вход по аккаунту



код для вставкиСкачать
Nanometer-scale Elasticity Measurements on
Organic Monolayers Using Scanning Force
Waruna Kiridena,1 Vijay Jain,1 P. K. Kuo2 and Gang-yu Liu1,*
1 Department of Chemistry, Wayne State University, Detroit, MI 48202, USA
2 Department of Physics, Wayne State University, Detroit, MI 48202, USA
Local elastic compliance of organic monolayers (octadecyltriethoxysilane/mica and alkanethiol/gold) has been
investigated with nanometer resolution by applying a force modulation technique to an atomic force microscope.
Systematic measurements were taken as a function of modulation frequency and amplitude, as well as the local
environment surrounding the surface. The topography and local elasticity of the monolayers are contrasted to the
bare substrate created by the tip of the atomic force microscope at high imaging force. Under ambient laboratory
conditions, the YoungÏs modulus of mica calculated from the elasticity images is lower than the organic monolayer.
Such an observation is not intuitive and can be explained by the thin Ðlm of water adsorbed on mica. Water
adsorption can change the microscope tip surface interaction. As a result, mica appears as a softer surface than the
organic layers. In addition, the elasticity is dramatically enhanced if the modulation frequency coincides with or is
close to the natural resonance frequency of the tips of the atomic force microscope. Measurements taken under
liquid provide more reproducible and accurate results because the resonance frequency is damped out and capillary
interactions are avoided. The measured YoungÏs modulus is also found to increase slightly with increasing modulation amplitude. ( 1997 by John Wiley & Sons, Ltd.
Surf. Interface. Anal. 25, 383È389 (1997)
No. of Figures : 7 No. of Tables : 1 No. of Refs : 31
KEY WORDS : atomic force microscopy (AFM) ; force modulation microscopy ; elasticity ; elastic compliance ; self-assembled
Atomic force microscopy (AFM) is widely used to
image the surface topography of various materials with
nanometer to molecular resolution.1,2 Most AFM scanners adapt an optical lever design, Ðrst reported by
Meyer and Amer.3 With a quadrant photodiode detector, such a conÐguration enables simultaneous measurement of surface topography and frictional force. Recent
e†orts have demonstrated a much broader application
of AFM, e.g. imaging local mechanical properties in
addition to friction.4h12 For instance, if the sample is
modulated rapidly, the surface viscoelastic properties
can be extracted under contact mode from the corresponding response of the cantilever through both the
amplitude and the phase.4,7,9 There have been a
number of reports4h12 that utilize this new operation
mode to acquire an elasticity image at the same time as
the topographic image is taken. The local elasticity provides a new contrast mechanism to the AFM images
and is particularly useful in imaging the soft
materials.4h12 For example, local elasticity images have
been used to map the distribution of di†erent hardness
* Correspondence to : Gang-yu Liu, Department of Chemistry,
Wayne State University, Detroit, MI 48202, USA.
¤ This work was Ðrst presented at Surface Analysis Ï96, a topical
conference of the AVS, 12È14 June 1996, Ann Arbor, Michigan.
Contract grant sponsor : National Science Foundation ; grant no.
CCC 0142È2421/97/060383È07 $17.50
( 1997 by John Wiley & Sons, Ltd.
sites in fully hydrated calciÐed tissues11 and to measure
the local viscoelasticity of organic thin Ðlms5h12 and
living cells.5,12
E†orts have also been made to extract local elasticity
quantitatively.9h11 However, it has been difficult in
reporting absolute or even relative values of YoungÏs
modulus because of experimental uncertainties and lack
of systematic studies. Here we address this issue by systematically investigating the elastic compliance of selfassembled
well-characterized self-assembled monolayers were used
for this study : octadecyltriethoxysilane monolayer on
mica (OTE/mica) and octadecylmercapton monolayer
on gold(111) [C S/Au(111)].13 It is known that organic
monolayers on 18
thin Ðlms are good model systems for
organic boundary lubricants.9 In addition, the local frictional force is found to depend sensitively upon the
local Ðlm composition and elasticity.9 In the present
work, the inÑuence of the following parameters has
been studied : modulation frequency and amplitude ; and
the environment surrounding the surface, such as
imaging under ambient laboratory conditions as against
in a butanol solution.
The OTE/mica monolayers were prepared essentially
following the procedure reported by Peanasky et al.14 A
prehydrolysis solution was prepared by dissolving 0.23
Received 18 November 1996
Accepted 27 December 1996
ml of OTE in 25 ml of THF containing 0.1 ml of 1.31 N
HCl. The solution was stirred for 48 h before a 20-fold
dilution. We have found that 1 h of immersion followed
by 2 h of baking tends to produce the most homogenous OTE monolayers on mica. The C S/Au(111)
monolayers were prepared according to a procedure
Ðrst reported by Nuzzo and Allara.15 About 1500 Ó
gold was evaporated onto a mica surface under high
vacuum while the mica was kept at 350 ¡C. The thin
Ðlm of gold was immersed in a 1 mM C SH solution
immediately after the evaporation. Thiols18
are known to
form commensurate self-assembled monolayers on
The atomic force microscope used in this investigation is a home-built deÑection-type microscope. The
conÐguration of carrying out simultaneous topography
and elasticity measurements is illustrated in Fig. 1. In
order to image the surface elasticity, the sample is
modulated along the surface normal or z-direction
using a sinusoidal signal at the desired frequencies ( f )
and amplitudes (*z). The modulation frequency is kept
sufficiently high (tens of kHz) above the response of the
feedback circuit (typically a few kHz) so that the topographic images will not be a†ected. The deformation of
the cantilever would also follow a sinusoidal function in
response to the z-modulation (see Fig. 1) due to its
contact with the sample. The magnitude and phase of
the cantilever deformation are processed by a lock-in
ampliÐer (Stanford Research System SR 830 DSP). The
data acquisition software can record the topography
and the two signals from the lock-in ampliÐer simultaneously. Therefore, the surface topography and the cantileverÏs responding amplitude and phase can be
acquired concurrently. Di†erent elasticity or hardness
manifest into di†erent contrast in both amplitude and
phase images.
A similar procedure (reported in Ref. 9) was used to
quantify the surface elasticity. First, the cantilever
modulation amplitude, p, is extracted from the amplitude image. Surfaces with known YoungÏs modulus can
be used to precalibrate the amplitude image. For this
study, we used the hardest surface, e.g. a diamond
crystal, to calibrate the modulation amplitude of the
cantilever deÑection. In this case, the amplitude and
phase of the cantilever modulation are equal to those of
the surface. For other surfaces, the cantilever would
modulate with di†erent amplitude and phase. The
changes in amplitude and phase of the cantilever with
respect to the sample modulation reÑect the viscoelastic
behavior of the sample surface.
The height modulation of the sample, *z, is known
and a static approximation is assumed to be valid.* The
surface compression (deformation), d, can then be calculated using
d \ *z [ p
The force variation *F should be of the same magnitude for both the cantilever and the sample surface, i.e.
*F \ k d \ k (*z [ p) \ k pz
where k and p represent, respectively, the force constant andc deformation variation of the cantilever, and k
and d are, respectively, the force constant and compres-s
sion variation of the sample surface.
Rearrangement of Eqn (2) results in the force constant of the sample
k \k
c d
Using the above k and Hertzian model, one can deduce
s E
the surface modulus
k 3 [email protected]
where R and F are the tip radius and average image
force, respectively.
* For very hard surfaces such as diamond, Eqn (1) is accurate. For
most other samples, Eqn (1) is widely used as an approximation.
Figure 1. Schematic diagram of how AFM and force modulation are used to image surface topography and elasticity simultaneously. The
sample is modulated with the desired frequency (f ) and amplitude (Dz ). The cantilever follows the same frequency (f ) as the sample, but
with a phase shift and a different amplitude (p ) due to the elastic response of the sample (f , d ).
( 1997 by John Wiley & Sons, Ltd.
During our experiments, both the modulation amplitude and the phase of the cantilever were recorded
simultaneously. In principle, one can extract the elastic
and viscous part of E. In this study, we focus our discussion on the elastic part of E by using only the amplitude images.
Elasticity of OTE/mica measured under ambient
laboratory conditions
The OTE monolayer surface is composed of nearcircular OTE domains separated by mica, which
appears as faint “scarsÏ in Fig. 2. On zooming into any
OTE domain, the topographic image of OTE shows no
long-range order or periodicity, which is consistent with
previous AFM studies.20h24 Increasing the imaging
force beyond 20 nN results in displacement of OTE
adsorbates. On zooming into the newly created “holesÏ,
mica periodicity is clearly visible, as shown in Fig. 2(c).
By fabricating a mica area inlaid in the OTE monolayer, one can compare and contrast the relative local
Topography, amplitude and phase response of
the cantilever to z-modulation are measured simultaneously, as shown in Fig. 3. The modulation frequency,
f, varied from 10 to 50 kHz. Figure 3 displays only the
images at f \ 23 and 40 kHz, respectively. In the topographic scan, the OTE layer (bright area) is 25 Ó above
the mica (750 Ó ] 750 Ó dark square hole) substrate. In
the amplitude and phase images, the value over the
OTE area is higher than the mica area at most of the
modulation frequencies except for f \ 23 ^ 1 kHz,
where the contrast is reversed, as shown in Fig. 3. The
brighter areas in amplitude images indicate a higher
To understand the frequency dependence of the elasticity measurements, the frequency spectra of the AFM
cantilever were taken. The natural resonance frequency
of the free cantilever is 32 kHz [Fig. 4(a)], with a Q
factor of 16. During the scan, the AFM tip was in
Figure 2. Topographic image of OTE/mica. Image (a) shows OTE domains separated by uncovered substrate areas, which appear as faint
scars. All the images were taken using a sharpened microcantilever from Park Scientific Instrument, with a force constant of 0.1 N mÉ1. The
image force is 2 nN and the total scan areas of images (a), (b) and (c) are 4000 à 4000 AŽ 2, 70 à 70 AŽ 2 and 100 à 100 AŽ 2, respectively. The
center square hole in (a) is a 750 à 750 AŽ 2 bare mica substrate fabricated by displacing OTE molecules using an AFM tip at an image force
higher than 10 N. Zooming into the hole, high-resolution image (b) reveal the periodicity of the mica. As shown in image (c), OTE has no
long-range order. The cursor plot (d) reveals the thickness of the OTE layer to be 25 À 1 AŽ .
( 1997 by John Wiley & Sons, Ltd.
Figure 3. Simultaneous AFM images (2500 à 2500 AŽ 2) of topography, elasticity amplitude and phase of a monolayer film of OTE/mica.
Images were taken under ambient laboratory conditions using the same cantilever as in Fig. 2. The center square (700 à 700 AŽ 2) hole is the
bare mica substrate fabricated using the same AFM tip. All the elasticity images show a higher Young’s modulus on the OTE domain than on
mica (one example is shown at f ¼ 40 kHz) except for one modulation frequency, f ¼ 23 kHz. The modulation amplitude is 55 AŽ in both
cases, with an average image force of 2 nN.
contact with the surface. The frequency (natural
resonance) was down-shifted due to the tip/surface
interactions. As shown in Figs 4(b) and 4(c), the highest
resonance frequency of the cantilever becomes 23 or 27
kHz when in contact with the mica or OTE surface,
respectively. If modulated at D23 kHz, i.e. the new
resonance frequency when the tip interacts with the
mica, the vibrational amplitude of the cantilever is
enhanced. This causes the “ÑipÏ in the contrast, as shown
in Fig. 3. At modulation frequencies other than 23 kHz,
Figure 4. The frequency spectra taken for a sharpened microcantilever (Park Scientific Instrument) with a force constant of 0.1 N mÉ1. (a)
When the cantilever is far away from the surface (free cantilever), the spectrum shows the natural resonance frequency of 32 kHz, which is
consistent with the manufacture’s value. (b) When in contact with the OTE layer, two resonance peaks are located at 10 and 27 kHz. (c)
When in contact with the mica surface (inside the hole), the resonance is shifted to 23 kHz and another resonance peak occurs at 4 kHz.
( 1997 by John Wiley & Sons, Ltd.
the measured elasticity over the mica area is smaller
than over the OTE areas. Such a result is very counterintuitive at Ðrst glance because the reported YoungÏs
modulus for mica is 137È205 GPa25 and that for OTE
is 0.1È9 GPa.7,9,26 These observations can be rationalized by the presence of a thin Ðlm of water on the
surface when imaged under ambient conditions.27
Because mica is hydrophilic and OTE is hydrophobic,
mica is more susceptible to moisture in the air.28h30 In
fact, optical studies30 have revealed the occurrence of a
water Ðlm of D0.3 nm on mica in ambient air. Therefore, the measured “water/micaÏ elasticity is lower than
the real mica due to the capillary interaction between
the tip and the thin Ðlm of water.27 We may also understand the observed contrast in Fig. 3 in terms of the fact
that the water Ðlm is softer than the mica. Another
possibility, although much less likely, is the presence of
weakly adsorbed contaminants, e.g. some OTE molecules. These OTE molecules may be loosely packed
and therefore exhibit a lower YoungÏs modulus. The
elasticity measurements under ambient conditions
suggest that z-elastic compliance depends sensitively on
the Ðrst few atomic layers of the surface. Therefore, the
presence of a water Ðlm or other contaminants could
signiÐcantly alter the apparent elasticity of the
materials. In addition, frequency spectra should be
taken locally when the tip is in contact with di†erent
surface areas. The results are very helpful in guiding the
choice of modulation frequencies. For instance, frequency must be o†-resonance in order to obtain accurate z-elasticity measurements. On the other hand, one
can selectively choose the resonance frequency to
enhance the image contrast of the materials of interest.
Elasticity measurements in 2-butanol
To verify that the above observation is due to water
adsorption on the surface under ambient conditions, we
have imaged the OTE/mica surface in 2-butanol solution. Imaging in liquid can e†ectively avoid the inÑuence of water adsorbate and the capillary force between
the tip and the surface, which is always present under
ambient laboratory conditions.27h30 In addition, the
displaced OTE molecules dissolve in the solvent instead
of possible formation of loosely packed adsorbates
inside the hole.
As a bonus, the resonance frequency of the cantilever
is “damped outÏ in liquid, as shown in Fig. 5. Figure 5
also represents a typical frequency spectrum of soft cantilevers in liquid, in which the resonance frequency is
down-shifted. Such a dramatic down-shift in resonance
frequency can be understood in terms of cantilever/
liquid interaction. Such interaction results in the extra
inertia and drag due to the fact the liquid must vacate
the space in front of the cantilever and Ðll in the space
behind it. In other words, the e†ective mass of the oscillator (cantilever) is increased in liquid. As a result, the
natural resonance frequency is decreased.
The topographic and elastic compliance images were
taken at various frequencies and amplitudes of the zmodulation. Despite the modulation conditions, mica
always exhibits a higher amplitude than the OTE
monolayers for the elasticity images in 2-butanol.
Figure 6 displays two examples, i.e. images taken at
f \ 20 and 40 kHz. As shown in Fig. 6, the image contrast in all three images is independent of the modulation frequency. To check the generality of this
observation, another monolayer was examined :
C S/Au(111). As shown in Fig. 7, the measured elas18 of the gold substrate is higher than the surroundticity
ing monolayer. These qualitative observations support
our conclusions that more accurate measurements of
local elasticity can be taken under liquid to avoid the
complication of capillary interactions. Another advantage of imaging under liquid is that a much gentler
image force (or average load) can be used during the
Figure 5. The frequency spectra taken in 2-butanol for the same cantilever as in Fig. 4. The resonance frequency is down-shifted so that no
obvious resonance is observed for all three situations : (a) free cantilever (non-contact) ; (b) cantilever in contact with OTE ; and (c)
cantilever in contact with mica.
( 1997 by John Wiley & Sons, Ltd.
Figure 6. Simultaneous AFM images (3500 à 3500 AŽ 2) of topography, elasticity and phase of OTE/mica taken in 2-butanol with an image
force of 0.5 nN. The 750 à 1200 AŽ 2 rectangular area is the bare mica surface. Mica shows a higher Young’s modulus under 2-butanol at all
modulation conditions. Two examples are included here : f ¼ 23 kHz, Dz ¼ 30 AŽ ; and f ¼ 40 kHz, Dz ¼ 30 AŽ respectively.
measurements, which reduces tip-induced perturbation
during scanning and, in most cases, results in higher
resolution images.
A more quantitative estimation of YoungÏs modulus
is summarized in Table 1.
The YoungÏs modulus of OTE calculated from the
amplitude images is within the range of the literature
reported values of 0.1È9 GPa.7,9,27 The YoungÏs
modulus of mica, on the other hand, is below the
reported values of 137È205 GPa.25 Several factors may
be responsible for these results. The calibration may be
slightly o† due to the possible contamination on the
diamond surface in 2-butanol solution, which could
introduce systematic error. One way to avoid such sys-
tematic error is to use an internal standard, e.g. mica or
gold, assuming that the YoungÏs modulus is known.
Using the reported bulk value and the substrate as
internal standard, we calculated the YoungÏs modulus of
organic layers. The result is within the uncertainty of
the values listed in Table 1. For unknown surfaces, one
must use an external calibration method as described
above because of the absence of the internal standard.
Another possibility is that it may not be valid to
assume that the microscopic and the surface hardness
should be equal to the macroscopic bulk hardness.31 In
fact researchers introduced “surface hardnessÏ as a new
material property.31 One may be able to sense the bulk
properties by increasing the amplitude of the modula-
Figure 7. Simultaneous AFM images (3000 à 2500 AŽ 2) of topography, elasticity and phase of C S/gold taken under 2-butanol. The
18 central 750 à 740 AŽ 2 square area is
sample modulation condition is f ¼ 40 kHz and Dz ¼ 55 AŽ , with an average image force of 0.5 nN. The
the fabricated Au(111) surface. As can be seen in the amplitude image, the Young’s modulus of gold is higher than that of the thiol
( 1997 by John Wiley & Sons, Ltd.
tion. As can be seen from Table 1, a larger z-modulation
amplitude (*z) in general results in higher values of the
measured surface YoungÏs modulus. We think that a
higher z-modulation amplitude senses deeper into the
bulk, and therefore the measured YoungÏs modulus
approaches the value of the bulk materials.
placement of adsorbates, we have demonstrated that
elasticity measurements can distinguish, on a nanometer
level, between materials of di†erent compositions. We
can selectively enhance the elasticity image contrast of
certain materials by modulating the sample at the resonance frequency of the tip when in contact with the area
of the selected material. Comparison between the
images taken in ambient air and those taken in 2butanol solution suggests that the elastic compliance
images are mainly sensitive to the Ðrst few atomic layers
at surfaces. For instance, a thin Ðlm of water on the
surface causes the measured YoungÏs modulus of mica
to be lower than that of OTE. Imaging in a liquid can
avoid the capillary force e†ect and provides more reproducible and accurate elasticity measurements. The
YoungÏs modulus extracted from the images using a
Hertzian model depends on the modulation amplitude.
In general, we observed that the measured surface
YoungÏs modulus approaches the bulk value with
increasing z-modulation amplitude.
Simultaneous measurement of surface topography and
the amplitude and phase of the cantilever response to
z-modulation has been performed on self-assembled
monolayer surfaces using AFM. Using the contrast
between the monolayer and substrate formed by the dis-
We thank Song Xu for his help in the preparation of C S/Au(111).
G.Y.L. gratefully acknowledges the Camille and Henry Dreyfus
Foundation for a New Faculty Award, and the Arnold and Mabel
Beckman Foundation for a Young Investigator Award. This work is
also supported by Wayne State University, the Institute of Manufacturing Research and National Science Foundation Grant CHE9510402.
Table 1. YoungÏs modulus calculated from elasticity images
using a Hertzian modela
*z (AŽ )
E substrate (GPa)
E layer (GPa)
C S/Au(111)
C S/Au(111)
a The uncertainty of the Young’s modulus is 11%, which represents the precision (not the accuracy) of the measurements.
1. For recent reviews, see : J. Frommer, J . Angew . Chem ., Int . Ed .
Engl . 31, 1298 (1992) ; H. Fuchs, J . Mol . Struct . 292, 29
2. P. Enzel, G. S. Henderson, G. A. Ozin and R. L. Bedard, Adv .
Mater . 7, 64 (1995).
3. G. Meyer and N. M. Amer, Appl . Phys . Lett . 53, 2400 (1988).
4. P. Maivald, H. J. Butt, S. A. C. Gould, C. B. Pratev, B. Drake,
J. A. Gurley, V. B. Elings and P. K. Hansma, Nanotechnology
2, 103 (1991).
5. M. Radmacher, R. W. Tillmann, M. Fritz and H. E. Gaub,
Science 257, 1900 (1992).
6. R. M. Overney, T. Bonner, E. Meyer, M. Reutschi, R. Luthi, L.
Howald, J. Frommer, H. J. Guntherodt, M. Fujihara and H.
Takano, J . Vac . Sci . Technol . B 12, 1973 (1994) ; R. M.
Overney, D. P. Leta, C. F. Pictroski, M. H. Rafailovich, Y. Liu,
J. Quinn, J. Sokolov, A. Eisenberg and G. Overney, Phy . Rev .
Lett . 76, 1272 (1996).
7. M. Radmacher, R. W. Tillmann and H. E. Gaub, Biophys J . 64,
735 (1993).
8. S. D. Akari, E. W. van der Vegte, P. C. M. Grim, G. F. Belder,
V. Koutsos, G. TenBrinke and G. Hadziioannou, Appl . Phys .
Lett . 65, 1915 (1994).
9. R. M. Overney, E. Meyer, J. Frommer, H. J. Guntherodt, M.
Fujihira, H. Takano and Y. Gotoh, Langmuir 10, 1281 (1994).
10. H. Y. Nie, M. Motomatsu, W. Mizutani and H. Tokumoto, Thin
Solid Films 273, 143 (1996).
11. J. H. Kinney, M. Balooch, S. J. Marshall, G. W. Marshall and
T. P. Weihs, J . Biomech . Eng . Trans . ASME 118, 133 (1996).
12. M. Fritz, M. Radmacher, N. Petersen and H. E. Gaub, J . Vac .
Sci . Technol . B 12, 1526 (1994).
13. A. Ulman An Introduction to Ultrathin Organic Films . Academic Press, CA (1991).
14. J. Peanasky, H. M. Schneider, S. Granick and C. R. Kessel,
Langmuir 11, 953 (1995).
( 1997 by John Wiley & Sons, Ltd.
15. R. G. Nuzzo and D. L. Allara, J . Am . Chem . Soc . 105, 4481
16. L. H. Dubois and R. G. Nuzzo, Annu . Rev . Phys . Chem . 43,
437 (1992).
17. L. Strong and G. M. Whitesides, Langmuir 4, 546 (1988).
18. N. Camillone, C. E. D. Chidsey, G. Y. Liu and G. Scoles, J .
Chem . Phys . 98, 3503 (1993).
19. G. E. Poirier and E. D. Pylant, Science 272, 1145 (1996).
20. X. D. Xiao, G. Y. Liu, D. H. Charych and M. Salmeron, Langmuir 11, 1600 (1995).
21. D. K. Schwartz, S. Steinberg, J. Israelchvili and J. A. Zasadzinski, Phys . Rev . Lett . 69, 3354 (1992).
22. M. Fujii, S. Sugisawa, K. Fukada, T. Kate and T. Seimiya,
Langmuir 11, 405 (1995).
23. T. Nakagama, K. Ogawa and T. Kurumizawa, Langmuir 10,
525 (1994).
24. K. Bierbaum, M. Grunze, A. A. Baski, L. F. Chi, W. Schrepp
and H. Fuchs, Langmuir 11, 2143 (1995).
25. C. L. Mantell (ed.), Engineering Materials Handbook , 1st Edn,
pp. 35–48. McGraw-Hill, New York (1958).
26. M. Salmeron, G. Neubauer, A. Folch, M. Tomitori, D. F.
Ogletree and P. Sautet, Langmuir 9, 3600 (1993).
27. L. Scandella, A. Schumacher, N. Kruse, R. Prins, E. Meyer, R.
Luthi, L. Howald and H. J. Guntherodt, Thin Solid Films 240,
101 (1994).
28. J. N. Israelachvili and H. Wennestrom, Nature 379, 219
29. M. Ruths, H. Yoshizawa, L. J. Fetlers and J. N. Israelachvili,
Macromolecules 29, 7193 (1996).
30. A. M. Homola, J. N. Israelachvili, P. M. McGuiggan and M. L.
Gee, Wear 136, 65 (1990).
31. R. M. Overney, TRIP 3, 359, (1995).
Без категории
Размер файла
4 347 Кб
Пожаловаться на содержимое документа