Bhushan B. Nanotribology and Nanomechanics: An Introduction

Подождите немного. Документ загружается.

182 Markus Morgenstern et al.

imaged with atomic resolution only at low temperatures due to its extremely low

corrugation, which was below the room-temperature noise level [7,8].

Another,evenmore striking,exampleis the spectroscopicresolutionin scanning

tunneling spectroscopy (STS). This depends linearly on the temperature [2] and is

consequently reduced even more at LT than the thermal noise in AFM. This pro-

vides the opportunity to study structures or physical eﬀects not accessible at room

temperature such as spin and Landau levels in semiconductors [9].

Finally, it might be worth mentioningthat the enhanced stiﬀness of most materi-

als at low temperatures (increased Young’s modulus) leads to a reduced coupling

to external noise. Even though this eﬀect is considered small [6], it should not be

ignored.

5.2.3 Stability

There are two major stability issues that considerably improve at low temperature.

First, low temperatures close to the temperature of liquid helium inhibit most of the

thermallyactivated diﬀusionprocesses.As a consequence,the samplesurfacesshow

a signiﬁcantly increased long-term stability, since defect motion or adatom diﬀusion

is massively suppressed.Most strikingly, evensingle xenonatoms depositedon suit-

able substrates can be successfully imaged [10,11], or even manipulated [12]. In the

same way, the low temperatures also stabilize the atomic conﬁguration at the tip end

by preventing sudden jumps of the most loosely bound, foremost tip atom(s). Sec-

ondly, the large cryostat that usually surrounds the microscope acts as an eﬀective

cryo-pump. Thus samples can be kept clean for several weeks, which is a multiple

of the corresponding time at room temperature (about 3–4h).

5.2.4 Piezo Relaxation and Hysteresis

The last important beneﬁt from low-temperature operation of SPMs is that artifacts

from the response of the piezoelectric scanners are substantially reduced. After ap-

plying a voltage ramp to one electrode of a piezoelectric scanner, its immediate

initial deﬂection, l

, is followed by a much slower relaxation, Δl, with a logarith-

mic time dependence. This eﬀect, known as piezo relaxation or creep, diminishes

substantially at low temperatures, typically by a factor of ten or more. As a conse-

quence, piezo nonlinearities and piezo hysteresis decrease accordingly. Additional

information is given by Hug et al. [13].

5.3 Instrumentation

The two main design criteria for all vacuum-based scanning probe microscope sys-

tems are: (1) to provide an eﬃcient decoupling of the microscope from the vac-

uum system and other sources of external vibrations, and (2) to avoid most internal

noise sources through the high mechanical rigidity of the microscope body itself. In

5 Low-Temperature Scanning Probe Microscopy 183

vacuum systems designed for low-temperature applications, a signiﬁcant degree of

complexity is added, since, on the one hand, close thermal contact of the SPM and

cryogen is necessary to ensure the (approximately) drift-free conditions described

above, while, on the other hand, good vibration isolation (both from the outside

world, as well as from the boiling or ﬂowing cryogen) has to be maintained.

Plenty of microscope designs have been presented in the last 10–15years, pre-

dominantly in the ﬁeld of STM. Due to the variety of the diﬀerent approaches, we

will, somewhat arbitrarily, give two examples at diﬀerent levels of complexity that

might serve as illustrative model designs.

5.3.1 A Simple Design for a Variable-Temperature STM

A simple design for a variable-temperature STM system is presented in Fig. 5.1

(similar systems are also oﬀered by Omicron (Germany) or Jeol (Japan)). It should

give an impression of what the minimum requirements are, if samples are to be in-

vestigated successfully at low temperatures. It features a single ultrahigh vacuum

(UHV) chamber that houses the microscope in its center. The general idea to keep

the set-up simple is that only the sample is cooled, by means of a ﬂow cryostat that

ends in the small liquid-nitrogen (LN) reservoir. This reservoir is connected to the

sample holder with copper braids. The role of the copper braids is to attach the LN

reservoir thermally to the sample located on the sample holder in an eﬀective man-

ner, while vibrations due to the ﬂow of the cryogen should be blocked as much as

possible. In this way, a sample temperature of about 100 K is reached. Alternatively,

with liquid-helium operation, a base temperature of below 30 K can be achieved,

while a heater that is integrated into the sample stage enables high-temperature op-

eration up to 1000K.

A typical experiment would run as follows. First, the sample is brought into the

system by placing it in the so-called load-lock. This small part of the chamber can

be separated from the rest of the system by a valve, so that the main part of the

system can remain under vacuum at all times (i.e., even if the load-lock is opened

to introduce the sample). After vacuum is reestablished, the sample is transferredto

the main chamber using the transfer arm. A linear-motion feedthrough enables the

storage of sample holders or, alternatively, specialized holders that carry replace-

ment tips for the STM. Extending the transfer arm further, the sample can be placed

on the sample stage and subsequently cooled down to the desired temperature. The

scan head, which carries the STM tip, is then lowered with the scan-head manip-

ulator onto the sample holder (see Fig. 5.2). The special design of the scan head

(see [14] for details) allows not only a ﬂexible positioning of the tip on any de-

sired location on the sample surface, but also compensates to a certain degree for

the thermal drift that inevitably occurs in such a design due to temperature gradi-

ents.

In fact, thermal drift is often much more prominent in LT-SPM designs, where

only the sample is cooled, than in room-temperature designs. Therefore, to beneﬁt

fully from the high stability conditionsdescribed in the introduction, it is mandatory

184 Markus Morgenstern et al.

Fig. 5.1. One-chamber UHV system with

variable-temperature STM based on a ﬂow cryostat

design (courtesy of RHK Technology, USA)

5 Low-Temperature Scanning Probe Microscopy 185

Heating/

cooling

stage

Sample

holder

Tip

Scan head

manipulator

Copper

braids

Scan

head

Fig. 5.2. Photograph of the

STM located inside the sys-

tem sketched in Fig. 5.1.

After the scan head has been

lowered onto the sample

holder, it is fully decoupled

from the scan head manipula-

tor and can be moved laterally

using the three piezo legs on

which it stands (courtesy of

RHK Technology, USA)

to keep the whole microscope at the exact same temperature. This is mostly realized

by using bath cryostats, which add a certain degree of complexity.

5.3.2 A Low Temperature SFM Based on a Bath Cryostat

As an example of an LT-SPM set-up based on a bath cryostat, let us take a closer

look at the LT-SFM system sketched in Fig. 5.3, which has been used to acquire

the images on graphite, xenon, NiO, and InAs presented in Sect. 5.5. The force mi-

croscope is built into a UHV system that comprises three vacuum chambers: one

for cantilever and sample preparation, which also serves as a transfer chamber, one

for analysis purposes, and a main chamber that houses the microscope. A specially

designed vertical transfer mechanism based on a double chain allows the lower-

ing of the microscope into a UHV-compatible bath cryostat attached underneath

the main chamber. To damp the system, it is mounted on a table carried by pneu-

matic damping legs, which, in turn, stand on a separate foundation to decouple it

from building vibrations. The cryostat and dewar are separated from the rest of the

UHV system by a bellow. In addition, the dewar is surrounded by sand for acoustic

isolation.

In this design, tip and sample are exchanged at room temperature in the main

chamber. After the transfer into the cryostat, the SFM can be cooled by either liquid

nitrogen or liquid helium, reaching temperatures down to 10K. An all-ﬁber inter-

ferometer as the detection mechanism for the cantilever deﬂection ensures high res-

olution, while simultaneously allowing the construction of a comparatively small,

rigid, and symmetric microscope.

Figure 5.4 highlights the layout of the SFM body itself. Along with the care-

ful choice of materials, the symmetric design eliminates most of the problems with

drift inside the microscope encountered when cooling or warming it up. The micro-

scope body has an overallcylindricalshape with a height of 13cm and a diameter of

186 Markus Morgenstern et al.

Electrical and fiber

feedthroughs

Getter

pumps

Analysis

chamber

Turbo

pump

Motor chain

drive

Main

chamber

Vertical

chain transfer

Preparation

chamber

Getter

pump

Turbo

pump

Table

Pneumatic

leg

Dewar

Cryostat

Copper

cone

SFM

Sand

Separate foundation

with pit

Fig. 5.3. Three-chamber

UHV and bath cryostat sys-

tem for scanning force mi-

croscopy, front view

6 cm and exact mirror symmetry along the cantilever axis. The main body is made

of a single block of macor, a machinable glass ceramic, which ensures a rigid and

stable design. For most of the metallic parts titanium was used, which has a tem-

perature coeﬃcient similar to macor. The controlled but stable accomplishment of

movements,such as coarse approach and lateral positioningin other microscope de-

signs, is a diﬃcult task at low temperatures. The present design uses a special type

of piezo motor that moves a sapphire prism (see the ﬁber approach and the sam-

ple approach labels in Fig. 5.3); it is described in detail in [15]. More information

regarding this design is given in [16].

5 Low-Temperature Scanning Probe Microscopy 187

Fiber

Macor

body

Fiber

approach

Piezo tube

Cantilever

stage

Sample

approach

Sapphire

prism

a) b)

Fig. 5.4. The scanning force microscope incorporated into the system presented in Fig. 5.3.

(a) Section along plane of symmetry. (b) Photo from the front

5.4 Scanning Tunneling Microscopy and Spectroscopy

In this section, we review some of the most important results achieved by LT-STM.

After summarizing the results, placing emphasis on the necessity for LT equipment,

we turn to the details of the diﬀerent experiments and the physical meaning of the

results obtained.

As described in Sect. 5.2, the LT equipment has basically three advantages for

scanning tunneling microscopy (STM) and spectroscopy (STS). First, the instru-

ments are much more stable with respect to thermal drift and coupling to external

noise, allowing the establishment of new functionalities of the instrument. In par-

ticular, the LT-STM has been used to move atoms on a surface [12], cut molecules

into pieces [17], reform bonds [18], and, consequently, establish new structures on

the nanometer scale. Also, the detection of light resulting from tunneling into a par-

ticular atom [19] and the visualization of thermally induced atomic movements[20]

partly require LT instrumentation.

Second, the spectroscopic resolution in STS depends linearly on temperature

and is, therefore,considerably reduced at LT. This providesthe opportunity to study

physicaleﬀects unaccessible at room temperature.Obvious examples are the resolu-

tion of spin and Landau levels in semiconductors [9], or the investigationof lifetime

broadening eﬀects of particular electronic states on the nanometer scale [21]. More

spectacularly, electronic wave functions have been imaged for the ﬁrst time in real

space usingan LT-STM [22], andvibrationallevels givingrise to additionalinelastic

tunneling have been detected [23] and localized within particular molecules [24].

188 Markus Morgenstern et al.

Third, and most obviously, many physical eﬀects, in particular, eﬀects guided

by electronic correlations, are restricted to low temperature. Typical examples are

superconductivity [3], the Kondo eﬀect [4], and many of the electron phases found

in semiconductors [25]. Here, LT-STM provides the possibility to study electronic

eﬀects on a local scale, and intensive work has been done in this ﬁeld, the most

elaborate with respect to high-temperaturesuperconductivity [26].

5.4.1 Atomic Manipulation

Although manipulation of surfaces on the atomic scale can be achieved at room

temperature [27], only the use of LT-STM allows the placement of individual atoms

at desired atomic positions [28].

The usual technique to manipulate atoms is to increase the current above a cer-

tain atom, which reduces the tip–atom distance, then to move the tip with the atom

to a desired position, and ﬁnally to reduce the current again in order to decouple the

atom and tip. The ﬁrst demonstrationof this technique was performed by Eigler and

Schweizer [12], who used Xe atoms on a Ni(110) surface to write the three letters

“IBM” (their employer) on the atomic scale (Fig. 5.5a). Nowadays, many laborato-

ries are able to move diﬀerent kinds of atoms and molecules on diﬀerent surfaces

with high precision. An example featuring CO molecules on Cu(110) is shown in

Fig. 5.5b–g. Basic modes of controlled motion, pushing, pulling, and sliding of the

molecules, have been established that depend on the tunneling current, i.e., the dis-

tance and the particular molecule–substrate combination [29]. It is believed that

the electric ﬁeld between the tip and molecule is the strongest force moving the

molecules, but other mechanisms such as electromigration caused by the high cur-

rent density [28] or modiﬁcations of the surface potential due to the presence of the

tip [30] have been put forth as important for some of the manipulation modes.

Meanwhile, other types of manipulation on the atomic scale have been devel-

oped. Some of them require inelastic tunneling into vibrational or rotational modes

of the molecules or atoms. They lead to controlled desorption [31], diﬀusion [32],

pick-up of molecules by the tip [18], or rotation of individual entities [33,34]. Also,

dissociation of molecules by voltage pulses [17], conformational changes induced

by dramatic change of the tip–molecule distance [35], and association of pieces

into larger molecules by reducing their lateral distance [18] have been shown. Fig-

ure 5.5h–mshowsthe productionof biphenylfromtwo iodobenzenemolecules[36].

The iodine is abstracted by voltage pulses (Fig. 5.5i,j), then the iodine is moved to

the terrace by the pulling mode (Fig. 5.5k,l), and ﬁnally the two phenyl parts are slid

along the step edge until they are close enough to react (Fig. 5.5m). The chemical

identiﬁcation of the componentsis not deducedstraightforwardlyfrom STM images

and partly requires detailed calculations of their apparent shape.

Low temperatures are not always required in these experiments, but they in-

crease reproducibility because of the higher stability of the instrument, as discussed

in Sect. 5.2. Moreover, rotation or diﬀusion of entities could be excited at higher

temperatures, making the intentionally produced conﬁgurationsunstable.

5 Low-Temperature Scanning Probe Microscopy 189

b) c) d)

e) f) g)

Fig. 5.5. (a) STM image of single Xe atoms positioned on a Ni(110) surface in order to

realize the letters IBM on the atomic scale (courtesy of Eigler, IBM); (b)–(f) STM images

recorded after diﬀerent positioning processes of CO molecules on a Cu(110) surface; (g)ﬁnal

artwork greeting the new millennium on the atomic scale((b)–(g) courtesy of Meyer, Berlin).

(h)–(m) Synthesis of biphenyl from two iodobenzene molecules on Cu(111): First, iodine

is abstracted from both molecules (i),(j), then the iodine between the two phenyl groups

is removed from the step (k), and ﬁnally one of the phenyls is slid along the Cu-step (l)

until it reacts with the other phenyl (m); the line drawings symbolize the actual status of the

molecules ((h)–(m) courtesy of Hla and Rieder, Ohio)

5.4.2 Imaging Atomic Motion

Since individualmanipulation processes last seconds to minutes, they probablycan-

not be used to manufacture large and repetitive structures. A possibility to construct

such structures is self-assembled growth. This partly relies on the temperature de-

pendence of diﬀerent diﬀusion processes on the surface. A detailed knowledge of

the diﬀusion parameters is required, which can be deduced from sequences of STM

images measured at temperatures close to the onset of the process of interest [37].

Since many diﬀusion processes have their onset at LT, LT are partly required [20].

Consecutive images of so-called hexa-tert-butyl-decacyclene (HtBDC) molecules

on Cu(110) recordedat T = 194K are shownin Fig. 5.6a–c[38]. As indicated by the

190 Markus Morgenstern et al.

Temperature (K)

1,000/t

–2

–4

–6

40 60 80 100 120

–16

–18

–20

–22

80 40 25 15 10

(cm

–1

)

Hop

rate

–1

)

–2

–4

–6

–8

–34

–36

–38

–40

–42

211 193 178 166

55 60 65 70

In h

(lns

–1

)

T (K)

In D(ln

(cm

–1

))

(kt)

–1

a) b) c)

(001)

–

01)

Fig. 5.6. (a)–(c) Consecutive

STM images of hexa-tert-

butyl decacyclene molecules

on Cu(110) imaged at

T = 194 K; arrows indi-

cate the direction of motion

of the molecules between two

images. (d) Arrhenius plot of

the hopping rate h determined

from images like (a)–(c)as

a function of inverse tem-

perature (grey symbols); the

brown symbols show the

corresponding diﬀusion con-

stant D; lines are ﬁt results

revealing an energy barrier

of 570 meV for molecular

diﬀusion ((a)–(d) courtesy of

M. Schuhnack and F. Besen-

bacher, Aarhus). (e) Arrhe-

nius plot for D (crosses)and

H(circles) on Cu(001). The

constant hopping rate of H

below 65 K indicates a non-

thermal diﬀusion process,

probably tunneling (courtesy

of Ho, Irvine)

arrows, the position of the molecules changeswith time, implying diﬀusion. Diﬀu-

sion parameters are obtainedfrom Arrhenius plots of the determined hoppingrate h,

as shown in Fig. 5.6d. Of course, one must make sure that the diﬀusion process is

not inﬂuenced by the presence of the tip, since it is known from manipulation ex-

periments that the presence of the tip can move a molecule. However, particularly

at low tunneling voltages, these conditions can be fulﬁlled.

Besides the determination of diﬀusion parameters, studies of the diﬀusion of

individual molecules showed the importance of mutual interactions in diﬀusion,

which can lead to concerted motion of several molecules [20], or, very interestingly,

the inﬂuence of quantum tunneling [39]. The latter is deduced from the Arrhenius

plot of hopping rates of H and D on Cu(001), as shown in Fig. 5.6e. The hopping

rate of H levels oﬀ at about 65K, while the hopping rate of the heavier D atom goes

down to nearly zero, as expected from thermally induced hopping.

5 Low-Temperature Scanning Probe Microscopy 191

Other diﬀusion processes such as the movement of surface vacancies [40] or

of bulk interstitials close to the surface [41], and the Brownian motion of vacancy

islands [42] have also been displayed.

5.4.3 Detecting Light from Single Atoms and Molecules

It had already been realized in 1988 that STM experimentsare accompaniedby light

emission [43]. The fact that molecular resolution in the light intensity was achieved

at LT (Fig. 5.7a,b) [19] raised the hope of performing quasi-optical experiments on

the molecular scale. Meanwhile, it is clear that the basic emission process observed

on metals is the decay of a local plasmon induced in the area around the tip by

inelastic tunneling processes [44, 45]. Thus, the molecular resolution is basically

a change in the plasmon environment, largely given by the increased height of the

tip with respect to the surface above the molecule [46]. However, the electron can,

in principle, also decay via single-particle excitations. Indeed, signatures of single-

particle levels are observed. Figure 5.7c shows light spectra measured at diﬀerent

tunneling voltages V above a nearly complete Na monolayer on Cu(111) [47]. The

plasmon-mode peak energy (arrow) is found, as usual, to be proportional to V,but

an additional peak that does not move with V appears at 1.6eV (p). Plotting the

light intensity as a function of photon energy and V (Fig. 5.7d) clearly shows that

this additional peak is ﬁxed in photon energy and corresponds to the separation of

quantum-well levels of the Na (E

−E

Light has also been detected from semiconductors [48], including heterostruc-

tures [49]. There, the light is mostly caused by single-particle relaxation of the in-

jected electrons, allowing a very local source of electron injection.

5.4.4 High-Resolution Spectroscopy

One of the most important modes of LT-STM is STS. It detects the diﬀerential con-

ductivity dI/ dV as a function of the applied voltage V and the position (x, y). The

dI/ dV signal is basically proportional to the local density of states (LDOS) of the

sample, the sum over squared single-particle wave functions Ψ

[2]

(V, x,y) ∝ LDOS(E, x, y) =



ΔE

|Ψ

(E, x, y)|

, (5.1)

where ΔE is the energy resolution of the experiment. In simple terms, each state

corresponds to a tunneling channel if it is located between the Fermi levels (E

)

of the tip and the sample. Thus, all states located in this energy interval contribute

to I, while dI/ dV(V) detects only the states at the energy E corresponding to V.

The local intensity of each channel depends further on the LDOS of the state at

the corresponding surface position and its decay length into vacuum. For s-like tip

states, Tersoﬀ and Hamann have shown that it is simply proportional to the LDOS

at the position of the tip [50]. Therefore, as long as the decay length is spatially

constant, one measures the LDOS at the surface (5.1). Note that the contributing