 |
Description  |
|
|
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to scanning probe microscopes, in particular the
scanning tunneling microscope (STM) and the atomic force microscope (AFM).
Identification of the chemical composition of areas in a microscope image
has been very difficult in the past. Certain elements can be identified by
their characteristic x-ray emission in an electron microscope. In the case
of the STM, identification has been limited to certain atoms which induce
well-understood surface states near the Fermi energy (the energy of the
tunneling electrons). In the case of the AFM, certain discrimination
between the composition of molecular adlayers has been possible, based on
differences in friction between the adlayers and the scanning tip. In the
present invention, we exploit the ability of a conductive tip to transfer
charge to and from molecules in surfaces at well defined potentials (being
the electrochemical reduction or oxidation potentials). The present
invention permits discrimination among, and identification of
electroactive molecules on the surface of a sample. The present invention
provides the first known method for identification of organic molecules in
a microscope with nanometer scale resolution.
2. The Prior Art
The scanning tunneling microscope (STM) is capable of atomic-resolution
imaging of a conductive surface [Binnig, G. and Rohrer, H., Reviews of
Modern Physics, vol. 59, pp. 615-626, 1987]. The atomic force microscope
can image single atoms in an insulating surface [Ohnesorge, F. and Binnig,
G., Science vol. 260, pp. 1451-1456, 1993]. However, neither technique is
well suited to identification of the composition of material in the gap
formed between the probe and an underlying substrate. In the case of the
STM, current is carried by electrons in the itinerant states of the metals
that constitute the tip or substrate. The composition of some intervening
material is only of significance to the extent that it modifies the
properties of those states. In certain very special cases, it has proved
possible to identify surface atoms, based on the manner in which they
modify the current carrying states near a surface [Feenstra et al.,
Physical Review Letters, vol. 58, pp. 1192-1195, 1987]. In the case of the
AFM, the intervening material plays a role in the friction between the
scanning probe when there are chemically-specific interactions between the
scanning probe and molecules under the probe, a phenomenon that has been
used to distinguish (but not identify) regions of different chemical
composition in a thin film [Overney et al., Nature, 359, pp. 133-135,
1992].
An alternative approach to chemical identification uses thin films
sandwiched between metal electrodes. A voltage is applied between the
electrodes so as to raise the energy of the electrons in one electrode
with respect to the other electrode. When the energy of electrons in one
electrode is coincident with an electronic state of a molecule in the thin
film between the electrode, an enhanced current flow occurs because of the
process of resonant tunneling, a quantum-mechanical phenomenon in which
the intermediate state in the gap serves to transport extra current.
Because the energy of the molecular state is characteristic of the
chemical species in the gap between the electrodes, the voltage at which
this extra current flows is characteristic and could, in principle, be
used to identify the chemical species. FIG. 1 shows a schematic
arrangement of such a solid-state tunnel junction 10. A voltage V is
applied by device 12 across two metal electrodes, 14 and 16. Each
electrode is coated with a thin insulating film (such as an oxide layer)
18 and 20 and a layer of molecules 22. A sensitive current measuring
device 24 records the current through the device. FIGS. 2A and 2B show the
energy of the electrons in the electrodes of FIG. 1 schematically in two
configurations: FIG. 2A is a diagram of voltage conditions where there is
no extra current due to resonant tunneling. The voltage applied across the
device, V.sub.1 is too little to raise the energy of the electrons in
electrode 14 so as to be coincident with the energy of the molecular state
E.sub.M. FIG. 2B is a diagram of the situation when the voltage is
adjusted to resonance. The electrons that carry current from electrode 14
now have an energy equal to E.sub.M. The voltage, V.sub.2 at which the
extra current "turns on" serves to identify the molecule in the gap. A
diagram of the current-voltage characteristic of such a device is shown in
FIG. 3. Conventionally, the step at V.sub.2 is detected by plotting the
first or second derivative of the current so that features are made
sharper.
While the above description has long been supposed to apply to tunneling
through molecules, some recent work shows that the situation is both more
complex, and yet more tractable in terms of achieving the desirable goal
of identifying a broad range of molecules by such a mechanism. Mazur and
Hipps [Journal of Physical Chemistry, submitted, 1994] have measured the
current-voltage characteristics of a number of devices containing
different organic molecules with states that lie some electron volts from
the energy of the electrons with no voltage applied across the device.
They have extracted the value of the voltage at which the extra current
turns on (V.sub.2 in FIG. 3) for a number of different organic molecules.
They find that the energy of the state, E.sub.M, at which increased
current flow is detected, is not the energy that would be measured for the
same molecule in the gas phase. It is, instead, the energy of the final
state that occurs when the molecule is electrochemically reduced or
oxidized. This is different from the energy of the isolated molecule for
two reasons. First, oxidation or reduction involves charging of the
molecule, a process that changes the energy of the states of the molecule.
In contrast, in resonant tunneling, the process described above, the
electron does not interact with the molecule for long enough to change its
energy. Second, the charged molecule is embedded in a dielectric medium.
In this case it is the insulating films 18, 20 and other molecules that
constitute the dielectric, but in an electrochemistry experiment, it is
the solvent used to dissolve the molecules. In either case, the medium
polarizes so as to reduce the energy of the charged state. This final step
is called `relaxation`. In any case, the charging (reduction) or
discharging (oxidation) of a molecule in a medium is a much more complex
process than resonant tunneling. However, the magnitude of the energy
shift caused by relaxation is usually big; i.e., many electron volts, so
that the states associated with oxidized or reduced molecules lie closer
to the energy of the electrons in the metal than the original,
unperturbed, states of the molecule. More importantly, from the standpoint
of the present invention, these state-energies are easily measured by the
conventional methods of electrochemistry. Many sources list standard
reduction and oxidation potentials for organic compounds.
Electrochemical potentials are conventionally stated as potentials relative
to the electrochemical potential of a standard `reference electrode`.
Thus, identification of molecules via their reduction or oxidation
potentials would seem to require an electrochemical cell containing such
an electrode as a reference. However, these reference electrodes function
because their potential is fixed. That is to say, a certain fixed amount
of work would have to be done to remove an electron from such an electrode
to a position at rest far from the electrode. This quantity is the work
function of the reference electrode. It is illustrated schematically in
FIG. 4 where the energy to take an electron from the reference electrode
is labeled .phi..sub.REF. The (known) oxidation and reduction potentials
are labeled E.sub.RER (OX) and E.sub.RER (RED). The work function of the
metal used for an electrode in a tunneling device is also usually a known
quantity, .phi..sub.METAL. Thus, the voltage for reduction (V.sub.2 (RED))
or oxidation (V.sub.2 (OX)) of molecules in a device like that shown in
FIG. 1 can be calculated if .phi..sub.REF is known. The standard for
reference electrodes is the Normal Hydrogen Electrode, NHE, and values for
the potential of other reference electrodes relative to the NHE are well
known. The work function of the NHE, .phi..sub.NHE, is still the subject
of some debate, although 4.8 eV is the currently accepted value [Trasatti,
S., Advances in Electrochemistry, ed. H. Gerischer, C. W. Tobias, Wiley
InterScience, New York, pp. 213-321 ].
Mazur and Hipps have used the value of .phi..sub.NHE =4.8 eV together with
the known oxidation and reduction potentials of several organic molecules
and the work function of lead (4.1 eV) to calculate the reduction
potential for these molecules between lead electrodes in a device such as
that shown in FIG. 1. The potential is calculated as V.sub.2 (RED), the
voltage that would have to be applied to the device in order to see a
step-like increase in current due to the reduction of the molecules. TABLE
1 includes a listing of the calculated V.sub.2 (RED), and the measured
voltage, V.sub.2, at which a step occurs in the current for six organic
molecules.
TABLE I
______________________________________
V.sub.2 V.sub.2
Molecule (RED) Calculated (Volts)
Measured (Volts)
______________________________________
Ni(acac).sub.2
2.47 2.50
coronene 2.62 2.51
anthracene
2.80 2.74
perylene 3.00 2.91
tetracene
3.12 3.11
pentacene
3.40 3.47
______________________________________
On the whole, there is rather good agreement between the calculated and
measured values. Thus, this step in current at V.sub.2 serves as a marker
that may be used to identify the organic compound. Clearly, this method
can be extended to other organic compounds and other electrodes. Compounds
that are reduced at more negative potentials could be studied on
electrodes with larger work functions. Thus, this method of chemical
identification is applicable to any compound that can be reduced (or
oxidized) on any metal suitable for use as an electrode.
The limitation of the prior art is that, in order to carry out
identification of molecules, they must be somehow inserted into a device
of the general layout shown in FIG. 1. This is not easy to do and not very
useful once done, for one must usually know the chemistry of the molecules
in advance in order to make a device such as that shown in FIG. 1.
SUMMARY OF THE INVENTION
The present invention provides a method and apparatus for high resolution
mapping of the chemical composition of a thin film using scanning probe
microscopy techniques. The sample to be studied is prepared as a thin film
disposed on a conductive backing electrode. A sensitive electrometer is
connected to the backing electrode to detect current passing through it.
According to a first aspect of the invention, a force sensing cantilever
is scanned relative to the sample surface a plurality of times.
Topographic information about the sample surface is obtained in a
conventional manner by studying deflections of the cantilever or feedback
current used to minimize deflections of the cantilever. Simultaneously, a
voltage is applied to the probe tip. This voltage, by means of a tunneling
current to the backing electrode, causes reduction and/or oxidation
reactions in the sample surface. On successive scans, different voltages
may be used. In this way, the tunneling current at each of a number of
different voltages for each location in the sample surface is obtained.
Because specific oxidation and reduction reactions take place only at well
defined voltages, it is possible, from the current measured at a certain
location and a certain applied voltage at that location, to deduce what
the chemical located at that location is. According to a second aspect of
the invention, a scanning tunneling microscope mechanism may be used
instead of a force sensing mechanism.
OBJECTS AND ADVANTAGES OF THE INVENTION
Accordingly, it is an object of the present invention to provide a
microscope which can determine the chemical composition of regions of a
sample surface.
It is a further object of the present invention to provide a microscope in
which the chemical composition of a thin film is mapped with high
resolution.
It is a further object to provide a microscope that will map the chemical
composition of the surface of an insulating film.
Still another object of the present invention is to provide a microscope
that will provide a simultaneous topographical and chemical display of the
surface of a sample under examination.
These and many other objects and advantages of the present invention will
become apparent to those of ordinary skill in the art from a consideration
of the drawings and ensuing description of the invention.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a schematic diagram of a solid state tunnel junction according to
the prior art.
FIG. 2A is a diagram showing energy levels in a tunnel junction according
to FIG. 1 with the tunnel junction not in a resonance condition.
FIG. 2B is a diagram showing energy levels in a tunnel junction according
to FIG. 1 with the tunnel junction in a resonance condition.
FIG. 3 is a current-voltage characteristic for the tunnel junction of FIG.
1.
FIG. 4 is an energy diagram showing the energy levels of a reference
electrode, molecular oxidation and reduction states and the work function
of a metal electrode.
FIG. 5 is a schematic diagram of a scanning force microscope according to
the present invention.
FIG. 6 is a diagram showing a thin-film sample assembly for use with the
scanning force microscope shown in FIG. 5.
FIG. 7 is an electrical schematic diagram of the electrometer of the
present invention.
FIG. 8 is a block diagram of the control system for the microscope of FIG.
5 according to a presently preferred embodiment of the present invention.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
Those of ordinary skill in the art will realize that the following
description of the present invention is illustrative only and is not
intended to be in any way limiting. Other embodiments of the invention
will readily suggest themselves to such skilled persons from an
examination of the within disclosure.
The essential elements of the microscope according to the present invention
are shown in FIG. 5. A force sensing cantilever (such as those sold by
Park Scientific Instruments of Sunnyvale, Calif.) 22 with a probe tip 23
is coated with a thin metal film, 24. In the preferred embodiment, the
cantilevers are DC ion-sputter coated. A thin layer (a few angstroms) of
chrome is first applied to improve adhesion of subsequent coatings. Next,
a layer of several hundred angstroms of gold is sputtered onto the
cantilevers. DC ion-sputter coated films generate less film stress and do
not bend the force sensing cantilevers significantly. The cantilever is
scanned over the upper surface of the sample 26 in a raster or equivalent
pattern by the scanner 28 which also adjusts the height of the cantilever
22 above the surface 26. According to a preferred embodiment of the
present invention, the cantilever 22 is grounded. This eliminates leakage
of the high voltage signals used in the scanner 28 (typically a
piezoceramic transducer controlled by high voltage signals).
Other metals besides gold may be used. If freshly prepared, even quite
reactive metals will only oxidize to a depth of a few angstores, so they
are still useful as tunneling electrodes. The advantage of using another
metal is that the onset currents for oxidation or reduction are all
shifted by the amount by which the work function differs. The data shown
in Table I are for lead electrodes, for which the work function is 4.1 eV.
Gold electrodes have a work function of 5.2 eV, so the voltages listed in
Table 1 would all be reduced by 1.1 volts. That is, in a gold-gold
electrode system, the measured voltages would be 1.4 V for Ni(acac).sub.2,
1.41 V for coronene, 1.64 V for anthracene, 1.81 V for perylene, 2.01 V
for tetracene and 2.37 V for pentacene. Changing the electrode metals
serves two functions: (1) it can be used to confirm assignments made with
the electrode system and (2) the range of materials that can be analyzed
in a given range of applied voltages can be extended by as much as the
work function of the electrodes can be changed. Data for work functions
for various metals can be found, for example, in the CRC Handbook of
Chemistry and Physics (CRC Press, Boca Raton, Fla.).
The sample 26 is prepared in the form of a thin film with a conductive
backing 30 as described further below. The conductive backing of "back
electrode" 30 is connected into a sensitive electrometer 34 (capable of
detecting electrical currents as low as 0.01 pA) and a bias voltage 32
applied to the backing electrode 30 (effectively between the probe tip 23
and the backing electrode 30). Displacements of the force-sensing
cantilever 22 are sensed by deflection of a laser beam 36 from a laser 38
which is reflected off the back 40 of the cantilever 22 and into a
position-sensitive detector 42.
The detailed layout of a microscope which incorporates most of these
features is described in our co-pending U. S. patent application Ser. No.
08/388,068, referred to above. In particular, it is important that the
sample surface 26 is clean and free of water. This is required to prevent
unwanted electrochemical reactions that limit the range of potential that
can be applied across the sample. This is achieved in the present
invention by enclosing the sample area in a hermetically-sealed chamber
such as that described in the above-referenced U.S. patent application
Ser. No. 08/388,068. Inert gasses, such as argon, may be flowed through
the chamber in order to keep the sample dry and clean.
Preparation of a film thin enough for efficient electron transfer is an
important step. If the film is prepared as a monolayer on a metal film, it
is straightforward to use. Examples of such preparation methods are
dipping a metal electrode into a Langmuir trough on which the sample is
floated as a surfactant, and use of standard chemical methods, such as the
use of alkyl-thiols on gold electrodes. Another method is to use standard
electrochemical methods to deposit a thin layer onto an electrode and then
to remove the electrode form the electrochemistry cell. In the most
general case, the sample is a solid, possibly a section of a biological
material such as a cell. In this case the standard microtome methods used
for transmission electron microscopy may be used to prepare a thin film.
Freezing and subsequent sectioning permits fabrication of films that are
thinner than 100.ANG., quite routinely. Such films must then be contacted
electrically. A procedure for doing this is illustrated in FIG. 6. The
sample 44 is placed in an evaporator and a thin film (on the order of
20.ANG.,) of gold 46 evaporated onto one side in order to establish a back
electrode. Placed onto a clean, flat gold film, the back electrode
spontaneously bonds to the underlying gold support. A suitable gold
support is made by evaporating a gold film 48 of a few thousand angstroms
thickness onto a mica sheet 50. This process is referred to as placing the
sample surface on a conductive backing electrode.
The electrometer used in a preferred embodiment is shown in FIG. 7. The
signal on line 52 from back electrode 30 is sent to a current-to-voltage
converter 54. A "dummy" signal on line 56 from a second lead, placed in
close proximity to line 52, is fed to a second, identical current to
voltage converter 58. The signals from both current to voltage converters
are subtracted in a differential amplifier 60 to give a final output
signal 62. This arrangement cancels any common-mode noise signals that are
common to the dummy signal lead and the signal lead, leading to a
reduction in noise.
The system is controlled as shown in FIG. 8. The signal on line 64 from the
position sensitive detector 42 is sent to a conventional scanning probe
microscope controller 66 (such as the NanoScope III available from Digital
Instruments of Santa Barbara, Calif.) which generates the x, y raster-scan
signals on line 68 that are used to position the tip 23 in the plane of
the sample 26.
The signal that controls the height (z-axis) of the tip 23 is controlled by
a signal on line 70 which also is used to form a topographical image of
the surface of the sample 26. The signal on line 62 from the
current-to-voltage converter (FIG. 7) is fed to a computer 72 (in FIG. 8)
as is the x-y raster scan signal on line 68. Computer 72 also generates a
series of bias voltages on line 74. On successive scans, this bias voltage
is incremented. The computer 72 generates an image on the chemistry
display 76 which shows the difference in current between the previous scan
and the current scan as a function of the position of the tip over the
surface. The computer is programmed to show larger current increments as
brighter regions. In this way, regions over which increased current flow
occurs at a particular voltage show up as bright patches on the screen.
These voltages are correlated to known oxidation or reduction potentials
in order to identify the molecules responsible for the increased
brightness. Other display mechanisms may also be used as would be apparent
to those of ordinary skill in the art, such as false color mapping of
locations associated with relatively high measured currents--these
locations, in conjunction with the bias voltage applied at the time,
correlate with the presence of a particular chemical substance. A
conventional topography display 78 is driven by the microscope controller
66 in a conventional manner. It is also possible to combine the topography
image on display 78 and the chemistry image on chemistry display 76 to
provide a single combined image showing both topography and chemical
composition of the sample surface.
A certain distortion of the response occurs because the onset voltage of
the reduction or oxidation process will be related to the position of the
molecules in the potential gradient (electric field) between the tip and
the back electrode. However, the onset for the most favorably placed
molecules will still occur at a well defined potential. The net effect of
this is to give a signal that is primarily sensitive to the molecules in
the surface of the film. Molecules at other positions contribute to
currents at higher voltage, but their spatial distribution usually
broadens the current steps associated with the onset of reduction or
oxidation so that sharp features are not seen. The resolution attainable
with this technique is somewhat better than the radius of curvature of the
coated tip (typically about 10 nm).
The current due to oxidation or reduction processes is limited by the rate
at which the reduced or oxidized molecules transfer charge to the backing
electrode. However, even if this is a very slow process (as in a thick
sample) the current transient that results from the initial charging or
discharging is still significant. To see how this can be, consider a tip
contacting an area of 10 nm diameter and being swept across the surface at
20,000 nm/s (a typical speed). The tip sweeps out an area of 2E5 nm.sup.2
(2.times.10.sup.5 nm.sup.2) each second. If there is one electroactive
molecule in each 20 by 20 angstroms of the surface and one electron is
transferred, then the corresponding current is about 0.03 pA which is
quite easy to detect with the electrometer of FIG. 8.
Generally, oxidation (or reduction) peaks are quite broad, that is to say,
linewidths are on the order of 0.1 V. The range of voltages that can be
scanned without dielecric breakdown is on the order of .+-.10 V. Thus, the
entire us | | |
