Bhushan B. Nanotribology and Nanomechanics: An Introduction

Подождите немного. Документ загружается.

120 Jason H. Hafner

3.2.7 Carbon Nanotube Tips

Carbon nanotubes are microscopic graphitic cylinders that are nanometers in diam-

eter, yet many microns in length. Single-walled carbon nanotubes (SWNT) consist

of singlesp

hybridizedcarbon sheets rolled into seamless tubes and have diameters

ranging from 0.7 to 3nm.

Carbon Nanotube Structure

Larger structures called multiwalled carbon nanotubes (MWNT) consist of nested,

concentrically arranged SWNT and have diameters ranging from 3 to 50 nm. Fig-

ure 3.8showsa model of nanotubestructure,as well as TEMimages of a SWNT and

a MWNT. The small diameter and high aspect ratio of carbon nanotubes suggests

their application as high resolution, high-aspect-ratio AFM probes.

Carbon Nanotube Mechanical Properties

Carbon nanotubes possess exceptional mechanical properties that impact their use

as probes. Their lateral stiﬀnesscan be approximatedfrom that of a solid elastic rod:

lat

3πYr

, (3.4)

Fig. 3.8. The structure of carbon nanotubes, including TEM images of a MWNT (left)and

aSWNT(right), from [16]

3 Probes in Scanning Microscopies 121

where the spring constantk

lat

represents the restoring force per unit lateral displace-

ment, r is the radius, l is the length, and Y is the Young’s modulus (also called the

elastic modulus) of the material. For the small diameters and extreme aspect ratios

of carbon nanotube tips, the thermal vibrations of the probe tip at room temperature

can become suﬃcient to degrade image resolution. These thermal vibrations can be

approximated by equating

T of thermal energy to the energy of an oscillating

nanotube:

T =

lat

, (3.5)

where k

is Boltzmann’s constant, T is the temperature, and a is the vibration am-

plitude. Substituting for k

lat

from (3.4) yields:

a =



3πYr

. (3.6)

The strong dependence on radius and length reveals that one must carefully control

the tip geometry at this size scale. Equation (3.6) implies that the stiﬀer the mate-

rial, i.e., the higher its Young’s modulus, the smaller the thermal vibrations and the

longer and thinner a tip can be. The Young’s moduli of carbon nanotubes have been

determined by measurements of the thermal vibration amplitude by TEM [18,19]

and by directly measuring the forces required to deﬂect a pinned carbon nanotube

in an AFM [20]. These experiments revealed that the Young’s modulus of carbon

nanotubes is 1–2 TPa, in agreement with theoretical predictions [21]. This makes

carbon nanotubes the stiﬀest known material and, therefore, the best for fabricating

thin, high-aspect-ratio tips. A more detailed and accurate derivation of the thermal

vibration amplitudes was derived for the Young’s modulus measurements [18,19].

Carbon nanotubes elastically buckle under large loads, rather than fracture or

plastically deform like most materials. Nanotubes were ﬁrst observed in the buck-

led state by transmission electron microscopy [17], as shown in Fig. 3.9. The ﬁrst

experimental evidence that nanotube buckling is elastic came from the application

of nanotubes as probe tips [22],described in detail below. A more direct experimen-

tal observation of elastic buckling was obtained by deﬂecting nanotubes pinned to

a low friction surface with an AFM tip [20]. Both reports found that the buckling

force could be approximated with the macroscopic Euler buckling formula for an

elastic column:

Euler

. (3.7)

The buckling force puts another constraint on the tip length: If the nanotube is too

long the bucklingforcewill be too low for stable imaging. Theelastic buckling prop-

erty of carbon nanotubes has signiﬁcant implications for their use as AFM probes.

If a large force is applied to the tip inadvertently, or if the tip encounters a large

step in sample height, the nanotube can buckle to the side, then snap back without

degraded imaging resolution when the force is removed, making these tips highly

robust. No other tip material displays this buckling characteristic.

122 Jason H. Hafner

1.2 nm

1 nm

Fig. 3.9. TEM images and a model of a buckled

nanotube, adapted from [17]

Manually Assembled Nanotube Probes

The ﬁrst carbon nanotube AFM probes [22] were fabricated by techniques devel-

oped for assembling single-nanotube ﬁeld emission tips [23]. This process, illus-

trated in Fig. 3.10, used puriﬁed MWNT material synthesized by the carbon arc

procedure. The raw material, which must contain at least a few percent of long

nanotubes (> 10µm) by weight, puriﬁed by oxidation to approximately 1% of its

original mass. A torn edge of the puriﬁed material was attached to a micromanipu-

lator by carbon tape and viewed under a high power optical microscope. Individual

nanotubes and nanotubebundles were visible as ﬁlamentsunder dark ﬁeld illumina-

tion. A commercially available AFM tip was attached to another micromanipulator

opposing the nanotube material. Glue was applied to the tip apex from high vacuum

carbon tape supporting the nanotube material. Nanotubes were then manually at-

tached to the tip apex by micromanipulation. As assembled, MWNT tips were often

too long for imaging due to thermal vibrations and low buckling forces described in

Sect. 3.2.7. Nanotubes tips were shortened by applying 10V pulses to the tip while

it was near a sputtered niobium surface. This process etched ∼ 100nm lengths of

nanotube per pulse.

The manually assembled MWNT tips demonstrated several important nanotube

tip properties [22]. First, the high aspect ratio of the MWNT tips allowed the accu-

rate imaging of trenches in silicon with steep sidewalls, similar to FIB and EBD

tips. Second, elastic buckling was observed indirectly through force curves (see

3 Probes in Scanning Microscopies 123

50 × 0.6

Fig. 3.10. A schematic drawing of the setup for manual assembly of carbon nanotube tips

(top) and optical microscopy images of the assembly process (the cantilever was drawn in for

clarity)

Fig. 3.11). Note that as the tip taps the sample, the amplitudedrops to zero and a DC

deﬂection is observed, because the nanotube is unbuckled and is essentially rigid.

As the tip moves closer, the force on the nanotube eventually exceeds the buckling

force. The nanotube buckles, allowing the vibration amplitude to partially recover,

Deflection

(3.5 nm/div)

Z-position (10 nm/div)

Amplitude

(2 nm/div)

Fig. 3.11. Nanotube tip buck-

ling. Top diagrams corre-

spond to labeled regions

of the force curves. As the

nanotube tip buckles, the

deﬂection remains constant

and the amplitude increases,

from [16]

124 Jason H. Hafner

and the deﬂection remains constant. Numeric tip trajectory simulations could only

reproduce these force curves if elastic buckling was included in the nanotube re-

sponse. Finally, the nanotube tips were highly robust. Even after “tip crashes” or

hundreds of controlled buckling cycles, the tip retained its resolution and high as-

pect ratio.

Manual assembly of carbon nanotube probe tips is straightforward, but has sev-

eral limitations. It is labor intensive and notamenable to mass production.Although

MWNT tips have been made commercially available by this method, they are about

ten times more expensivethan silicon probes. The manualassembly method has also

been carried out in an SEM, rather than an optical microscope [24]. This eliminates

the need for pulse-etching, since short nanotubes can be attached to the tip, and the

“glue” can be applied by EBD. But this is still not the key to mass production, since

nanotube tips are made individually. MWNT tips provided a modest improvement

in resolution on biological samples, but typical MWNT radii are similar to that of

silicon tips, so they cannot provide the ultimate resolution possible with a SWNT

tip. SWNT bundles can be attached to silicon probes by manual assembly. Pulse

etching at times produces very high resolution tips that likely result from expos-

ing a small number of nanotubes from the bundle, but this is not reproducible [25].

Even if a sample could be prepared that consisted of individual SWNT for man-

ual assembly, such nanotubes would not be easily visible by optical microscopy or

SEM.

CVD Nanotube Probe Synthesis

The problems of manual assembly of nanotube probes discussed above can largely

be solved by directly growing nanotubes onto AFM tips by metal-catalyzed chem-

ical vapor deposition (CVD). The key features of the nanotube CVD process are

illustrated in Fig. 3.12. Nanometer-scale metal catalyst particles are heated in a gas

700–1000°C

Fig. 3.12. CVD nanotube synthesis. Ethylene reacts with a nanometer-scale iron catalyst

particle at high temperature to form a carbon nanotube. The inset in the upper right is a TEM

image showing a catalyst particle at the end of a nanotube, from [16]

3 Probes in Scanning Microscopies 125

mixture containing hydrocarbon or CO. The gas molecules dissociate on the metal

surface, and carbon is adsorbed into the catalyst particle. When this carbon precipi-

tates, it nucleates a nanotube of similar diameter to the catalyst particle. Therefore,

CVD allows control over nanotube size and structure, including the production of

SWNTs [26] with radii as low as 3.5 Angstrom [27].

Several key issues must be addressed to grow nanotube AFM tips by CVD:

(1) the alignment of the nanotubes at the tip, (2) the number of nanotubes that grow

at the tip, and (3) the length of the nanotube tip. Li et al. [28] found that nano-

tubes grow perpendicular to a porous surface containing embedded catalyst. This

approach was exploited to fabricate nanotube tips by CVD [29] with the proper

alignment, as illustrated inFig. 3.13.A ﬂattened area of approximately1–5µm

was

created on Si tips by scanning in contact mode at high load (1 µN) on a hard, syn-

thetic diamond surface. The tip was then anodized in HF to create 100 nm-diameter

pores in this ﬂat surface [30]. It is important to only anodize the last 20–40µmof

the cantilever, which includes the tip, so that the rest of the cantilever is still reﬂec-

tive for use in the AFM. This was achieved by anodizing the tip in a small drop

of HF under the view of an optical microscope. Next, iron was electrochemically

deposited into the pores to form catalyst particles [31]. Tips prepared in this way

were heated in low concentrations of ethylene at 800

◦

C, which is known to favor

the growth of thin nanotubes [26]. When imaged by SEM, nanotubes were found

Fig. 3.13. Pore-growth CVD nanotube tip fabrication. The left panel, from top to bottom,

shows the steps described in the text. The upper right is an SEM image of such a tip with

a small nanotube protruding from the pores (scale bar is 1 µm). The lower right is a TEM of

a nanotube protruding from the pores (scale bar is 20 nm), from [16]

126 Jason H. Hafner

to grow perpendicular to the surface from the pores as desired (Fig. 3.13). TEM re-

vealed that the nanotubes were thin, individual, multiwalled nanotubes with typical

radii ranging from 3–5nm. If nanotubes did not grow in an acceptable orientation,

the carbon could be removed by oxidation, and then CVD repeated to grow new

nanotube tips.

These “pore-growth” CVD nanotube tips were typically several microns in

length – too long for imaging – and were pulse-etched to a usable length of

< 500nm. The tips exhibitedelastic buckling behaviorand were very robustin imag-

ing. In addition, the thin, individual nanotube tips enabled improved resolution [29]

on isolated proteins. The pore-growth method demonstrated the potential of CVD

to simplify the fabrication of nanotube tips, although there were still limitations. In

particular, the porous layer was diﬃcult to prepare and rather fragile.

An alternative approach for CVD fabrication of nanotube tips involves direct

growth of SWNTs on the surface of a pyramidal AFM tip [32,33]. In this “surface-

growth” approach, an alumina/iron/molybdenum-powdered catalyst known to pro-

duce SWNT [26] was dispersed in ethanol at 1mg/mL. Silicon tips were dipped in

this solution and allowed to dry, leaving a sparse layer of ∼100nm catalyst clusters

on the tip. When CVD conditions wereapplied, single-walled nanotubesgrewalong

the silicon tipsurface.At a pyramidedge,nanotubescan either bend to alignwith the

edge,or protrude fromthe surface.If the energyrequired to bendthe tube and follow

the edge is less than the attractive nanotube-surface energy, then the nanotube will

follow the pyramid edge to the apex. Therefore, nanotubes were eﬀectively steered

toward the tip apex by the pyramid edges. At the apex, the nanotube protruded from

the tip, since the energetic cost of bending around the sharp silicon tip was too high.

The high aspect ratio at the oxide-sharpened silicon tip apex was critical for good

nanotube alignment. A schematic of this approach is shown in Fig. 3.14. Evidence

for this model came from SEM investigationsthat show that a very high yield of tips

contains nanotubes only at the apex, with very few protruding elsewhere from the

pyramid. TEM analysis demonstrated that the tips typically consist of small SWNT

bundles that are formed by nanotubes coming together from diﬀerent edges of the

pyramid to join at the apex, supporting the surface growth model described above

(Fig. 3.14). The “surface growth” nanotube tips exhibit a high aspect ratio and high

resolution imaging, as well as elastic buckling.

The surface growth method has been expanded to include wafer-scale produc-

tion of nanotube tips with yields of over 90% [34], yet one obstacle remains to

the mass production of nanotube probe tips. Nanotubes protruding from the tip are

several microns long, and since they are so thin, they must be etched to less than

100nm. While the pulse-etching step is fairly reproducible, it must be carried out

on nanotube tips in a serial fashion, so surface growth does not yet represent a true

method of batch nanotube tip fabrication.

Hybrid Nanotube Tip Fabrication: Pick-up Tips

Another method of creating nanotube tips is something of a hybrid between assem-

bly and CVD. The motivation was to create AFM probes that have an individual

3 Probes in Scanning Microscopies 127

b) c)

Fig. 3.14a–c. Surface-growth

nanotube tip fabrication.

(a) Schematic represents the

surface growth process in

which nanotubes growing on

the pyramidal tip are guided

to the tip apex. (b),(c)Images

show (b) SEM (200-nm-scale

bar) and (c) TEM (20-nm-

scale bar) images of a surface

growth tip, from [16]

SWNT at the tip to achieve the ultimate imaging resolution. In order to synthe-

size isolated SWNT, they must be nucleated at sites separated farther than their

typical length. The alumina-supported catalyst contains a high density of catalyst

particles per 100 nm cluster, so nanotube bundles cannot be avoided. To fabricate

completely isolated nanotubes, isolated catalyst particles were formed by dipping

a silicon wafer in an isopropyl alcohol solution of Fe(NO

)

.Thiseﬀectively left

a submonolayer of iron on the wafer, so that when it was heated in a CVD fur-

nace, the iron became mobile and aggregated to form isolated iron particles. During

CVD conditions, these particles nucleated and grew SWNTs. By controlling the

reaction time, the SWNT lengths were kept shorter than their typical separation,

so that the nanotubes never had a chance to form bundles. AFM analysis of these

samplesrevealed1–3nm-diameterSWNT and un-nucleatedparticleson the surface

(Fig. 3.15). However, there were tall objects that were diﬃcult to image at a density

of about 1 per 50µm

. SEM analysis at an oblique angle demonstrated that these

were SWNTs that had grown perpendicular to the surface (Fig. 3.15).

In the “pick-up tip” method, these isolated SWNT substrates were imaged by

AFM with silicon tips in air [9]. When the tip encountered a vertical SWNT, the

oscillation amplitude was damped, so the AFM pulled the sample away from the tip.

This pulled the SWNT into contact with the tip along its length, so that it became

attached to the tip. This assembly process happened automatically when imaging in

tapping mode – no special tip manipulation was required. When imaging a wafer

with the density shown in Fig. 3.15, one nanotube was attached per 8µm × 8 µm

scan at 512×512 and 2Hz, so a nanotube tip could be made in about 5min. Since

the as-formed SWNT tip continued to image, there was usually no evidence in the

128 Jason H. Hafner

(μm)

0 2.5 7.5

–10.0

–7.5

–5.0

–2.5

5.0

10.0

Fig. 3.15. Atomic force mi-

croscopy image (top)of

a wafer with isolated nano-

tubes synthesized by CVD.

The SEM view provides ev-

idence that some of these

nanotubes are arranged verti-

cally

Z (50 nm/div)

Time (10 ms/div)

Fig. 3.16. Pick-up tip assembly of nanotube probes. The top illustrates the nanotube pick-

up process that occurs while imaging vertical nanotubes in an AFM, including a trace of

the Z-position during a pick-up event. The lower left TEM images show single nanotubes

(diameters 0.9nmand2.8 nm) on an AFM tip fabricated by this method, adapted from [9]

topographic image that a nanotube had been attached. However, the pick-up event

was identiﬁed when the Z-voltage suddenly stepped to larger tip-sample separation

due to the eﬀective increase in tip length.

Individual SWNT tips must be quite short, typically less than 50nm in length,

for reasons outlined above. Pulse-etching, which removes 50–100nm of nanotube

3 Probes in Scanning Microscopies 129

Deflection (3.7 nm/div)

Z-position (5 nm/div)

Fig. 3.17. The process by which nanotube tips can be shortened in AFM force curves. The

hysteresis in the deﬂection trace (bottom) reveals that ∼ 17 nm were removed

length at a time, lacked the necessary precision for shortening pick-up tips, so the

tips were shortened through force curves. A pick-up tip force curve is shown in

Fig. 3.17. When the tip ﬁrst interacted with the sample, the amplitude decreased

to zero, and further approach generated a small deﬂection that ultimately saturated.

However, this saturation was not due to buckling. Note that the amplitude did not

recover, as in the buckling curves. This leveling was due to the nanotube sliding

on the pyramidal tip, which was conﬁrmed by the hysteresis in the amplitude and

deﬂection curves. If the force curve was repeated, the tip showed no deﬂection or

amplitude drop until further down, because the tip was essentially shorter.

Pick-up SWNT tips achievethe highest resolutionof all nanotube tips, since they

are always individuals rather than bundles. In tapping mode, they produce images

of 5nm gold particles that have a full width of ∼ 7 nm, the expected geometrical

resolution for a 2 nm cylindrical probe [9]. Although the pick-up method is serial

in nature, it may still be the key to the mass production of nanotube tips. Note that

the original nanotube tip length can be measured electronically from the size of the

Z-piezo step. The shortening can be electronically controlled through the hysteresis

in the force curves. Therefore, the entire procedure (including tip exchange) can be

automated by computer.

3.3 Scanning Tunneling Microscopy

Scanning tunneling microscopy(STM) was the original scanning probe microscopy

and generally produces the highest resolution images, routinely achieving atomic

resolution on ﬂat, conductive surfaces. In STM, the probe tip consists of a sharp-

ened metal wire that is held 0.3 to 1 nm from the sample. A potential diﬀerence of

0.1V to 1V between the tip and sample leads to tunneling currents on the order of