so. Therefore, by measuring the change in the excitation sinusoidal signal, we can evaluate the dissipative energy through the nonconservative interactions. It is noted that we can simultaneously measure the tunneling current passing between the tip and the sample by applying a bias voltage between them.

When the tip and the sample are conductive, and a bias voltage V_S is applied, ∆f can be decomposed into ∆f_vdW, ∆f_ele, and ∆f_SR that are caused by the van der Waals force, the electrostatic force, and the short-range forces, respectively, as follows:

（1）　

Meanwhile, the dissipation energy D can be decomposed into D_int, D_Joule, and D_SR that are caused by the internal friction of the cantilever, the displacement current passing the resistance in the electric circuit involving the tip and the sample, resulting in Joule heat, and the nonconservative short-range interactions between the tip and the sample in close proximity, respectively, as follows:

（2）

Now we focus on ∆f_ele and D_Joule at wide separations where the short-range forces can be ignored. Here, we define the contact potential difference V_CPD of the sample with respect to the potential of the tip. When the resistance in the circuit can be represented by resistance R_J, the relationship between ∆f_ele and D_Joule can be expressed as follows:

（3）

（4）

where f₀, A, and k are the resonance frequency of the cantilever at free, its oscillation amplitude, and spring constant, respectively; V_S is the sample bias voltage, r_tip is the tip radius, and d is the closest tip–sample separation during the cantilever oscillating. Comparing eq. (3) with eq. (4), it is appreciable that both ∆f_ele and D_Joule are proportional to d^–1/2 and (V_S–V_CPD)². By dividing eq. (4) by eq. (3), the equation below is obtained:

（5）

From eq. 5, it is apparent that ∆f_ele is proportional to D_Joule independently of the distance and bias voltage. The proportional coefficient contains R_J, and the other parameters are estimated from other measurements. Thus, by simultaneous measurements of ∆f_ele and D_Joule, we can evaluate R_J from the proportional coefficient between them.

For a sample of Si(111)-(7×7) and a Si tip, we measured ∆f_ele and D_Joule at a separation of a few nm under sweeping bias voltage, as shown in the figures below.

From the plots in the figures, we made the plots of D_Joule as a function of ∆f_ele in the following figure. The plots show the linear relationship we anticipated, clearly indicating the proportional relationship between ∆f_ele and D_Joule. From the proportional coefficient, it is possible to evaluate the resistance R_J in the region where the displacement current passed.

Here, we discuss on which resistance in the circuit was measured. We compared the values obtained when we externally inserted a resistor in-series in the circuit and also deposited a thin film on the sample surface to change the surface resistance. As a consequence, we reached a conclusion that we measured the resistance around the tip apex and the sample region underneath its surface with a depth of the tip radius. We expect that this measurement of dissipation energy can infer the surface resistance of sample in the region of a few tens nm depth in noncontact.

Energy dissipation unveils atomic displacement

• Phys. Rev. B, 97, 115428 (2018).

Since the advent of STM in 1981, the reconstructed surface of Si(111)-(7×7) was successfully observed with atomic resolution by many researchers using STM and FM-AFM. In the STM and FM-AFM images, the individual Si adatoms on the surface were observed. The atomic arrangement of Si(111)-(7×7) was incontrovertibly explained by dimer-adatom-stacking fault (DAS) model.

In the past, the increase in dissipation energy over the adatom was reported, corresponding to pulling up of the adatom through the nonconservative interaction with the AFM tip. Our finding dissipation energy over the hollow sites was about 0.2 eV/cycle, that is, 0.2 eV per one cycle of the cantilever oscillation. This value was in agreement to the difference in Si coordination energy between the adatom site and the hollow site, which is the most stable and the meta stable site, respectively. We further conducted the experiment using a partially hydrogen (H)-terminated surface of Si(111)-(7×7). Therein, the dangling bonds of the adatoms and the rest atoms were partially terminated with H atom, leading to loss of reactivity of the dangling bond. As a result, we confirmed that, when both adatom and its neighboring rest atom were not terminated with H, the dissipation energy increased at the middle site between the adatom and the rest atom, that is the hollow site. We discussed the mechanism as follows: when the Si tip atom was brought closer to the hollow site, the bonding between the Si tip atom and the Si adatom was formed, and partly broke the back bond of the Si adtom to the surface. Further tip approach on the cantilever oscillation moves the Si adatom toward the hollow site, as if the tip flips the tetrahedron of Si atoms on the surface, and stabilized the Si adatom over the hollow site. Next, on the tip retraction, the bond between the Si tip atom and the adatom was broken. Afterward, the adatom at the hollow site returned to the original adatom site with the most stable coordination. At the moment the extra bonding energy is released, corresponding to the difference in the coordination energy between the hollow site and the adatom site. This atomic motion seemed to be repeated by the tip approach and retraction on its oscillation. The force–distance curve would have hysteresis because the bonding configuration around the Si adatom was altered between the approach and the retraction on the cantilever oscillation. The attractive force on the tip retraction would be stronger than that on the tip approach, because the adatom at the hollow site is located just under the Si tip. The area of the hysteresis in the force–distance curve should be equal to the dissipation energy as a mechanical work, which we measured. At present, for confirmation, we are simulating the atomistic motion on the basis of the first-principle calculations to evaluate the dissipation energy during one cycle of the tip oscillation.

nanometer-thick water film

On hydrophilic solid surfaces, water is adsorbed, whose thickness depends on the humidity of the environment. For example, at a low humidity, one layer of water molecules is adsorbed on the surface of a cubic crystal of salt (NaCl) that looks smoothly trickled. This layer has no ability of dissolving the salt, differently from bulk water. Meanwhile, in a salt-saturated water solution, at the interface between the remained salt crystals and the solution, water molecules form a hydration layer with Na⁺ and Cl^– ions on the crystals. Besides ionic crystals, the hydration layer is formed on a hydrophilic solid surface. In a simplified picture to represent the hydration layer, the water molecules are depicted to form a stable structure without mobility. But, in fact, the lifetime of being in a stable hydrated structure is short; the water molecules in bulk water are very often exchanged with the water molecules hydrated on the surface.

What happens at the interface between the bulk water and the solid surface under high humidities, where the water adsorbed layer has about 10-molecule thickness, i.e. a few nanometers thickness? We, our nano physics lab, revealed that the water layer exhibits quasi-stabilized mechanical properties; the mobility of water molecules in the layer is lower and the mechanical properties of the water molecules are between liquid and solid, similar to ice structures. Thus, we are exploring the thin water layers on the solid surfaces; we name them nanometer-thick water films.

Nanometer-thick water film on mica

• Scientific Reports, 7, 4054 (2017).

ReTuned Fork (RTF) force sensor

For a force sensor in FM-AFM, commercial Si cantilevers have been often used, which are oscillated at their resonance frequency. For the FM-AFM operated in air, we have also used them. On the other hand, to detect the frequency shift ∆f at a very small amplitude with a high sensitivity, we developed a retuned tuning fork (RTF) sensor from a quartz tuning fork for a watch.

To improve the force sensitivity for the FM-AFM, a high Q-value of the force sensor is indispensable; high Q-value means low loss mechanical energy on the oscillation. In the tuning fork, its two prongs have the same resonance frequency, which have been tuned in the fabrication process. The high Q-value is achieved in anti-phase oscillation mode of the two prongs. For the FM-AFM, we have to attach a tip on the one prong, resulted in deterioration of the Q-value due to their unmatched resonance frequencies between the two prongs. Thus, we cut a tiny part at the end of the one prong, where the tip is attached. The mass of the part is almost equal to the mass of the tip and the adhesive. This can provide the same resonance frequency of the two prongs. In our hand-made process, we fabricated the RTF sensor, in which the resonance frequency was retuned with a high Q-value.

Fabrication procedure of RTF sensor

• H. Ooe, T. Sakuishi, M. Nogami, M. Tomitori, T. Arai, Appl. Phys. Lett., 105, 043107 (2014).

The prong size of the quartz tuning fork is 2×0.2×0.1 mm. We ground a tiny part of the end of prong. Afterward, we attach a sharpened tip of 0.1 mm in length to there. These processes are done by our hands under the stereo optical microscope.

Image Gallery

Developed Instruments

UHV LT FM-AFM/STM_1

UHV LT FM-AFM/STM with retuned fork (RTF) force sensor 1

This FM-AFM/STM equipped with a liquid helium tank cryostat. A retuned fork (RTF) force sensor is used.

At Room 103-2 in Natural Science and Technology Hall 5.

UHV FM-AFM/STM_2

UHV FM-AFM/STM with retuned fork (RTF) force sensor 2

A retuned fork (RTF) force sensor with a W tip is used. The W tip is sharped by flame etching.

At Room 103-1 in Natural Science and Technology Hall 5.

UHV FM-AFM/STM_3

UHV FM-AFM/STM with piezoresistive Si cantilever

This FM-AFM/STM was first build up by Prof. Arai. It has a fluorescent screen for FIM/FEM in the UHV chamber.  FIM/FEM images of tips can be obtained.

It is placed at the collaborator’s Lab in Japan Advanced Institute of Science and Technology (JAIST).

ambient FM-AFM_4

ambient FM-AFM 4

This FM-AFM system was developed with the support by the SENTAN Program of the Japan Science and Technology Agency.

At Room 410-411 in Natural Science and Technology Hall 5.

ambient FM-AFM_5

ambient FM-AFM 5

The FM-AFM head was installed in a thermostatic chamber. A retuned fork (RTF) force sensor is used.

At Room 106 in Natural Science and Technology Hall 5.