Carotene as a molecular wire: Conducting atomic force microscopy

Article abstract of DOI:10.1021/jp9831278

A conducting atomic force microscope was used to measure the electrical properties of carotenoid molecules attached to a gold electrode. The thiolated carotene molecules were embedded in insulating n-alkanethiol self-assembled monolayers. At a contact force of a few nanoNewtons, a carotenoid molecule behaves ohmically with a resistance of approximately 4.2 ± 0.7 GΩ, over a million times more conductive than an alkane chain of similar length. Modes of electron transport are discussed.

Full text of DOI:10.1021/jp9831278

4006
J. Phys. Chem. B 1999, 103, 4006-4010
Carotene as a Molecular Wire: Conducting Atomic Force Microscopy
†
‡
‡
‡
‡
‡
‡
G. Leatherman, E. N. Durantini, D. Gust, T. A. Moore, A. L. Moore, S. Stone, Z. Zhou,
†
†
,†
P. Rez, Y. Z. Liu, and S. M. Lindsay*
Department of Physics and Astronomy and Department of Chemistry and Biochemistry,
Arizona State UniVersity, Tempe, Arizona 85287
ReceiVed: July 22, 1998; In Final Form: September 16, 1998
A conducting atomic force microscope was used to measure the electrical properties of carotenoid molecules
attached to a gold electrode. The thiolated carotene molecules were embedded in insulating n-alkanethiol
self-assembled monolayers. At a contact force of a few nanoNewtons, a carotenoid molecule behaves ohmically
with a resistance of approximately 4.2 ( 0.7 GΩ, over a million times more conductive than an alkane chain
of similar length. Modes of electron transport are discussed.
Introduction
Current proposals for molecular wires include organic
1
molecules with large, delocalized π-electron systems. The
prototypical natural molecular conductor of this type is the
carotenoid polyene. Much of the color in plants, animals, and
microorganisms is due to the absorption of light by carotenoids,
whose optical properties arise from electronic states that are
delocalized over a chain of some twenty carbon atoms connected
2,3
by alternating single and double bonds. Carotenoids are readily
oxidized electrochemically,4 implying that they can be charged
at a metal surface, and in their ionic states, charge is delocalized
over the length of the conjugated chain. Photoinduced electron
transfer through carotenoids is facile.6 These facts suggest that
single carotenoids fixed between metallic contacts could function
as wires. In this work, conducting atomic force microscopy has
been used to contact individual carotenoids embedded in an
insulating alkanethiol monolayer on a gold surface with a
controlled force so as to measure the resistance of the molecule.
The scanning tunneling microscope has been used to measure
,5
,7
8,9
the electrical properties of small molecules and nanostructures
Figure 1. Schematic illustration of the experiment. Structural models
are for the carotenoid embedded in 1-docosanethiol attached to a gold
surface. Current flow is measured between a biased Pt-coated AFM
cantilever and a gold substrate. The tip is grounded and a bias applied
to the substrate. The sample is submerged in an insulating solvent in
a clean Teflon container.
such as carbon nanotubes.10 Oxidation currents have been
11
detected from single molecules in solution. Atomic “wires”
have been fabricated by stacking Xe atoms on top of one another
12
with an STM tip. The resistance of such a two atom “wire”
7
was 10 Ω. A combined STM-microwave dielectric microscope
was used to detect conduction in 4,4′-di(phenylene-ethynylene)-
Experimental Section
1
3
benzenethiolate. Molecules have been strung across micro-
fabricated junctions14 and coated onto microcracks (“break-
junctions”) in electrodes.15 These methods present some
difficulties. In STM-based experiments, the contact force is not
Synthesis. The carotenoid, (7′-apo-7′-(4-iodomethylphenyl)-
â-carotene), was synthesized as described below and mixed with
either 1-dodecanethiol (12 carbon atoms, Aldrich Chemical Co.)
or 22-carbon diluent, 1-docosanethiol (see below), in toluene
solutions. The alkanethiols formed ordered monolayers in which
carotenoids were embedded as illustrated in Figure 1, which
shows the overall layout. The structure of the carotenoid is
illustrated below:
controlled. The electronic properties of a molecule are sensitive
to the effects of deformation1
6-18
so it is important to sense
force and current simultaneously. This has been accomplished
in this work by the use of a conducting atomic force microscope
(AFM) to image and contact individual molecules embedded
in an otherwise insulating film. Contamination by water adlayers
19
and molecular oxygen can lead to spurious conductivity signals
so our measurements were carried out under an inert atmosphere
with the sample and probe covered by an insulating liquid.
*
Corresponding author. E-mail: Stuart.Lindsay@ASU.EDU.
Department of Physics and Astronomy.
Department of Chemistry and Biochemistry.
†
The thiols were synthesized from the corresponding halide,
‡
20
7′-apo-7′-(4-iodomethylphenyl)-â-carotene
or 1-bromo-
1
0.1021/jp9831278 CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/21/1998

Carotene as a Molecular Wire
J. Phys. Chem. B, Vol. 103, No. 20, 1999 4007
docosane, by nucleophilic displacement of the halide with an
excess of thiourea in THF. The thiols were obtained by base
hydrolysis of the thiouronium salt.
All materials were characterized by H and 13C NMR (COSY,
HMBC, and HMQC). The spectra were recorded on a Varian
Unity spectrometer at 500 MHz. The samples were dissolved
in deuteriochloroform with tetramethylsilane as an internal
standard. Mass spectra were obtained with a matrix-assisted
time-of-flight spectrometer (Vestec Laser Tec Research Instru-
ment using a sulfur matrix). Measured mass/charge ratios (m/
z) are listed for each compound. 1-Bromodocosane from Aldrich
were used as received. Tetrahydrofuran (THF) was distilled from
lithium aluminum hydride under an argon atmosphere and used
at once.
1
Figure 2. (A) STM image of film made from a toluene solution
containing 1 mM 1-dodecanethiol + 0.003 mM carotenethiol: (B) as
in A with 0.015 mM carotenethiol. Images taken in constant current
mode at 5 pA and +800 mV. Height scale for both images (shown in
A) is 0 to 3 nm.
7′-Apo-7′-(4-mercaptomethylphenyl)-â-carotene (1). A por-
tion of 7′-apo-7′-(4-iodomethylphenyl)-â-carotene (40 mg, 0.063
mmol) was added to an excess of thiourea (200 mg) in THF
(20 mL). The reaction mixture was stirred under nitrogen at
room temperature for 20 min. TLC (silica gel, hexane/3% ethyl
acetate) indicated that all of the iodocarotene was consumed
and the polar thiouronium salt was formed. The reaction mixture
was filtered. Powdered KOH (250 mg) was added to the filtrate
and the mixture was vigorously stirred under nitrogen at room
temperature for 45 min. The crude mixture was diluted with 50
mL of water and extracted with CH2Cl2 (3 × mL). The organic
phase was dried with MgSO4, filtered and the solvent evaporated
under reduced pressure. Flash chromatography on silica gel
(
hexane/20% CH2Cl2) afforded 28 mg (83%) of pure carotene-
+
1
thiol (1). MS m/z: 536 M (536 calculated for C38H48S). H
NMR (500 MHz, CDCl3) δ [ppm] 1.02 (s, 6H, 16-CH3, 17-
CH3), 1.47 (m, 2H, 2-CH2-), 1.61 (m, 2H, 3-CH2-), 1.71 (s,
3
2
Figure 3. Coverage of carotenethiol (left scale, open squares) and
particle size (right scale, filled circles) as a function of carotenoid
solution concentration.
H, 18-CH3), 1.97 (s, 9H, 19-CH3, 20-CH3, 20′-CH3), 2.01 (m,
H, 4-CH2-), 2.03 (s, 3H, 19′-CH3), 3.62 (s, 2H, Ar-CH2-
SH), 6.14 (d, 1H, J ) 15.5 Hz, 8-CH-), 6.15 (d, J ) 8.5 Hz,
H, 10-CH-), 6.16 (d, 1H, J ) 15.5 Hz, 7-CH-), 6.25 (d, 1H,
J ) 10.5 Hz, 14′-CH-), 6.27 (d, 1H, J ) 8.0 Hz, 14-CH-),
Sample Preparation and Imaging. Solutions of the caro-
1
tenoid and either 1-dodecanethiol or 1-docosanethiol in freshly
21
distilled toluene were placed on freshly flame-annealed films
6
.35 (d, 1H, J ) 11.0 Hz, 10′-CH-), 6.37 (d, 1H, J ) 15.0 Hz,
2-CH-), 6.41 (d, 1H, J ) 15.0 Hz, 12′-CH-), 6.55 (d, 1H, J
22
of Au(111) on mica and left to adsorb for ca. 1 h. Samples
1
)
1
were then rinsed with freshly distilled toluene and imaged under
the same solvent in a dry-nitrogen environment.
The conducting AFM (PicoSPM, Molecular Imaging, Tempe,
AZ) has a force-sensing cantilever coated with platinum and
connected to a low-noise current amplifier. The amplifier has a
16.0 Hz, 7′-CH-), 6.62-6.66 (m, 4H, 15′-CH-, 15-CH-,
1-CH-, 11′-CH-), 6.88 (d, 1H, J ) 16.0 Hz, 8′-CH-), 7.17
1
3
(
d, 2H, J ) 7.5 Hz, ArH), 7.37 (d, 2H, J ) 7.5 Hz, ArH). C
NMR (125.75 MHz, CDCl3) δ [ppm] 12.70 (19-C, 20-C, 19′-
C, 20′-C), 19.10 (3-C), 21.61 (18-C), 32.93 (4-C), 34.30 (1-
C), 39.45 (2-C, 16-C, 17-C), 43.17 (21-C), 124.47 (11′-C),
sensitivity of 1V/nA and a noise level of 0.01 pA/ Hz.
x
Cantilevers, of nominal force constant 0.05N/m, were supplied
sputter-coated with platinum by Molecular Imaging. The AFM
is operated in the conventional constant force contact mode to
produce a topographic image of the sample, while the current
signal is fed to a second input channel of the controller,
simultaneously producing a conductivity map.
125.25 (11-C), 126.32 (C-Ar), 126.68 (7-C), 126.87 (7′-C),
129.34 (5-C), 129.71 (C-Ar), 130.13 (15-C), 130.40 (15′-C),
130.67 (13′-C), 130.73 (10-C), 132.34 (14′-C), 132.50 (13-C),
133.20 (14-C), 133.23 (10′-C); 133.77 (8′-C), 135.40 (9′-C),
136.00 (C-Ar), 137.00 (C-Ar), 137.19 (12-C), 137.68 (8-C),
137.87 (6-C), 138.34 (9-C), 138.47 (12′-C).
1
-Docosanethiol. It was synthesized as described for 1 using
Results
1
-bromodocosane (500 mg, 1.28 mmol) and thiourea (2 g) in
THF (100 mL). The reaction mixture was stirred under nitrogen
at 80 °C for 18 h. TLC (silica gel, hexane) indicated that all of
the alkyl bromide was consumed. The suspension was filtered,
and the THF solution was stirred under nitrogen with powdered
KOH (2 g) at 80 °C for 30 min. The crude mixture was diluted
with 150 mL of water and extracted with CH2Cl2 (3 × 100
mL). Flash chromatography on silica gel (hexane) yielded 354
STM images of films made from solutions with increasing
concentrations of carotenoid in 1-dodecanethiol showed an
increasing density of bright spots over the otherwise normal
background images owing to the ordered alkanethiol monolay-
ers. Typical images at two concentrations are shown in Figures
2A and B. Note that the apparent size of the spots does not
appear to change with dose while the coverage increases in these
images. The spot size is highly variable, being dominated by
the geometry of the tips. Data from many such runs (summarized
in Figure 3) show that the size is, on average, about 2 nm,
independent of carotenoid concentration. Figure 3 also shows
how the number of bright spots increases smoothly with
+
mg (81%) of 1-docosanethiol. MS m/z: 342 M (342 calculated
1
for C22H46S). H NMR (500 MHz, CDCl3) δ [ppm] 0.86 (t,
3
1
H, J ) 6.6 Hz, -CH3), 1.18-1.36 (m, 38H, -CH2-), 1.32 (t,
H, J ) 7.2 Hz -SH), 1.60 (q, 2H, J ) 7.2 Hz CH2-CH2-
SH), 2.52 (q, 2H, J ) 7.2, -CH2-SH).

4008 J. Phys. Chem. B, Vol. 103, No. 20, 1999
Leatherman et al.
such as this were well fitted by an inverse square law
A
F )
(1)
2
(
Z - Z0)
where Z is the distance from the nominal contact point (where
the tip snaps into contact). Curves could be fitted with Z0 )
-
300 nm, so we fixed this parameter and extracted A as a
function of tip bias. These values for A as a function of bias
are shown in Figure 7B. They fit a parabola, as expected for a
25
Coulomb interaction. The center of the parabola is offset from
zero volts indicating that the film is intrinsically positively
charged. We observed similar behavior at many points on the
film, regardless of the carotenoid concentration (which is small
in any case) so we conclude that this charge is an intrinsic
property of the alkanethiol monolayer. It is more difficult to
understand the large negative value for Z0 as this would locate
the charge well within the metal if a point-charge model were
applied. This difficulty might be remedied by a more accurate
model of the charge distribution on the tip, as the weighted
center of the tip charge undoubtedly lies some distance above
its apex. Thus, when the tip contacts the surface, the center of
the charge distribution is some distance above the surface.
The electrostatically generated force was compensated for
by adjusting the set-point of the microscope each time the
voltage was changed. Specifically, the extra (attractive) deflec-
tion owing to the voltage change was noted and the tip pulled
back by an equivalent amount so as to compensate for the added
electrostatic force. The effect is quite small for the tips used in
these experiments, but becomes more important when sharper
tips are used.
Figure 4. AFM image of a film on Au(111) prepared from a toluene
solution containing 1 mM 1-docosanethiol. Height scale is 0 to 50 nm.
carotenoid concentration up to 2.5 × 10-5 M in the application
solution. These data indicate that over this range of concentra-
tion, the carotenoid probably resides in the film as isolated
molecules and not as aggregates. We cannot rule out the
possibility of a small aggregate (e.g., two molecules) of constant
size, but this seems unlikely.
Attempts to obtain reproducible current-voltage character-
istics from carotenoids in dodecanethiol films were not suc-
cessful, presumably because the much-longer carotenoid mol-
ecules protrude from the surrounding film and were moved or
damaged during the scans. Reproducibility was greatly improved
when carotenoid molecules were embedded in monolayers of
1
-docosanethiol, molecules of similar length to that of the
carotenes (see Figure 1).
These much longer alkanes form densely packed monolayers
which display the pits observed in monolayers made from
23
shorter alkanethiols and a typical AFM image of pure
-docosanethiol is shown in Figure 4. The film is quite ordered
1
at the molecular level, even at high levels of carotenoid doping
as shown by the molecular-resolution scan in Figure 5A. The
average alkanethiol nearest-neighbor distance is close to 0.5 nm,
as expected, but the lattice is distorted from the near-hexagonal
packing found in films made from shorter alkanethiols possibly
because these longer molecules are not in an equilibrium packing
geometry.
The pure 1-docosanethiol films were insulating up to the
highest voltages ((2V) and forces (6nN) applied and could not
be imaged by STM. In contrast, films doped with carotenoid
showed large current peaks in the conducting AFM images
Analysis of these images must take account of the current
distribution around each spot in the image. We expect that the
current will vary from a maximum at the center of the tip
because of the variation in the molecular distortion as the
molecule passes under different parts of the tip that are
24
1
7
embedded in the film. This is illustrated in the inset in Figure
8. For a tip with axial symmetry, the number of pixels, ∆N, at
a given radius, r, from its axis varies as ∆N ) Br where B is
a constant. Given a power-law dependence of current on force
(proportional to h, the indentation of the film, inset in Figure
8), and a monotonic decrease of local force with increasing r
p
(Figures 5B and C). The density of these peaks was on the same
(decreasing h) we obtain I ≈ E - FN where E, F, and p are
order, at the concentrations used, as those found for the
carotenoids in 1-dodecanethiol (Figure 2). Two successive
current images taken at 0.25 and 0.40V are shown in Figure 5
for the film prepared from a 0.025mM carotenoid solution.
Individual “bright spots” in this scan can be imaged repeatedly,
and show a marked increase in intensity when the bias is
increased from 0.25 to 0.40V.
constants. In practice, a simple power-law fits the data quite
well. This is illustrated by two representative data sets in Figure
8 where we have plotted the natural logarithm of the number
of pixels in a given current interval against the natural logarithm
of the current (in picoamps) for a given spot at two values of
substrate bias. The intercept at loge(∆N) ) 0 corresponds to
the maximum current under the tip Im. While Im could have
been selected from the data, the fitting procedure has the
advantage that all of the pixels in a spot are used in deducing
its value, and more reproducible data are obtained. The same
analysis cannot be applied to whole images because the number
of molecules contacted tends to decrease with repeated scanning,
presumably because of tip-induced damage. Accordingly, a
histogram of the number of pixels corresponding to a given
current interval was constructed for each current spot in an
image, and Im was deduced from a regression analysis of a plot
of the logarithm of the current vs the logarithm of the number
of pixels. Many (ca. 50) spots were analyzed at each value of
bias. We changed the bias in a randomized sequence, stepping
from positive to negative values, and checking that values were
consistent with data recorded earlier in order to ensure that the
Quantitative analysis of the “current spots” associated with
the carotenoids is complicated by two factors. First, the current
through a molecule is known to depend on the force used to
1
7,18
contact the molecule.
Application of a bias to the tip will
2
5
generate an electrostatic force between the tip and sample
resulting in a spurious bias dependence of the current. Second,
the shape and overall intensity of the spot clearly depends on
the geometry of the tip.
The dependence of current on force is illustrated in Figure 6
which shows a measure (Im, see below) of the peak spot intensity
plotted against the set-point force. The presence of an electro-
static force between the cantilever and sample is evident on
examination of the force-distance plot obtained on approaching
the surface. A typical example is shown in Figure 7A. Curves

Carotene as a Molecular Wire
J. Phys. Chem. B, Vol. 103, No. 20, 1999 4009
Figure 5. (A) Molecular resolution AFM image of a film prepared from a toluene solution containing 1 mM 1-docosanethiol + 0.025mM carotenoid.
B) Current image of the same film at substrate bias ) +0.25 V. (C) As in (B) with bias ) +0.40 V. Height color scale shown for A is 1 nm. Scale
bar for B and C is 0 to 16 pA. Horizontal scale scale for all 3 images is given in A. Contact force for all three images was set to 3 nN.
(
Figure 6. Maximum current (I
substrate bias ) +0.3 V for carotenoid in 1-docosanethiol (error bars
are ( standard error).
m
) as a function of contact force for
Figure 8. Natural log of current vs the natural log of the number of
pixels in a given current interval. Data are for one spot at two values
of substrate bias (as marked). The current at log
e
m
(∆N) ) 0 is I , the
maximum current under the apex of the tip. The insert shows the model
of film deformation on which this extrapolation is based. h is the
maximum indentation of the film at the apex of the tip.
Figure 7. Electrostatic interaction between tip and sample. (A) Shows
long-range attractive force for +300 mV sample bias. Solid line is fit
to inverse square law. (B) Shows interaction strength parameter A as
extracted from fits to force distance curves. Solid line is fit to a parabola
with zero charge at -0.9 V.
Figure 9. Data for I as a function of substrate bias for three different
m
sample preparations (open and closed circles and crosses). Contact
forces are 3 nN (O and ×) and 8.5 nN (b). The error bars are (1
standard error calculated from the measured distributions. The thick
solid line is a best fit to a simple ohmic model and the thin lines are
fits to a two-step electron-transfer model for values of R and λ as shown
and a potential difference between the mean Fermi energy and oxidation
level of 0.28V.
data were reproducible. We checked that the shape of the I-V
curve we deduced in this way was independent of the power-
law model by analyzing the data for each molecule in terms of
a current density over an arbitrarily chosen 2 nm diameter circle,
obtaining a similar form for the current-voltage plot.
Results for three different sample preparations using three
different tips are shown in Figure 9.
in Figure 9) over a range of (1V. We could measure no current
through the 1-docosanethiol films, but, given the measured
10
24
resistance of decanethiol monolayers (3 × 10 Ω ) and a
26
current-decay coefficient per CH2, â, of 1.06 we estimate their
16
resistance to be about 10 Ω. Thus, the carotenethiol molecules
Discussion
are over a million times more conductive than an alkane chain
4
The carotenethiol molecules do indeed act as molecular wires,
behaving like a 4.2 ( 0.7 GΩ resistor (the heavy straight line
of similar length, but still 10 times less conductive than an
2
7
ideal quantum wire.

4010 J. Phys. Chem. B, Vol. 103, No. 20, 1999
Leatherman et al.
Carotenethiol is easily oxidized electrochemically (+0.53V
(but not all) methods of contacting a molecule. The use of a
conducting AFM combined with the insertion of the molecule
into an insulating matrix makes determination of the electrical
properties of linear conducting molecules straightforward. We
anticipate that this technique will lead to a deeper understanding
of the mechanisms of molecular conduction as different mo-
lecular systems are studied in the future.
vs SCE) and we expect that oxidation (hole transport) plays a
role in the enhanced conductivity of carotene. We have recently
8
reviewed the tunneling behavior of electroactive molecules. Han
8
et al. give expressions for the tunnel current as a function of
bias for a one-step model (the electron tunnels through the
unrelaxed molecular levels without occupying them) and a two-
step model (the electron occupies the molecular orbitals long
enough for relaxation to occur). The key parameters are E0, the
formal potential relative to the mean Fermi energy of the tip
and substrate, λ, the electronic reorganization energy, and R,
the fraction of the tip-substrate bias that appears across the
molecule-to-substrate gap (1 - R being the fraction dropped
across the tip-to-molecule gap). Dropping the approximation
Acknowledgment. This work was supported by the NSF
(Grants DBI-9513233, CHE-9709272, and INT-9600282) and
Molecular Imaging Corporation. We acknowledge useful dis-
cussions with Nongjain Tao.
References and Notes
8
of high bias used in our earlier work, the expression for net
current flow for the two-step model is
(1) Joachim, C., Roth, S., Eds. Atomic and Molecular Wires; Kluwer:
Dordrecht, 1997; Vol. 341.
(
2) Platt, J. R. Electronic Structure and Excitation of Polyenes and
R R - RtmR
sm mt
ms
Porphyrins. In Radiation Biology; Hollaender, A., Ed.; McGraw-Hill: New
York, 1956; pp 71-123.
I(V) ∝
(2)
Rsm + Rtm + Rms + Rmt
(3) Kohler, B. E. Electronic Properties of Linear Polyenes. In Conju-
gated Polymers; Br e´ das, J. L., Silbey, R., Eds.; Kluwer: Dordrecht, 1991;
pp 405-434.
where R is a tunneling rate and the subscripts m, s, and t refer
to the molecule, substrate and tip. The R values are calculated
from the integrals given by Han et al.8 The first step in
estimating E0 is to obtain the Fermi energy of the combined
tip-substrate system in volts vs SCE. The geometric mean of
the values for the work function of Au(111) (5.31V) and
polycrystalline Pt (5.65V) yields the work function of the
combined system. The work function of the SCE is 4.67 V,
giving the Fermi energy of the tip-substrate system as +0.81
V SCE. The formal potential for the first oxidation of the
carotenethiol was measured by cyclic voltammetry to be +0.53V
vs SCE so that E0 ) +0.28V. Calculated curves for some values
of R and λ (as marked) are shown in Figure 9. The symmetric
behavior of the curve (I(V) ≈ I(-V)) restricts R to a value near
(
4) Jeevarajan, A. S.; Khaled, M.; Kispert, L. D. J. Phys. Chem. 1994,
98, 7777-7781.
(5) Sereno, L.; Silber, J. J.; Otero, L.; Bohorquez, M. D. V.; Moore,
A. L.; Moore, T. A.; Gust, D. J. Phys. Chem. 1996, 100, 814-821.
6) Benniston, A. C.; Goulle, V.; Harriman, A.; Lehn, J. M.; Marczince,
B. J. Phys. Chem. 1994, 98, 8, 7798-7804.
7) Wasielewski, M. R.; Johnson, D. G.; Svec, W. A.; Kersey, K. M.;
(
(
Cragg, D. E.; Minsek, D. E. Long-distance photoinduced electron transfer
through polyene molecular wires. In Photochemical Energy ConVersion;
Norris, J., Meisel, D., Eds.; Elsevier: New York, 1989; pp 135-147.
(8) Han, W.; Durantini, E. N.; Moore, T. A.; Moore, A. L.; Gust, D.;
Rez, P.; Leatherman, G.; Seely, G. R.; Tao, N.; Lindsay, S. M. J. Phys.
Chem. 1997, 101, 10719-10725.
(
9) Datta, S.; Tian, W.; Hong, S.; Reifenberger, R.; Henderson, J. I.;
Kubiak, C. P. Phys. ReV. Lett. 1997, 79, 2530-2533.
10) Collins, P. G.; Zettl, A.; Bando, H.; Thess, A.; Smalley, R. E.
0
.5 (poor fits are obtained below 0.4 or above 0.6). This is
(
surprising, given the difference between the molecule-substrate
and molecule-tip junctions, but a similar result was obtained
Science 1997, 278, 100-102.
(11) Fan, F. R. F.; Bard, A. J. Science 1995, 267, 871-874.
9
by Datta et al. for tunneling through xylyldithiol. It contrasts
(12) Yazolani, A. I.; Eigler, D. M.; Lang, N. D. Science 1996, 272,
921-1924.
1
with a value near unity for tunneling through porphyrin
8
(13) Bumm, L. A.; Arnold, J. J.; Cygan, M. T.; Dunbar, T. D.; Burgin,
molecules attached to gold. Our data do not resolve the fine
T. P.; Jones, L.; Allara, D. L.; Tour, J. M.; Weiss, P. S. Science 1996, 271,
705-1707.
14) Bockrath, M.; Cobden, D. H.; McEuen, P. L.; Chopra, N.; Zettl,
structure needed to discriminate between extreme values for λ.
For the same reason, we cannot discriminate between the one-
and two-step model.
1
(
A.; Thess, A.; Smalley, R. E. Science 1997, 275, 1922-1925.
Finally we note that there is considerable variation in the
current measured from molecule to molecule and from the same
molecule from scan to scan (reflected in the relatively large
standard error bars in Figure 9). This suggests that a significant
fraction of the observed resistance originates from the tip-to-
molecule contact, an observation at odds with the conclusion
that R ) 0.5. Further systematic studies of the dependence of
the resistance on molecular structure are needed to elucidate
this point.
(15) Reed, M. A.; Zhou, C.; Muller, C. J.; Burgin, T. P.; Tour, J. M.
Science 1997, 278, 252-254.
(
16) Lindsay, S. M.; Sankey, O. F.; Li, Y.; Herbst, C. J. Phys. Chem.
990, 94, 4655-4660.
17) Durig, U.; Zuger, O.; Michel, B.; Haussling, L.; Ringsdorf, H. Phys.
ReV. 1993, B48, 1711-1717.
18) Joachim, C.; Gimzewski, J. K.; Schlittler, R. R.; Chavy, C. Phys.
1
(
(
ReV. Lett. 1995, 74, 2102-2105.
(19) Anselmetti, D.; Baratoff, A.; Guntherodt, H. J.; Gerber, C.; Michel,
B.; Rohrer, H. J. Vac. Sci. Technol. 1994, B12, 1677-1680.
(
20) Gust, D.; Moore, T. A.; Moore, A. L.; Liddell, P. A. Methods
Enzymol. 1992, 213, 87-100.
21) Wandlowski, T.; Lampner, D.; Lindsay, S. M. J. Electroanal. Chem.
Conclusions
(
Carotenoids are excellent conductors of electrons when
compared with saturated hydrocarbons, facilitating measure-
ments of the electrical properties of individual molecules in an
insulating matrix. Reproducible current-voltage data are ob-
tained over a range of (1V when the samples are held in an
environment that is free of moisture and molecular oxygen. To
a first approximation, the molecules act as ohmic resistors,
implying that half of the applied potential appears at the
molecular orbital. This behavior has been observed in another
molecular system, and it may therefore be characteristic of some
1996, 404, 215-226.
(22) DeRose, J. A.; Lampner, D. B.; Lindsay, S. M. J. Vac. Sci. Technol.
1
993, A11, 776-780.
23) McDermott, C. A.; McDermott, M. T.; Green, J. B.; Porter, M. D.
J. Phys. Chem. 1995, 99, 9, 13257-13267.
(
(24) Poirier, G. E.; Tarlov, M. J. Langmuir 1994, 10, 0, 2853-2856.
(25) Martin, Y.; Abraham, D. W.; Wickramasinghe, H. K. Appl. Phys.
Lett. 1988, 52, 1103-1105.
26) Carter, M. T.; Rowe, G. K.; Richardson, J. N.; Tender, L. M.;
(
Terrill, R. H.; Murray, R. W. J. Am. Chem. Soc. 1995, 117, 7, 2896-2899.
(27) Landauer, R. Philos. Mag. 1970, 21, 863-867.