Thermal Dip Pen Nanolithography For Direct Writing of Polymer

Thermal Dip Pen Nanolithography for Direct Writing of Polymer-NanoParticle Composites
Expert Reviews
Written by:
Jeff Morse, Ph.D
Lee et. al. demonstrated a maskless, additive approach for direct writing of a range of NP–polymer
compositions using thermal dip-pen nanolithography (tDPN). This technique represents a facile strategy
for deposition of a wide range of NP composite materials in order to create dense nanopatterned films.
Reviewed by Jeff Morse, Ph.D, National Nanomanufacturing Network
 Lee WK, Dai Z, King WP, and Sheehan PE. 2009. Maskless Nanoscale Writing of Nanoparticle-Polymer Composites and Nanoparticle Assemblies Using
Thermal Nanoprobes. Nano Letters 10(1): 129-133. DOI: 10.1021/nl9030456.
Controlled nanoparticle (NP) assemblies have a range of applications for electronics, optical devices, and sensors. While NP positioning with atomic force
microscope tips has been demonstrated to provide extreme accuracy, patterning rates are not viable for high rate nanomanufacturing. Particle driven and
templated assemblies of NPs provide an additional means of creating precision NP compositions, but require custom chemistries for each NP material and
compositional makeup or a templated patterning to facilitate assembly. Furthermore, the overall fabrication becomes quite complex as chemistry, patterning,
additive, and subtractive process steps must be integrated, with each step increasing the probability for cross contamination of the composite matrix or
degradation of the NPs.
A polymer-nanoparticle composite is deposited from a heated cantilever tip. The

deposited composite may be used as is or the polymer removed with an oxygen
plasma. The nanoparticles can be either focused into a row or dispersed depending
on the size and chemical treatment of the nanoparticles.Dip pen nanolithography
(DPN) is a process that enables patterning of NP inks and composites using
solvent-based solution dispersions. As such, NP compositions are limited to specific
stable inks that can be annealed and dried after deposition and patterning. An
emerging nanomanufacturing tool uses thermally-controlled deposition from a
heated tip, known as thermal dip-pen nanolithography (tDPN). In this approach, the
deposition of nanoscale quantities of materials that are otherwise difficult to deposit
directly, such as metals, semiconductors, and polymers, is possible using a heated
atomic force microscope (AFM) probe tip that melts a solid ink and then directs the
ink flow onto a surface.
Recently, Lee et. al. demonstrated a maskless, additive approach for direct writing
of a range of NP–polymer compositions. In this work, the authors investigated the
controlled assembly of the NP and NP-polymer composites by adjusting tip
temperature and speed. In order to demonstrate the versatility of this approach,
several materials compositions were investigated, including poly(methyl
methacrylate) (PMMA) with Fe2O3 NPs, polyethylene with CdSe/ZnS quantum dots,
metallorganic tris(8-hydroxyquinolinato)aluminum mixed with the piezoelectric
polymer P(VDF-TrFE), and Au NPs mixed with the undoped conductive polymer
PDDT. The authors achieved direct write line-widths of the NP-polymer composites
on the order of 100nm. In order to obtain NP patterns, the polymer matrix was
selectively removed using an oxygen plasma etch step. As an example, after
removal of the PMMA matrix, it was observed that the remaining Fe2O3 NP trace was significantly narrower than the original composite pattern, in this case the
pattern was reduced from 360 nm to 10 nm in width. Thus NP concentration or focusing is possible.
In order to better understand and possibly control the NP focusing effect, the authors suggested that the critical balance between NP diffusion and shear flow as
the particles are dispersed enables the concentration and alignment for NP assembly. The effect would ultimately have dependencies on solution chemistry, NP
size and functionality, and solution viscosity, which could be controlled in part by tip temperature and write speed. The authors demonstrated that write speeds of
~200 µm/sec for polymer blends, but only <2 µm/sec for NP-polymer composites. This technique represents a facile strategy for deposition of a wide range of NP
composite materials in order to create dense nanopatterned films. The process has further demonstrated critical features of 10nm for patterned NPs exploiting the
shear flow focusing effect, and 78 nm for as deposited NP-polymer composites. This approach further provides a complimentary tool for more complicated
integration approaches incorporating top down and bottoms up processes.
Nanoparticle polymer composites containing metal, semiconductor, magnetic, and optically active nanoparticles were deposited onto multiple substrates from a
heatable atomic force microscope tip. The nanoparticle nanostructures were functional as deposited or could be etched with an oxygen plasma, revealing single
nanoparticle lithographic resolution. Many types of nanoparticles can be patterned with the same technique, without the need to tailor the substrate chemistry and
without solution processing.
Classic, liquid, and matrix-assisted dip-pen nanolithography for materials research
Jian Zhong,*a Gang Sun*b and Dannong Heac
Author affiliations
Abstract
As a powerful atomic force microscopy-based nanotechnological tool, dip-pen nanolithography (DPN) has provided an ideal direct-write “constructive”
lithographic tool that allows materials to be patterned from DPN tips onto a surface with high registration and sub-15 nm resolution. In the past few decades, DPN
has been enormously developed for studying the patterning of inorganic, organic, and biological materials onto a variety of substrates. The focus of this review is
on the development of three types of DPN: classic, liquid, and matrix-assisted DPN. Such development mainly includes the following aspects: the comparisons of
three types of DPN, the effect factors and basic mechanisms of three types of DPN, and the application progress of three types of DPN.
Patents
Find prior art
Discuss this application
View PDF
Download PDF
Try the new Google Patents, with machine-classified Google Scholar results, and Japanese and South Korean patents.
Publication number US20100071098 A1
Publication type Application
Application number US 12/465,616
Publication date 18 Mar 2010
Filing date 13 May 2009
Priority date 13 May 2008
Also published as EP2291849A2, 20 More »
Inventors Chad A. Mirkin, 9 More »
Original Assignee Northwestern University, Nanolnk, Inc.
Export Citation BiBTeX, EndNote, RefMan

Patent Citations (17), Referenced by (13), Classifications (17)
External Links: USPTO, USPTO Assignment, Espacenet
Scanning probe epitaxy

US 20100071098 A1
ABSTRACT
A dual tip probe for scanning probe epitaxy is disclosed. The dual tip probe includes first and second tips
disposed on a cantilever arm. The first and second tips can be a reader tip and a synthesis tip,
respectively. The dual tip probe further includes a rib disposed on the cantilever arm between the first and
second tips. The dual tip probe can also include a strain gauge disposed along the length of the cantilever
arm.
IMAGES(50)
CLAIMS(22)
1. A dual tip micro probe, comprising:
a micro cantilever arm comprising first and second arm ends;
a first tip disposed on the arm adjacent to the second arm end; and
a second tip disposed on the arm adjacent to the first tip; and
a rib disposed on the cantilever arm between the first and second tips.
2. A dual tip micro probe, comprising:
a micro cantilever arm comprising first and second arm ends;
a first tip disposed on the arm adjacent to the second arm end, wherein the first tip is a non-synthesis tip;
a second tip disposed on the arm adjacent to the first tip, wherein the second tip is a synthesis tip; and
a strain gauge disposed along a length of the cantilever arm.
3. The dual tip probe of claim 1 wherein the arm comprises first and second arm sections disposed
between the first and second arm ends, the first arm section having a width greater than the second arm
section, and the first and second tips being disposed on the second arm section.
4. The dual tip probe of claim 1, wherein the first and second tips are inverted pyramids comprising a
base and an apex, wherein the base is operatively coupled to the arm.
5. The dual tip probe of claim 4, wherein the second tip further comprises an aperture formed at the apex.
6. The dual tip probe of claim 5, wherein the aperture has a diameter in a range of about 5 nm to about
200 nm.
7. The dual tip probe of claim 1, wherein the second tip is a hollow tip.
8. The dual tip probe of claim 1, wherein the first tip and the second tip have the same thickness.
9. The dual tip probe of claim 1, wherein the first tip has a first thickness and the second tip has a second
thickness different from the first thickness.
10. The dual tip probe of claim 1, wherein the first tip is a non-synthesis tip and the second tip is a
synthesis tip for forming nanostructures on or in a substrate.
11. The dual tip probe of claim 10, wherein the first tip is a surface topology measurement tip.
12. The dual tip probe of claim 1, wherein the second tip is a synthesis tip selected from the group
consisting of an aperture probe synthesis tip, a catalyst tipped synthesis tip, an elastomer gel synthesis
tip, a polymer synthesis tip, a high temperature synthesis tip, and an electric field control synthesis tip.
13. The dual tip probe of claim 12, wherein the synthesis tip is a high temperature synthesis tip and the
high temperature synthesis tip comprises a resistive heater disposed on the tip.
14. The dual tip probe of claim 1, wherein the first tip is formed from a material selected from the group
consisting of silicon, silicon oxide, silicon nitride, gold, silver, copper, and tungsten.
15. The dual tip probe of claim 1, wherein the second tip is formed of a material selected from the group
consisting of doped polysilicon, a doped diamond, a metal, a hard metal, and a metal oxide.
16. The dual tip probe of claim 15, wherein the metal is selected from the group consisting of Au, Al, Ag,
Ni, Fe, Pt, Os, Ru, In, Ir, W, and Cr.
17. The dual tip probe of claim 15, wherein the hard metal is selected from the group consisting of TiN,
TiC, WC, and TaN.
18. The dual tip probe of claim 15, wherein the metal oxide is selected from the group consisting of InO
and IrO2.
19. The dual tip probe of claim 1, wherein the arm is formed from a material selected from the group
consisting of silicon, silicon nitride, silicon oxide, silver, gold, aluminum, tungsten, and copper.
20. The dual tip probe of claim 1, wherein the first and second tips are separated by a distance in a range
of 1 μm to 50 μm, measured as the distance from the apex of the first tip to the apex of the second tip.
21. The dual tip probe of claim 1, further comprising a strain gauge disposed along a length of the
cantilever arm.
22. The dual tip probe of claim 2, further comprising a rib disposed on the arm between the first and
second tips.
DESCRIPTION
CROSS-REFERENCE TO RELATED APPLICATION
 [0001]
The benefit under 35 U.S.C. §119 (e) of U.S. Provisional Patent Application No. 61/052,864 filed
May 13, 2008, and U.S. Provisional Patent Application No. 61/167,853, filed Apr. 8, 2009, the
disclosures of which are incorporated herein by reference, is hereby claimed.
STATEMENT OF GOVERNMENTAL INTEREST
 [0002]
This invention was made with government support under Grant No. N6601-08-1-2044 awarded
by the Space and Navel Warfare Systems Center. The government has certain rights in the
invention.
BACKGROUND
 [0003]
1. Field of the Disclosure
 [0004]
The disclosure generally relates to a probe for scanning probe epitaxy. In particular, the
disclosure relates to a probe for scanning probe epitaxy having a first and second tip disposed on
a cantilever arm and a rib disposed between the first and second tip. The disclosure further
relates to a probe for scanning probe epitaxy having first and second tips disposed on a
cantilever arm and a strain gauge disposed along the length of the cantilever arm.
 [0005]
2. Brief Description of Related Technology
 [0006]
There a variety of known tip-based methods of synthesizing nanostructures on a surface.
Different capabilities are needed for the synthesis of different nanostructures. For example, for
the synthesis of quantum dots, the ability to directly delivery a reactive chemical reagent to a
second reagent on a surface in order to make a binary structure is generally needed. The
synthesis of quantum dots and nanoparticles can also utilize the application of an electric field
that transforms a tip into a nanoevaporator capable of depositing nanoscopic amounts of a high
vapor pressure material on a surface. Control of the tip temperature up to hundreds of degrees
can be utilized to facilitate the direct catalytic growth of solid state nanostructures like carbon
nanotubes and semiconductor nanowires. There are no currently available methods for realizing
such capabilities with commercially or even academic laboratory-available scanning probe
systems.
SUMMARY
 [0007]
In accordance with an embodiment of the invention a dual tip micro probe includes a micro
cantilever arm comprising first and second arm ends, a first tip disposed on the arm adjacent to
the second arm end, a second tip disposed on the arm adjacent to the first tip, and ribs disposed
on the cantilever arm between the first and second tips.
 [0008]
In accordance with an embodiment of the invention a dual tip micro probe includes a micro
cantilever arm comprising first and second arm ends, a first tip disposed on the arm adjacent to
the second arm end, and a second tip disposed on the arm adjacent to the first tip, and a strain
gauge disposed along a length of the cantilever arm.
BRIEF DESCRIPTION OF THE DRAWINGS
 [0009]
FIGS. 1A and 1B are schematics of dual tip probes in accordance with embodiments of the
disclosed invention. FIG. 1B illustrates a dual tip probe having strain gauges. FIG. 1C is a
schematic of a dual tip probe in accordance with an embodiment of the disclosed invention
illustrating the relationship between probe dimensions, cantilever bending, and distance between
the second tip and the substrate.
 [0010]
FIGS. 2A and 2B are scanning electron microscopy (SEM) images of dual tip probes in
accordance with embodiments of the disclosed invention.
 [0011]
FIGS. 3A-3B are AFM images of in situ correction of a pattern using a dual tip probe in
accordance with an embodiment of the disclosed invention. FIG. 3D is AFM height images of the
patterns of FIG. 3A. FIG. 3E is an SEM image of the pattern of FIG. 3A and an energy dispersive
x-ray spectrum of the pattern.
 [0012]
FIG. 4A is an SEM image of an ultra sharp tip having a 10 nm radius of curvature. FIG. 4B is a
graph showing the convoluted tip results of the tip of FIG. 4A. FIG. 4C is SEM images of a sharp
tip before and after writing, showing the wear on a tip that occurs during writing.
 [0013]
FIG. 5 is a schematic of a cantilever arm of a dual tip probe in accordance with an embodiment of
the disclosed invention having two sections of different widths and lengths.
 [0014]
FIG. 6A is an optical microscopy image of a dual tip probe in accordance with an embodiment of
the disclosed invention having a cantilever arm with two sections of different widths. FIG. 6B is an
optical microscopy image of a dual tip probe in accordance with an embodiment of the disclosed
invention having a rib disposed between the first and second tips.
 [0015]
FIG. 7 is an SEM image of a dual tip probe in accordance with an embodiment of the invention
having an aperture in the second tip.
 [0016]
FIG. 8A is an SEM image of a thermal dual tip probe in accordance with an embodiment of the
claimed invention. FIG. 8B is an optical image of the dual tip probe of FIG. 8A. FIG. 8C is a
simulated model of a thermal dual tip probe in accordance with an embodiment of the
invention. FIG. 8D is a simulated temperature distribution of a thermal dual tip probe. FIG. 8E is a
thermal microscopy image of a thermal tip of a thermal dual tip probe. FIG. 8F is a graph of tip
temperature verses voltage comparing simulation results with measured results. FIG. 8G is a
temperature distribution obtained from the simulation shown in FIG. 8F. FIG. 8H is a graph
showing the maximum temperature verses the voltage bias for a thermal dual tip probe. FIG. 8I is
a schematic of a thermal tip having a resistive heater disposed on the tip.
 [0017]
FIG. 9A is a cross-sectional schematic of an electric field controlled dual tip probe in accordance
with an embodiment of the disclosed invention, showing a wire disposed within the hollow tip
down to a point in proximity to the aperture at the apex of the tip, and an evaporatable material
disposed within the hollow portion of the tip. FIG. 9B is a schematic of evaporation of a metal
from electric field controlled dual tip probe in accordance with an embodiment of the disclosed
invention as a result of an applied electric field between the tip and surface.
 [0018]
FIG. 10A is a schematic of a bending of a dual tip cantilever in connection with force-distance
calibration of a dual tip probe in accordance with an embodiment of the disclosed invention. FIG.
10B is graph showing the resulting force distance curve.
 [0019]
FIG. 11A-11C are graphs showing ANSYS modeling of the electric field and thermal gradients as
a function of distance between the tip and the substrate. FIG. 11D is a graph showing ANSYS
modeling of temperature diffusion from a tip.
 [0020]
FIGS. 12A and 12B are graphs illustrated the dependence of the thermal gradient on distance
between the tip and substrate. FIG. 12A illustrates the change of electric field intensity with
distance from the substrate under fixed voltage bias (U=20V). FIG. 12B illustrates the voltage
bias required to achieve an electric field intensity of 5×10 9 V/m for a given distance between the
tip and the substrate.
 [0021]
FIGS. 13A-13D are graphs showing the electric field gradient as a function of tip surface
topology.
 [0022]
FIGS. 14A-14D are force-distance curves illustrating how stiffness can affect the performance of
a dual tip probe in accordance with an embodiment of the invention.
 [0023]
FIG. 15A is a schematic illustrating the synthesis of a carbon nanotube (CNT) from a heated
tip. FIG. 15B is a schematic illustrating synthesis of a CNT from a tip using a heated substrate.
 [0024]
FIG. 16A is a schematic illustrating formation of nanostructures using probe assisted deliver of
chemical reagents to nanoreactors on a surface. FIG. 16B is an AFM height image of nanowells
formed by phase separation of immiscible polymers. FIG. 16C is an AFM height image of
nanowells formed using electron-beam lithography. FIG. 16D is an SEM image showing nanowell
formation by oxidation of anodic aluminum oxide.
 [0025]
FIG. 17 is a schematic generally illustrating the mold and transfer method for forming dual tip
probes in accordance with an embodiment of the disclosed invention.
 [0026]
FIG. 18 is a schematic illustrating a method of forming a dual tip probe in accordance with an
embodiment of the disclosed invention.
 [0027]
FIG. 19A is a schematic illustrating formation of a nanoparticle array using block copolymer
nanostructure templates in accordance with an embodiment of the disclosed invention. FIG.
19B is an AFM image of the block copolymer nanostructure template formed in accordance with
the method illustrated in FIG. 19A. FIG. 19C is SEM images of a nanoparticle array formed in
accordance with the method illustrated in FIG. 19A.
 [0028]
FIG. 20A is a schematic illustrating formation of a nanowire using block copolymer nanostructure
templates in accordance with an embodiment of the disclosed invention. FIG. 20B is an AFM
image of the block copolymer nanostructure template formed in accordance with the method
illustrated in FIG. 20A. FIG. 20C is an SEM image of a nanowire formed in accordance with the
method illustrated in FIG. 20A.
 [0029]
FIG. 21 is a schematic of the mold and transfer method for forming a dual tip probe in accordance
with the method described in Example 1.
 [0030]
FIG. 22 is an SEM of a dual tip probe illustrating a dual tip probe formed in accordance with the
method described in Example 1.
 [0031]
FIG. 23 is an SEM image of an array of silicon nitride dual tip probes formed in accordance with
the method described in Example 2.
 [0032]
FIGS. 24A and 24B are atomic force microscopy images comparing the imaging capability of a
dual tip probe in accordance with the disclosed invention (FIG. 24B) to a conventional single tip
(FIG. 24A).
 [0033]
FIG. 25A is an AFM height image of a 10 particle per line gold nanoparticle pattern formed using
an AFM tip. FIG. 25B is an AFM image of two gold hexagon shaped structures formed by pulsed
evaporation of gold from an AFM tip.
 [0034]
FIGS. 26A-26D are SEM images of Au patterned structures generated by applying a 20 V tip bias
voltage to a gold coated AFM tip. FIG. 26A is an SEM image of Au nanoparticles evaporated onto
a silicon dioxide surface. FIG. 26B is a low SE mode image showing the raised topology of a 3
line pattern of Au. FIGS. 26C and 26Dshow Au patterns formed by varying the pulse rates of the
applied voltage. FIG. 26C has 10 s−1 pulse rate and FIG. 26D has a 4 s−1 pulserate.
 [0035]
FIG. 27 is a schematic of an embodiment of hardware which can be used for scanning pulsed
evaporation of metal from a probe tip.
DETAILED DESCRIPTION
 [0036]
Scanning Probe Epitaxy (SPE) is the atom by atom growth of nanostructures from a surface
through the controlled delivery of chemical reagents to that surface under environmental control.
SPE can enable the tip-based synthesis of carbon nanotubes, semiconductor nanowires,
nanoparticles, quantum dots, and other printed indicia or patterns with control over the
architecture (e.g., length, diameter, and composition) of each nanostructure or pattern and control
over the orientation and spacing of the nanostructures on a surface. Tip-based synthesis
reactions can occur, for example, on a substrate where the tip delivers the chemical reagents to
the substrate. Alternatively, the reaction can occur at the tip surface where reagents in the gas
phase are delivered to the tip and a controlled pulse of energy can release the nanostructure from
the tip to a surface site or substrate of interest.
Dual Tip Probe Structure
 [0037]
Referring to FIGS. 1A and 1B and 2, a micro probe 10 having a dual tip architecture can include a
cantilever arm 12, a first tip 14 disposed on the cantilever arm 12, and a second tip 18 disposed
on the cantilever arm 12adjacent to the first tip 14. The apices of the first and second tips can be
formed on independent monolithic base structures connected to the cantilever arm, or the apices
of the first and second tips can be formed on a common monolithic base structure (see e.g., FIG.
7). The micro probe, also referred to herein as “the dual tip probe” can include, for example, a
non-synthesis tip, such as a reader tip, and a synthesis tip (e.g., one that creates features by
addition, subtraction, or alteration of material) as the first and second tips, as shown in FIG. 1A.
Alternatively, the dual tip can include two reader tips or two synthesis tips. Any other known or
suitable tip types can be included as one or both of the tips on the dual tip probe. The dual tip
probe can further include, for example, one or more stiffness ribs (as shown in FIG. 6B) disposed
on the cantilever arm, for example, between the first and second tips The dual tip probe can
further include one or more strain gauges (as shown in FIG. 1B) disposed on the cantilever arm.
 [0038]
The inclusion of both a reader tip and a synthesis tip on a single cantilever arm can allow for the
simultaneous or substantially simultaneous (a) measurement of the topology of a surface and (b)
synthesis of nano structures or printing of indicia on the substrate. This can allow for in situ
correction of an error. For example, the dual tip probe can be used to detect, with the first or
reader tip, an error in a printed indicia. The error can be, for example, an omission in the printed
indicia. The error can also be, for example, an additional printed feature, such as an extra printed
feature or an extraneous feature not introduced via printing (e.g., a flaw in the substrate, or other
extraneously introduced material). The second tip can then be used to correct the error either by
printing a correction indicia spatially corresponding to the printing omission or by removing the
additional feature. An extra printed feature can be removed, for example, by etching the extra
printed feature, for example, by depositing an etchant with the second tip onto the extra printed
feature. Where the printed indicia is a circuit, for example, the error can be a gap in the circuit.
For example, where the printed indicia is a circuit and the error is a gap in the circuit, the second
tip can be used to print or form a conducting nanowire in the gap to reconnect the circuit. The
error in a circuit can also be, for example, an extra conductive wire or dot that erroneously
couples portions of the circuit. This error can be corrected, for example, by depositing with the
second tip an etchant or other material to remove the extra wire or dot from the circuit pattern.
Any known tip based printing and removal methods can be used for correction of a detected error
in a printed indicia. For example, it is well known that certain metal salts (e.g., metal halides) are
more volatile than the parent metal. Thus, extra printed metal or extraneous metal can be
removed by depositing a suitable material that reacts with the metal to form a volatile species that
evaporates from the substrate.
 [0039]
The dual tip probe can also be used, for example, for in situ correction of a printed indicia while
printing the indicia using the second tip. A printed indicia can be printed using the second tip and
simultaneously or substantially simultaneously characterized using the first tip to detect errors in
the printed indicia. The writing of the printed indicia and the detection of errors using the
synthesis and reader tips, respectively, can also occur, for example, sequentially. Then, as
described above, the error can be corrected by reprinting with the second tip.
 [0040]
FIG. 3 illustrates in situ correction of a pattern using the dual tip probe. As illustrated in FIGS. 3A
and 3B, a first Au pattern was printed using the second tip. Characterization of the Au pattern
using the first tip detected error in the patterns, namely omission of printed Au between the
square patterns. As illustrated in FIG. 3C, this error was corrected using the second tip by printing
the additional Au patterns.
 [0041]
The dual tip structure can also allow for simultaneously working in both contact and non-contact
mode. For example, the synthesis tip can operate in a non-contact mode, while the reader tip can
be in contact with the surface and operate in contact mode. This can aid in preventing synthesis
tip wear due to contact with the substrate. This can be especially useful with sharp synthesis tips
that are more susceptible to wear when operated in contact mode. FIG. 4 shows a sharp tip both
before and after writing to illustrate the wear on the tip. Tip wear can diminish the resolution of the
tip.
 [0042]
Referring back to FIGS. 1 and 2, the cantilever arm can extend the entire length of the micro
probe and operatively couple the first and second tips. The cantilever arm can be formed, for
example, from silicon nitride, silicon oxide, or polysilicon. The cantilever arm can also be formed,
for example, from a metal, such as, silver, gold, aluminum, tungsten, and copper. The cantilever
arm can be designed to bend in response to an applied force. The cantilever arm can further
include electric leads for applying a bias and/or electrical pulses to an electric field controlled or
thermal synthesis tip. The electric leads can be, for example, printed onto the cantilever arm.
 [0043]
The cantilever arm has first and second ends, with the first and second tips disposed adjacent the
second end. The cantilever arm can have any suitable length, for example, in a range of 100 to
500 μm. Other suitable lengths include, for example, ranges of 100 μm to 400 μm, 150 μm to 350
μm, 100 μm to 300 μm, 200 μm to 500 μm, 200 μm to 400 μm, and 200 μm to 300 μm. The
length can be for example, about 100, 150, 200, 250, 300, 350, 400, 450, or 500 μm. The
cantilever arm can have any suitable thickness, for example, in a range of 1 μm to 100 μm. Other
suitable thickness include ranges of 10 μm to 80 μm, 20 μm to 60 μm, and 30 μm to 50 μm. The
thickness can be for example, about 1, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75,
80, 85, 90, 95, or 100 μm.
 [0044]
Referring to FIG. 2, the cantilever arm can have a constant width along the entire length of the
cantilever arm. Alternatively, the cantilever arm can have a varied width comprising two, three,
four or more discrete sections, or tapered sections or cantilever arm. For example, as shown
in FIGS. 5 and 6A, the cantilever arm 26 can have two sections 20 and 22 disposed between the
first and second ends 24 and 28, respectively. The first and second tips 30 and 32, respectively,
can be disposed, for example, on the second section 22. The first section 20of the cantilever
arm 26 can have a width W1 and a length L1, and the second section 22 of the cantilever arm
can 26 have a width W2 and a length L2. The length L1, L2 and width W1, W2 of the first and
second sections 20and 22 can be different. For example, the length L1 and width W1 of the first
section 20 can be greater than the length L2 and width W2 of the second section 22. The length
L1 of the first section 20 can be, for example, in a range of 50 μm to 400 μm. Other suitable
lengths L1 include, for example, 50 μm to 350 μm, 75 μm to 200 μm, and 100 μm to 300 μm.
Length L1 can be, for example, about 50, 60, 70, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170,
180, 190, 200, 210, 220, 230, 240, 250, 260, 270, 280, 290, 300, 310, 320, 330, 340, 350, 360,
370, 380, 390, or 400 μm. The length L2 of the second section 22 can be, for example, in a range
of 25 μm to 100 μm. Other suitable lengths L2 include, for example, 30 μm to 70 μm, 40 μm to 60
μm, and 25 μm to 50 μm. Length L2, can be, for example, about 25, 30, 35, 40, 45, 50, 55, 60,
65, 70, 75, 80, 85, 90, 95, or 100 μm. The widths W1, W2 of the first and second
sections 20 and 22 can be, for example, in a range of 5 μm to 100 μm. Other suitable widths
include, for example, 10 μm to 90 μm, 20 μm to 80 μm, 30 μm to 70 μm, and 40 μm to 60 μm.
One or both of the widths, W1, W2, can be, for example, about 5, 10, 15, 20, 25, 35, 40, 45, 50,
55, 60, 65, 70, 75, 80, 85, 90, 95, or 100 μm. A micro probe having a cantilever arm 26 having a
second section with a narrower width W2 than the width W1 of a first section can have increased
mechanical stiffness between first and second tips as compared to a probe having a uniform
width along the entire cantilever. Variation of the width of sections of the cantilever arm can be
used as means for varying the stiffness of the probe in selected regions, and finer control over the
bending of the cantilever arm after contact of the first tip with the surface end during increased
applied force to bring the second tip in closer proximity to the surface. Variation of the other
dimensions of the cantilever arm include, for example, total cantilever arm length, first and
second section length L1 and L2, cantilever arm thickness and cantilever arm materials can also
affect stiffness and bending performance. Thus, these variables can be modified to achieve the
desired force-distance relationship for a dual tip probe.
 [0045]
With reference to FIG. 5, for example, the first tip 30 is disposed adjacent to the second end 28 of
the cantilever arm 26. The first tip 30 can be designed, for example, as a non-synthesis or reader
tip. For example, the first tip can be a surface topology measurement tip. The reader tip can
operate, for example, in contact mode to measure and characterize the topology of a surface of a
substrate and/or structures printed on the substrate. The first tip 30 can be brought into contact
with a substrate, for example, by applying a force on the cantilever arm 26to bend the cantilever
arm and displace the first tip 30 towards the surface. The first tip can be formed, for example,
from a dielectric materials such as, for example, silicon, silicon nitride, or silicon dioxide. The first
tip can also be formed, for example, from a metal, such as Au, Ag, Cu, and W. The first tip can
have a thickness in a range of 1 μm to 20 μm, for example. Other suitable thickness included for
example, about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, and 20 μm. The first
tip can have, for example, a pyramidal shape with a base disposed on the cantilever arm and an
apex terminating in a point, and thus the thickness is measured as the distance from the base of
the tip disposed on the cantilever arm to the apex. The point can be, for example, a sharp point or
a rounded, e.g. spherical point. The tip can also have a cylindrical shape, wherein the major axis
of the cylinder projects out from the cantilever arm, e.g. in perpendicular fashion. The first tip can
be any known or suitable reader type tip having any other suitable geometry. For example, the
first tip can be an atomic force microscopy tip or a scanning microscopy tip. The first tip can be for
example a solid tip or a hollow tip, and preferably is a solid tip.
 [0046]
The second tip 32 is disposed on the cantilever arm 26 adjacent to the first tip 30. The second
tip 32 can be adjacent the first tip 30 and in line with the major axis of the cantilever arm 26. In
the alternative, the second tip 32 can be offset from the first tip 30 toward the first end 24 of the
cantilever arm 26 by a distance x, as illustrated in FIG. 5. The distance x can be in a range for
example, of 1 μm to 50 μm. Other suitable distances x include, for example, 1 μm to 40 μm, 5 μm
to 30 μm, and 10 μm to 20 μm. The distance x can be, for example, about 1, 2, 3, 4, 5, 6, 7, 8, 9,
10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,
36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, or 50 μm.
 [0047]
The second tip can be designed as a synthesis tip for additive fabrication, such as synthesizing
nanostructures and/or printing (e.g. depositing) indicia or patterns. The second tip can also be
designed as a tip for subtractive fabrication to remove features from a substrate. For example, the
second tip can be used to deposit an etchant to remove a portion or a feature of a substrate. The
tip can also be used, for example, to deposit a material that reacts with a metal feature on a
substrate that reacts with the metal to form a volatile species, such as a metal salt, for example, a
metal halide, that evaporates from the substrate. As used herein, “synthesis tip” refers to a tip
with either additive fabrication capabilities (forming structures onto a substrate), subtractive
fabrication capabilities (i.e. removing structures from a substrate), or alteration capabilities (e.g.,
reaction, phase change, magnetic properties).
 [0048]
The second tip can operate in either contact mode, in which the second tip is in contact with a
substrate, or preferably in non-contact mode, in which the second tip is disposed above the
substrate. The second tip can be formed, for example, from a conductive material such as, for
example, doped polysilicon, doped diamond, a metal, a hard metal, or a metal oxide. Suitable
metals include Au, Al, Ni, Fe, Pt, Os, Ru, Jr, In, W, Ag, and Cr. Suitable hard metals include TiN,
TiC, WC, and TaN. Suitable metal oxides can be, for example, InO and IrO2. The second tip can
have a thickness in a range of 1 μm to 20 μm for example. Other suitable thickness include, for
example, about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, and 20 μm. The first
and second tips can have the same thickness or can have different thicknesses. For example, the
first tip can be thicker than the second tip (see, for example, FIG. 1 a). Referring to FIGS. 2 and
4, the tips, and preferably the second tip, can have a pyramidal shape with a base disposed on
the cantilever arm and an apex terminating at a point. The point can be, for example, a sharp
point or a rounded spherical point. The second tip can also have, for example, a cylindrical or
spherical shape. The second tip can have a rough or a smooth surface. For example, a second
tip can have a pyramidal shape with a rough surface having grain boundaries. The grain
boundaries can be in a range of 100 nm to 50 μm. Other suitable ranges include 200 nm to 40
μm, 300 nm to 30 μm, 400 nm to 20 μm, 500 nm to 10 μm, 600 nm to 5 μm, 700 nm to 1 μm, 100
nm to 500 nm, 150 nm to 400 nm, 200 nm to 300 nm, 1 μm to 50 μm to 40 μm, and 10 μm to 30
μm. The grain boundaries can be, for example, 100 nm, 200 nm, 300 nm, 400 nm, 500 nm, 600
nm, 700 nm, 800 nm, 900 nm, 1 μm, 5 μm, 10 μm, 15 μm, 20 μm, 25 μm, 30 μm, 35 μm, 40 μm,
45 μm, or 50 μm.
 [0049]
Referring to FIG. 7, the second tip can also be formed to have an aperture at the apex for delivery
of ink material through the tip. The second tip can be a hollow tip (partially or fully hollow) or a
solid tip. For example, as shown in FIG. 7, the second tip can include a channel for delivery of an
ink material through the channel.
 [0050]
In addition to the second tip architectures described above, the synthesis tip can be any known or
suitable synthesis or writing tip, such as those used with dip pen nanolithography and atomic
force microscopy. For example, the synthesis tip can be a high temperature tip (as illustrated
in FIG. 8), an electric field controlled tip (as illustrated in FIG. 9), a catalyst tipped tip, an aperture
evaporator tip (as illustrated in FIG. 7), an atomic force microscopy tip, an elastomeric gel tip, or a
polymer tip. The synthesis tip can also be used, for example, for dip pen nanolithography (DPN).
To load the synthesis tip with ink for DPN, a nanowell can be provided having the ink therein. The
nanowell can further include a resting surface for the non-synthesis tip, such that the non-
synthesis tip is not loaded with ink, while the synthesis tip remains aligned with the nanowell. The
non-synthesis tip can be placed on the resting surface and a force can be applied to the
cantilever arm to bend the cantilever arm and displace the synthesis tip into the nanowell, thereby
loading it with ink.
 [0051]
Alternatively, a substrate mold created during the mold and transfer process for fabricating the
dual tip probes (as is described in detail herein) can be used as ink wells for loading the dual tip
probe for writing, such as for DPN. As a result of the fabrication process, the openings of the
substrate mold are inherently aligned with the first and second tips (e.g., in relative location,
shape, dimension, etc.). This relationship can be particularly advantageous when loading an
array of dual tip probes formed from a substrate mold. One of both of the openings can be filled
with an ink, depending whether one or both of the first and second tips are to be loaded with ink.
The first and second tips can be aligned with the openings and then the probe can be lowered
toward the substrate mold such that the first and second tips are at least partially disposed within
the first and second openings to contact an ink contained within at least one of the openings. This
method can be particularly useful for loading only a single tip of the dual tip probe with the ink, as
the mold inherently separates the tips into the two openings, isolating the non-synthesis tip from
the ink.
 [0052]
Referring to FIG. 6B, the dual tip probe can further include stiffness-enhancing ribs disposed on
the cantilever arm, for example between the first and second tips. The ribs preferably comprise
an additional layer of cantilever arm material or other stiffness-enhancing material disposed on
the cantilever arm, for example in linear fashion. The ribs preferably are disposed on the same
side of the cantilever arm from which the tips project, and in the alternative or in addition can be
disposed on the opposite side of the cantilever arm. The ribs can be oriented between the first
and second tips parallel with respect to the major axis of the cantilever arm, perpendicular with
respect to the major axis of the cantilever arm, or any range in between, preferably parallel with
respect to the major axis of the cantilever arm which has a length greater than its width (see FIG.
5). The ribs can have a length, for example, in a range of 1 μm to 10 μm. Other suitable lengths
include, for example, about 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 μm. The ribs can have a width, for
example, in a range of alum to 5 μm, and a thickness, for example, of 0.1 μm to 5 μm. Other
suitable widths include, for example, about 0.1, 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, and 5 μm. Other
suitable thickness include, for example, about 0.1, 0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, and 5 μm. The
ribs can have any suitable cross-sectional shape, including, for example, linear, circular,
rectangular, triangular, and semi-circular.
 [0053]
The distance between the apex of the second tip and the substrate can be controlled using the
dual tip design. For example, the relationship between the applied force on the cantilever arm and
the distance between the second tip and the substrate of a dual tip probe can be calibrated using
a force-distance curve as shown in FIGS. 10A and 10B to determine the relationship between the
amount of applied force and the distance between the second tip and the substrate. A dual tip
probe can be calibrated by bringing the probe towards the surface of the substrate until the first
tip is in contact with the substrate, applying additional force to thereby bend the cantilever arm
between the tips and optionally between the second tip and base of the cantilever arm to displace
the second tip closer to the substrate. The amount of cantilever bending can be sensed, for
example, from a laser deflection from the cantilever by means known in the art. The amount of
cantilever bending can also be sensed using one or more strain gauges disposed on the
cantilever arm, for example as illustrated in FIG. 1B. Referring to FIG. 1C, the change in distance
between the second tip and the substrate can be calculated based on the amount of bending of
the cantilever (Δz) and the dimensions of the dual tip probe, using the following relationship:
 [0000]
Δ h2=Δz (1-L2L1)
 [0000]
wherein Δh2, Δz, L1, and L2 are defined as illustrated in FIG. 1C.
 [0054]
With reference to FIGS. 10A and 10B, the knee in the approach curve followed by a sharp rise in
force is indicative of contact. The sharp dip is associated with attractive capillary forces that
cause the tip to snap into contact with the substrate, and the rise in force indicates the increased
force necessary to push the second tip towards the substrate after contact has been made by the
first tip. Generally, the flat portion of the curve indicates that the neither tip has contacted the
substrate. The relationship between applied force and the distance between the second tip and
the substrate can be determined from the force-distance curves using the amount of cantilever
bending and the difference in thickness between the first and second tips.
 [0055]
Once the force-distance relationship is known, the distance or gap between the second tip and
the substrate can be modulated, e.g. during a writing process, by varying the applied force. The
distance between the second tip and the substrate can affect the writing method and can be used
to vary dimensions of the nanostructure synthesized. For example, when a second tip is a
thermal tip or an electric field controlled tip, the distance between the second tip and the
substrate can affect the thermal or electric field gradient between the substrate and the second
tip. Changes in the gradient can be, for example, used to alter dimensions of the synthesized
nano structures.
 [0056]
Referring to FIG. 11, the behavior of the thermal or electric field gradient can be modeled using,
for example, ANSYS finite element analysis and computational fluid dynamics software known in
the art. FIGS. 11A-11Cillustrate a simulation of localized electric field around a tip as a function of
the distance between the tip and the substrate. As illustrated in FIGS. 11B and 11C, when the
gap is less than 100 nm, the electrical field is generally focused inside a 200 nm area. As
illustrated in FIG. 11A, when the gap is about 200 nm, the field begins to diffuse. Preferably, the
gap is less than about 100 nm in order to deposit materials and make nanometer dimensioned
structures on the substrate. FIG. 11D illustrates temperature diffusion from the tip. A high
temperature zone (about 600° C.) is localized around the tip. The temperature drops more rapidly
with increasing distance away from the tip. Referring to FIG. 12A, the change of electric field
intensity with distance from the substrate (i.e. gap) under a fixed voltage bias (20V) is
illustrated. FIG. 12B illustrates the voltage bias required to obtain an electric field intensity of
5×109 V/m for a given gap size.
 [0057]
Referring to FIGS. 13A-13D, the gradient can also be varied, for example, by changing the tip
shape. FIG. 13illustrates the effect of various tip surfaces, including a nanorod tip having a
smooth surface and a 50 nm radius, a pyramid tip having a smooth surface spherical end having
a 50 nm radius, a nanorod tip having a rough surface with 10 nm grain diameter, and a pyramid
tip having a spherical end with a rough surface and a 10 nm grain diameter. The graphs of FIGS.
13A-13D were generated by contacting each of the tips with a 7 nm thick SiO 2layer and applying
a voltage bias of 16V. As shown in FIGS. 13A-13D, the rough surface generally resulted in
increased electric field intensities with the rough nanorod tip having the highest electric field
intensity. Other parameter modifications such as the amount of the applied field and/or the
temperature can be used to modify the gradient.
 [0058]
The dual tip probe can further include means for adjusting the stiffness of the probe. Referring
to FIGS. 14A-14Dand as described elsewhere herein, adjustment of the stiffness can change the
force-distance relationship of the probe. The stiffness of the probe can be adjusted, for example,
by varying the dimensions of the probe, including the lengths L1 and L2 and widths W1 and
W2 of the cantilever arm, the thickness of the cantilever arm, the thickness of the first and second
tips, the distance between tips, the materials of construction, and the inclusion of one or more ribs
disposed on the cantilever arm, for example between the first and second tip.
 [0059]
Referring to FIGS. 14A and 14B, variation of the width of the first section of the cantilever arm
can result in different force-distance behavior. FIGS. 14A and 14C show in each figure the force-
distance curve during approach (application of force) as the top line, and during retraction as the
bottom line. FIG. 14B shows the force-distance curve during approach (application of force) as
the bottom line, and during retraction as the top line. In FIG. 14D, the approach and retraction
curves substantially overlap, except that the retract curve shows a sharp decrease in relative
force at a relative distance of about −4 μm. The dimension of the designs tested (Design 110and
Design 130) in FIGS. 14A and 14B are included in Table 1 of Example 2. Generally, the width of
the first section of Design 110 was about 20 μm smaller than the width of the first section of
Design 130. As shown in FIG. 14A, for Design 110, it can be difficult to differentiate on a force
distance curve distinct points when the reading and the synthesis tips contact the substrate. For
initial contact of Design 110 between relative distances −0.75 and −1.00 μm, the stiffness was
about −19.8 nN/μm, while further contact between relative distances −1.28 and −1.87 μm, the
stiffness was about −67.4 nN/μm. As shown in FIG. 14B, for Design 130, variation of the stiffness
of the probe can make it easier to detect distinct points of contact for the first and second tip.
In FIG. 14B, when the first tip made contact, the stiffness was about 10.8 nN/μm, and when the
second tip made contact the stiffness increased by more than two fold to −28.1 nN/μm. From this
curve, it can e determined that there is a 200 nm vertical distance between where the first tip
makes contact and the second tip makes contact with the substrate. As shown in FIGS. 14C and
14D, the addition of ribs can increase the stiffness value. In fact, the addition of two ribs can
increase the stiffness value of a probe by an order of magnitude. With such stiff tips it can be
difficult to discern distinct contact points for the first and second tips.
Typical Synthesis Conditions for Nanostructures
 [0060]
The dual tip probe can be designed to synthesize a variety of nanostructures and patterns.
Synthesis of various nanostructures using the dual tip probe can be done using synthesis
conditions as are well-known in the art. For example, quantum dots, such as CdS and CdSe
quantum dots are typically synthesized at a temperature in excess of about 200° C. using, for
example, organometallic precursors in, for example, an inert atmosphere. Quantum dots can also
be synthesized using, for example, an ambient atmosphere. Carbon nanotubes are typically
synthesized at a temperature in excess of about 550° C., using for example, catalytic
nanoparticles. The catalytic nanoparticles can include, for example, Fe, Ni, and Co nanoparticles.
The carbon nanotubes can be synthesized in a hydrocarbon environment, such as, for example,
CH4, C2H2, or C2H5OH, using a carrier gas, such as, for example, Ar. Silicon semiconducting
nanowires are typically synthesized at a temperature in excess of 400° C., using a catalytic
nanoparticle, such as, for example, Au. The silicon semiconducting nanowires can be
synthesized in a SiH4 and H2 environment. InP semiconducting nanowires are typically
synthesized at a temperature in a range of 240° C. to 300° C. Catalytic nanoparticles, such as Bi,
can be used for synthesis of the nanowires, in an environment, for example, of polydecene
solutions of In(myristate) and P(SiMe3). The synthesis tip of the dual tip probe can be adapted to
use the above-described processing conditions for the formation of various nanostructures.
Evaporator Synthesis Tip
 [0061]
Referring to FIG. 9, the synthesis tip can be a variety of synthesis-type tips, including, for
example, an evaporator synthesis tip with electric field control or thermal capabilities. The thermal
and/or electric field evaporator tip can be used to form a variety of nanostructures, including, for
example, nanoparticles, nanowires, nanodiscs, quantum dots, nanotubes, nanopatterns, and
combinations thereof. The dimensions of the nanostructure can be adjusted, for example, by
varying the applied voltage and the pulse width, while placement of the nanostructure is governed
by tip movement.
 [0062]
The evaporator synthesis tip can be used, for example, for direct metal deposition onto a
substrate. Direct metal deposition can be useful for formation of nanowires and carbon nanotube
catalysis, plasmonic structures, and in circuitry repair. Field-induced deposition from an electric
field controlled synthesis tip enables control of the feature size by varying pulse width and pulse
bias voltage. Metal evaporation can occur, for example, under negative bias, with voltages, for
example, in a range of −8 V to −100 V. Other suitable voltages include, for example, −10 V to −90
V, −20 V to −80 V, −30 V to −70 V, and −40 V to −60V, The voltage can be for example, about
−8, −9, −10, −15, −20, −25, −30, −35, −40, −45, −50, −55, −60, −65, −70, −75, −80, −85, −90,
−95, and −100 V.
 [0063]
The evaporator tip can also be used, for example, to synthesize semiconducting nanowires and
carbon nanotubes (CNT). Metallic precursor for nanowire and CNT growth can be delivered to a
surface using field induced evaporation from the evaporator tip. The as-deposited precursor can
then be exposed to a gaseous environment and heated to induce growth the nanowires and/or
carbon nanotubes.
 [0064]
For example, gold nanoparticles can be used as a catalyst for epitaxial growth of semiconducting
silicon nanowires. Gold nanoparticles can be deposited onto the synthesis tip and then
transferred to the substrate using field induced evaporation. A Cr layer can be first deposited onto
the synthesis tip as an adhesion layer. The Cr layer can have a thickness, for example, in a range
of 5 nm to 50 nm. Other suitable thickness include, for example from 10 nm to 40 nm, 15 nm to
35 nm, 20 nm to 30 nm. The Cr layer thickness can be, for example, about 5, 6, 7, 8, 9, 10, 15,
20, 25, 30, 35, 40, 45, or 50 nm. The gold layer can have a thickness, for example, in a range of
50 nm to 500 nm. Other suitable thicknesses include, for example, from 60 nm to 400 nm, 70 nm
to 300 nm, 80 nm to 200 nm, and 100 nm to 200 nm. The gold layer can have a thickness, for
example, of about 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 110, 120, 130, 140, 150, 160, 170,
180, 190, 200, 210, 220, 230, 240, 250, 260, 270, 280, 290, 300, 310, 320, 330, 340, 350, 360,
370, 380, 390, 400, 410, 420, 430, 440, 450, 460, 470, 480, 490, or 500 nm. The gold
nanoparticles can be deposited onto the substrate using the dual tip probe. Gold nanoparticles
can be deposited, for example, by applying a bias to the tip. The bias can be in a range, for
example, of 8 V to 100 V. Other suitable voltages include, for example, 10 V to 90 V, 20 V to 80
V, 30 V to 70 V, and 40 V to 60V, The voltage can be for example, about 8, 9, 10, 15, 20, 25, 30,
35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, and 100 V. The bias can also be applied in short
pulses to the tip. For example, pulses in a range of 1 to 100 ms can be used. Other suitable pulse
times include, for example, 5 ms to 80 ms, 10 ms 70 ms, 20 ms to 60 ms, and 30 ms to 50 ms.
The pulse time can be for example, about 1, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70,
75, 80, 85, 90, 95, or 100 ms. The pulses can be controlled using, for example LABVIEW
software (National Instruments, Austin, Tex.) that operates a pulse generator through a GPIB
interface. Upon introduction of a gas, such as silane gas, for example, formation of the nanowires
can proceed through the vapor-liquid-solid mechanism. The diameter of the nanowire can depend
upon the size of the gold nanoparticle precursor.
 [0065]
Referring to FIG. 7, the evaporator tip can further include an aperture formed in the writing portion
of the tip (e.g., at the apex of the tip), through which the writing material can be deposited. The tip
can be, for example, a hollow tip and the writing material can be contained within the tip and can
be caused to exit the tip upon application of an applied voltage. The rate of deposition can
depend upon the size of the orifice and the magnitude and duration of the applied field.
 [0066]
The dual tip design allows for the synthesis through pulsed evaporation in a non-contact mode.
The first tip can operate in contact mode to provide in situ characterization of the surface and/or
the structures formed, while the evaporator synthesis tip (i.e. the second tip) remains disposed
above the substrate. As one advantage, this architecture decreases or avoids the consumption,
wear, and change in morphology or dimension of the synthesis tip, which can have one or more
benefits such as improving resolution, feature size control, and reproducibility. This architecture
can also allow for the extension of pulsed scanning evaporation deposition of non-conducting
precursors that have lower vapor pressure than that of the tip metal, such as, for example,
stoichiometric solid precursors including bulk CdSe and CdS solids, and decomposable
precursors including CoCl2, FeCl2, and NiCl2.
Thermal Catalyst-Tipped Synthesis Tip
 [0067]
Referring to FIG. 8, the synthesis tip can be, for example, a heated tip. The heated tip can further
include a catalyst, such as, for example, catalytic nanoparticles, disposed on the tip. The catalytic
nanoparticles can include, for example, Fe, Ni, Co, Au, and Bi nanoparticles. The tip can be
heated, for example, by including a resistive heater on the second tip. Thermal tips can heat to
temperatures, for example, in a range of 100° C. to 700° C. Other suitable temperatures include,
for example, 150° C. to 600° C., 200° C. to 500° C., and 300° C. to 400° C. The temperature can
be, for example, 100° C., 150° C., 200° C., 250° C., 300° C., 350° C., 400° C., 450° C., 500° C.,
550° C., 600° C., 650° C., or 700° C. As shown in FIG. 8D, the highest temperatures are localized
on and around the tip. FIGS. 8D, 8E, and 8G further illustrate that the increased temperature is
localized around the tip, with the temperature decreasing more rapidly with increasing distance
from the tip. It can be desirable that the tips be able to heat up quickly. For example, it can be
desirable to have tips that reach in excess of about 300° C. in about 15 seconds. The thermal tip
can behave as a nanoscale evaporator that deposits nanoscale quantities of metal in a similar
fashion as a macroscopic thermal metal evaporator. The thermal tips can also behave as a
chemical vapor deposition (CVD) system whereby nanowires and carbon nanotubes are grown
directly from the tip.
 [0068]
The resistive heater can be formed by patterning metal wires onto the second probe, which can
be for example a silicon nitride probe. The metal wires can include, for example, a primary
conductor, such as Au wires, a diffusion barrier, such Pt wires, and an adhesive, such as Cr. The
resistive heater can be wire bonded onto the tip. The temperature of the tip can be determined
using the following relationship by applying a bias across the restive heater:
 [0000]
R(T)=R 0(1+αT)
 [0069]
wherein R(T) is the resistant at temperature T, R0 is the resistance at a reference temperature

(i.e. room temperature), and α is the temperature coefficient of the resistance. The resistance
increases as the applied power increases. Preferably, the tips can be heated to a temperature in
a range of 100° C. to 700° C.
 [0070]
Referring to FIG. 15, a heated catalyst-tipped tip can be used, for example, for formation of
CNTs. An increase in temperature of the catalyst-tipped synthesis tip can be used to initiate
growth of the CNT. The temperate can be increased to be in a range of 200° C. to 700° C. Other
suitable temperatures include, for example, 200° C. to 600° C., 250° C. to 500° C., and 300° C. to
400° C. The temperature can be, for example, 200° C., 250° C., 300° C., 350° C., 400° C., 450°
C., 500° C., 550° C., 600° C., 650° C., or 700° C. A subsequent temperature drop at the tip can
be used to terminate growth. Thus, the length of a CNT can be controlled by controlling the
temperature of the tip. Further, the localized temperature control at the tip will minimize
competing thermal decomposition of the catalyst and outgassing of the system. As illustrated
in FIG. 11D, the temperature can substantially drop at increasing distances away from the tip,
which can allow for localization of the CNT growth at the tip. Thus, CNT growth can be localized
to the tip even when other portions of the probe are coated with the catalyst. The position of the
CNT on the substrate can be controlled by the movement of the tip on the substrate. CNT growth
can also be initiated by a catalyst-tipped tip without heat control by heating the substrate to a
temperature in a range of 200° C. to 700° C. Other suitable temperatures include, for example,
200° C. to 600° C., 250° C. to 500° C., and 300° C. to 400° C. The temperature can be, for
example, 200° C., 250° C., 300° C., 350° C., 400° C., 450° C., 500° C., 550° C., 600° C., 650° C.,
or 700° C.
 [0000]
Nanostructures formed in Nanoreactor Wells
 [0071]
Referring to FIG. 16, a dual tip probe having any of a variety of synthesis type tips as discussed
above can be used to form nanostructures in a nanoreactor well. Nanoreactor wells can be
synthesized using a variety of known techniques, such as self assembly, for example via phase
separation, of copolymers, electro-beam lithography, and anodically oxidized aluminum
templates, which results in a series of different nanoreactors, each with unique, tailorable
properties. Referring to FIG. 16B-16D, nanowells were formed by phase separation of immiscible
polymers, electron-beam lithography, and oxidation of anodic aluminum oxide, respectively. A
probe, such as the dual tip probe, can then be used to deposit a material into the nanoreactor
wells to form, for example, quantum dots, nanowires, and carbon nanotubes.
Method of Making The Dual Tip Probe
 [0072]
Referring to FIGS. 17 and 18, the probes having a dual tip structure can be formed using, for
example, a mold and transfer process. The mold and transfer process utilizes a substrate as a
template for probe formation. The substrate can be, for example, a silicon wafer. The substrate
can have a thickness in a range of 50 to 1000 μm. Other suitable thickness include from 60 μm to
900 μm, 80 μm to 800 μm, 100 μm to 600 μm, 200 μm to 500 μm, and 300 μm to 400 μm. The
substrate can have a thickness, for example, of about 50, 100, 150, 200, 250, 300, 350, 400, 450,
500, 550, 600, 650, 700, 750, 800, 850, 900, 950, or 1000 μm. The substrate can be precleaned,
for example, by rinsing with a solvent, preferably an organic solvent (e.g., acetone, methanol,
isopropyl alcohol, or any combination thereof).
 [0073]
First and second cavities are formed in the substrate. The cavities can be formed, for example, by
anisotropically etching the substrate. The substrate can be etched using a mask patterned with
two openings defining the first and second cavities. The mask can be formed, for example, by
depositing a mask layer onto the substrate and patterning the mask to form the two openings.
The mask layer can be, for example, a silicon oxide layer. The mask layer can have a thickness
in a range of 1000 Å to 10000 Å. Other suitable thicknesses include, for example 1100 Å to 9000
Å, 1200 Å to 8000 Å, 1400 Å to 7000 Å, 1600 Å to 6000 Å, 1800 Å to 8000 Å, 2000 Å to 6000 Å,
and 4000 Å to 5000 Å. The mask layer can have a thickness for example, of about 1000, 1500,
2000, 2500, 3000, 3500, 4000, 4500, 5000, 5500, 6000, 6500, 7000, 7500, 8000, 8500, 9000,
9500, and 10000 Å. The mask layer can be, for example, thermally grown on the substrate. For
example, a 5000 Å silicon oxide layer can be thermally grown on a silicon wafer at 1110° C. for
about 13 hours to form the mask layer. The openings can be formed, for example, using electron-
beam lithography or can be etched using, for example, wet chemical etching. The wet chemical
etching can use HF as the etching solution. The size of the openings can be correlated to the
depth of the subsequently formed cavities. For example, if the first opening is formed larger than
the second opening, a first cavity that is deeper than the second cavity will be subsequently
formed. This can be further correlated to the thickness (height, as measured from the base
connected to the cantilever arm) of the subsequently formed tips. Accordingly, the thickness
difference between the first and second tips can be controlled by controlling dimensions of the
openings formed in the mask.
 [0074]
The cavities can then be etched into the substrate, for example, by a wet chemical etching
process, using, for example KOH. The cavities have a shape corresponding to the desired tip
shape. For example, the cavities can have a pyramidal shape where it is desired to form
pyramidal tips. The mask can then be removed, using for example, BOE. The cavities can also be
formed, for example, by first patterning square openings onto a substrate, for example an
oxidized <100> silicon wafer, and then immersing the substrate can then be immersed in an etch
solution, such as KOH, to anisotropically etch pyramidal pits into the substrate. Etching is
generally terminated at <111> for a <100> silicon wafer, which can prevent over-etching of the
substrate.
 [0075]
A sacrificial layer can then be formed on the substrate including the cavities. The sacrificial layer
can be for formed, for example, by oxidation, chemical vapor deposition, low pressure chemical
vapor deposition, or physical vapor deposition. The sacrificial layer can be formed, for example,
from metals such as, copper, permalloy, tungsten, titanium, aluminum, silver, gold, oxides, such
as silicon oxide, silicon dioxide, silicon oxynitride, and zinc oxide, nitrides, such as silicon nitride
and titanium nitride, polymers, such as poly(dimethylsiloxane) (PDMS), polyimide, parylene,
elastomers, such as silicone and rubber, and photoresists such as SU-8. The sacrificial layer can
have a thickness, for example, in a range of 500 Å to 5000 Å. Other suitable thicknesses include,
for example, 600 Å to 4000 Å, 700 Å to 3000 Å, 800 Å to 2000 Å, and 900 Å to 1000 Å. The
sacrificial layer can have a thickness, for example, of about 500, 600, 700, 800, 900, 1000, 1500,
2000, 2500, 3000, 3500, 4000, 4500, or 5000 Å.
 [0076]
A first tip layer for forming the first tip can then be deposited onto the sacrificial layer. The first tip
layer is then patterned to remove at least a portion of the first tip layer disposed outside the first
cavity. The first tip layer can also be patterned such that only a portion of the first tip layer
disposed in the first cavity remains to form the first tip. The first tip layer can, for example, be
photolithographically patterned and chemically etched to form the first tip. If the first tip is an
imaging tip, then the first tip layer preferably includes, for example, a dielectric layer, such as a
silicon nitride layer or a silicon dioxide layer, which can electrically isolate the tip from the
substrate during imaging. Other suitable materials for imaging tips are known in the art and are
described elsewhere herein. The first tip layer can be etched, for example, using a plasma
etching.
 [0077]
A second tip layer for forming the second tip can be deposited onto the sacrificial layer. The
second tip layer can be deposited, for example, by low pressure chemical vapor deposition The
second tip layer can then be patterned to remove at least a portion of the second tip layer
disposed outside the second cavity. The second tip layer can also be patterned such that only a
portion of the second tip layer disposed in the second cavity remains to form the second tip. The
second tip layer can, for example, be photolithographically patterned and chemically etched to
form the second tip. If the second tip is a writing tip, then the second tip layer can be, for
example, a conductor, such as a doped polysilicon layer or a metal layer, such as gold or
aluminum. Other suitable materials for the writing and synthesis tips are known in the art and are
described elsewhere herein.
 [0078]
A cantilever arm layer can be deposited on the sacrificial layer including the first and second
layers. The cantilever arm layer can be optionally patterned to remove at least a portion of the
cantilever arm layer disposed on the patterned first and second tip layers. Alternatively, the first
and/or second tip layers can be patterned such that one or both of the first and/or second tip
layers extend outside of the first and/or second cavities to form the cantilever arm. The tip layers
and the cantilever arm layer can be deposited so that at least a portion of the layers overlap. A
handle wafer can then be attached to the cantilever arm layer. The sacrificial layer can then be
selectively etched to remove the dual tip probe from the substrate.
 [0079]
Where a thermal or electric field controlled dual tip probe is desired, a conductive material can be
deposited in or around the second tip. For example, an electrical biasing layer, an electrical
insulating layer, and an electrical conductor or heater can be sequentially deposited on top of the
second tip layer. One or more of the electrical biasing layer, the electrical insulating layer, and the
electrical conductor layer can optionally be formed to extend within at least a portion of the
second tip opening. These layers can be used, for example, to form the cantilever arm. A handle
can be attached to the cantilever arm and wires can be attached to the handle, for example, to
provide an electrical feed through. A thermal tip can also be formed by attaching a resistive
heater as described above to the one or both of the first and second tips, using for example, a
wire bonding method.
 [0080]
Referring to FIG. 18, the method can further include forming an aperture in the second tip in order
to form the second tip as an aperture evaporation tip. The size of the aperture can be controlled
to control the dimensions of the evaporated material deposited onto a substrate. The aperture
can have a width in a range of 5 nm to 200 nm. Other suitable widths include, for example, 10 nm
to 150 nm, 20 nm to 100 nm, and 40 nm to 50 nm. The aperture can have a width for example, of
about 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 110, 120, 130,
140, 150, 160, 170, 180, 190, or 200 nm. The aperture can be formed, for example, using
focused ion beam (FIB) etching or electrochemical plating. FIB etching can be used to etch
patterns with a line resolution in a range of 30 nm to 50 nm. See Wang et al., 87 Appl. Phys. Lett.
054102 (2005). The size of the aperture can be further reduced (i.e., the tip built up around the
aperture), if necessary, using, for example, selective thermal oxidation or material deposition.
 [0081]
The mold and transfer method can provide one or more advantages, including, for example, tip
uniformity within an array of tips, a variety of tip and cantilever material combinations can be
used, various tip and cantilever materials can be integrated into the same array, providing
multiplexed functions, uniform tip sharpness, a master substrate that can be reused thereby
reducing cost of fabrication over time, a master substrate that can be used as ink wells for loading
one or more tips with an ink or other material, and high fabrication yield and uniformity. The
shape and dimensions of the tip can be controlled and pre-determined in the mold and transfer
process by photolithography and self-limiting etching process. The sharpness of the tip can be
controlled and determined by the etched cavities in the substrate and the subsequently deposited
tip layers. The conventional mold and transfer process has been used to realize a million pen
probe array (single tipped probes) with 100% yield.
Software Platform for Pulsed Evaporation Synthesis

 [0082]
A manual application of voltage pulses to the evaporator synthesis tip and human-assisted
repetition of experiments can introduce error and become prone to human errors. A software
platform can be used to automate and enable precise experiments of evaporation from an
electric-field controlled tip using voltage pulses. The software can enable high-precision control of
the movement of the evaporator synthesis tip and application of voltage pulses on the evaporator
synthesis tip. The software can also enable the monitoring of the current flowing through the
tip/surface interface.
 [0083]
Referring to FIG. 27, a system for scanning pulsed evaporation of metals can use various
hardware including a voltage pulse generator, a current preamplifier, an oscilloscope, a data
acquisition board (DAQ), a computer (general purpose processor or logic chip, or a general
purpose computer), and communication interfaces associated therewith for communication
between the components.
 [0084]
The software can include a set of two processes running in two different physical locations.
Process 1 is a program. The program source code can be programmed into, for example,
Labview (National Instruments), a graphical interface programming language. The Process 2 can
be run, for example, over an AFM control software, and the source code can be programmed, for
example, using Nanoscript functions provided by the AFM manufacturer and general C
programming language. The two processes wait and exchange signals with the other process for
timing and scheduling the whole progress of hardware (HW) control.
 [0085]
Process 1 controls the data acquisition board (DAQ) to acquire the READY signal (voltage pulse)
from the AFM controller. Process 1 further controls the data acquisition board (DAQ) to send out
an OK signal to the AFM controller, so the AFM controller stops waiting and proceeds with the
next instructions. Process 1 also controls the function generator, and sets the amplitude, pulse
width, pulse period, pulse number, and pulse shooting.
 [0086]
Process 2 controls the vertical and lateral transitional movements of the probe. Process 2 control
can be designed, for example, to control the vertical and lateral transitional movements of a dual
tip probe having first and second tips as described above. Process 2 control can further control
the amount of cantilever bending to operate the first tip in contact mode while simultaneous
operating the second tip in non-contact mode and modulating the distance between the second
tip and the substrate. Process 2 also controls a Signal Access Module (SAM) to send out the
READY signal to Process 1. Process 2 controls the SAM to read in the OK signal from
Process 1 to stop waiting and proceed with the next instructions.
 [0087]
The following exemplary flow description describes the role and the signaling sequence between
different units of the system during two succeeding “feedback on” events. The probe controlled in
the flow description can be any known probe, for example, an AFM probe or the above-described
dual tip probe. The two separate process (Process 1 and 2) can run on different computers
(PC1 and PC2), and can exchange signals to schedule the events.
Flow Description (Between Two Successive Feedback-Ons)
 [0088]
 [0000]
1 The system gets into feedback. The tip gets in contact with the substrate, and the
feedback loop maintains a constant set point value.
2 The process 2 sends a ready signal to process 1 and enters into waiting mode,
where it waits for an OK signal form process 1.
3 Process 1, knowing that the tip is in feedback with the substrate, makes the
function generator apply the pulses with predefined amplitude, period and width.
4 Process 1, after having finished with applying pulses, sends an OK signal to the
AFM controller.
5 The AFM controller, having received an OK signal from process 2, turns off
feedback and lifts the probe from the substrate, and moves the probe to the next
position (go there).
6 After reaching the next position, the AFM lowers the tip down 90% the distance it
lifted before, and from there turns on the feedback loop. The probe makes a soft
approach, eventually reaching a constant set point controlled by the feedback loop.

Synthesis of Nanostructures using Block Copolymer Template
 [0089]
Referring to FIGS. 19 and 20, nanostructures can be formed using for example a block copolymer
as a template. The template can be formed by patterning a block copolymer on a substrate using
any known writing method such as dip pen nanolithography. The dual tip probe can be used, for
example, to pattern the block copolymer. Once patterned on the substrate, the block copolymer
can phase separate to form the template having at least one smaller dimension than the initially
formed pattern. For example, the block copolymer can be patterned and then allowed to phase
separate to shrink the size of the templated pattern. The block copolymer can be, for example,
polystyrene-b-poly(2-vinylpyridine), polystyrene-b-polyethylene oxide, polystyrene-b-
poly(methylmethacrylate), polystyrene-b-poly(4-vinylpyridine), or polystyrene-b-
poly(methylmethacrylate). The template can be loaded with the nanostructure precursor material,
such as, for example, a metal, such as gold aluminum, silver, platinum, palladium, or silicon. The
block copolymer can then be removed, for example, by etching the polymer. For example, the
block copolymer can be removed using oxygen or hydrogen plasma etching or by exposure to UV
light. Referring to FIGS. 19 and 20, the block copolymer can be patterned to form a nanoarray or
nanostructures, such as an array of nanoparticles or a nanowire. The block copolymer template
method can be used to form nanostructures having narrower widths or diameters in a range of 1
nm to 50 nm. Other suitable ranges include for example, from 5 nm to 40 nm, 10 nm to 30 nm,
and 15 nm to 20 nm. The width or diameter of the nanostructure can be, for example, 1, 2, 3, 4, 5,
6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32,
33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49 or 50 nm. Other nanostructures
that can be formed using the block copolymer template include, for example, quantum dots,
nanodiscs, and nanotubes.
EXAMPLES
 [0090]
The following examples are provided for illustration and are not intended to limit the scope of the
invention.
Example 1Method of Forming a Dual Tip Probe

 [0091]
Referring to FIG. 21, a dual tip probe was formed by a mold and transfer process. Silicon wafers
were rinsed using acetone, methanol and isopropyl alcohol. Then, a 5000 Å silicon oxide layer
was thermally grown on the silicon wafers at 1100° C. for 13 hrs 25 min. After the openings were
etched using HF for 6 mins into the silicon wafers, the dual-tip cavities were etched using KOH.
Optical microscopy was used to verify that the tip etching was complete. Subsequently, the oxide
layer was stripped using BOE. A 2500 Å layer of Cu were sputtered onto the patterned wafer as a
sacrificial layer. Then, an 1500 Å layer of low-stress silicon nitride was deposited using STS
plasma enhanced chemical vapor deposition (PECVD), and the first silicon nitride first (reader) tip
was patterned and etched.
 [0092]
A 500 nm Au layer was lifted off as the seed layer for the dual-tip cantilever and holder. A 10 μm
thick AZ4260 layer was patterned as the mold and 1 μm NiFe was electroplated inside the
cantilever and probe holder area following 100 nm Au deposition as the adhesion layer inside the
probe holder area. The 500 μm thick SU8 holder was spun on the wafer at 500 rpm for 30 s. The
thickness of the spun SU-8 2075 was about 500 μm. The coated wafer was prebaked; the
hotplate was ramped up from room temperature to 105° C. using 150° C./hr ramp and soaked for
15 hr. Following the prebake, the SU8 coated wafer was exposed for 2880 mJ. Then, the wafer
was post-exposure baked by ramping the temperature from room temperature to 105° C. using
150° C./hr. The wafer was soaked at 105° C. for 0.5 hr, and then the temperature was ramped
down at 15° C./hr to room temperature. Finally, the SU8 coated wafer was developed for 1 hour
to complete the fabrication process. The dual-tip probes were released by immersing the wafers
in an aqueous acetic acid/peroxide solution (Acetic Acid:H2O2:H2O=1:1:10) for 4 hours.
 [0093]
Referring to FIG. 22, the cantilevers of the metal dual-tip probes formed in accordance with the
above-described method demonstrated some bending. Bending of the cantilever can lead to
difficulty aligning a laser on the cantilever. Without intending to be bound by theory, it is believed
that the bending resulted from the bimetallic structure of the gold seed layer and the NiFe
electroplating layer, which possessed intrinsic strain causing the cantilever to have unbalanced
stress. The unbalanced stress can also result in unpredictability between the bending of individual
cantilevers in an array.
Example 2Method of Forming a Dual Tip Probe
 [0094]
The above-described mold and transfer method was used to make a dual tip probe, with two
modifications to the method: (1) A Cu sacrificial layer is chosen as the electroplating seed layer
so that the metal cantilever has only a single NiFe layer, and (2) the second metal probe was
protected using a photoresist during the NiFe electroplating so that the thickness of the second
metal probe was precisely controlled by the thermal evaporation. Referring to FIG. 23, the
bimorph structure of the dual tip probes of Example 1 was replaced by a single layer of silicon
nitride, which can aid in eliminating the bending problem identified with the dual tip probes of
Example 1. Without intending to be bound by theory, it is believed that the bending problem was
not experienced with the dual tip probes formed by the method of Example 2 because the stress
in the use of only a single NiFe layer in the cantilever is even. Further, it is believed that the
electroplated NiFe area inside the holder was limited to increase the adhesion of the SU8 holder.
Arrays of holes were designed inside the SU8 holder to release thermal stress, thereby avoiding
the peeling issue. While some cantilevers exhibited some bending, it is believed that this was a
result of a fast current ramping rate during NiFe electroplating, which generated heat, causing the
NiFe layer to peel off towards the edges of the prove holder. It is believed that this problem can
be avoided by ramping the electroplating current more slowly.
 [0095]
Table 1 illustrates the dimensions of various dual tip probes formed in accordance with the
above-described method.
 [0000]
TABLE 1
Dimension of Dual Tip Probes Formed Using
the Method In Accordance with the

Invention.
Tip
Number
Size
L1 L2 W1 W2
Design (μm) of Ribs
(μm) (μm) (μm) (μm)
1_1_0 124 40 40 40 5 0
1_2_0 124 40 30 30 5 0
TABLE 1
1_3_0 124 40 20 20 5 0
1_4_0 124 40 15 15 5 0
2_1_0 124 40 40 35 5 0
2_2_0 124 40 40 30 5 0
2_3_0 124 40 40 25 5 0
2_4_0 124 40 40 20 5 0
2_5_0 124 40 40 15 5 0
2_1_2 124 40 40 40 8 0
3_1_2 124 40 40 40 11 0
1_1_3 124 40 40 40 5 1
2_1_3 124 40 40 40 5 2
3_1_3 124 40 40 40 5 3
4_1_3 124 40 40 40 5 4
1_1_1 124 40 40 40 5 1
2_1_1 124 40 40 40 5 2
3_1_1 124 40 40 40 5 1
4_1_1 124 40 40 40 5 2
Example 3Method of Making a Thermal Probe
 [0096]
Referring to FIG. 8, the dual tips probes can include a thermal tip as the synthesis tip. The
thermal tip was formed by attaching a resistive heater to the synthesis tip. The resistive heater
was formed by patterning metal wires onto the silicon nitride tip. The metal wires included a 10
nm Cr wire, a 30 nm Pt wire, and a 400 nm Au wire. The Au was used as the primary conductor,
the Pt was used as a diffusion barrier, and the Cr was used as an adhesive layer. Referring
to FIG. 8A, the metal wires were wire bonded to the tip using the standard technique. The probe
was processed through wafer bonding, tetramethylammonium hydroxide (TMAH) etching, and
oxide etching, leaving a 100 mm diameter silicon frame with all the die locations attached
securely, but nevertheless able to be diced quickly, and with the cantilevers all free and clear.
Optical microscopy confirmed that the tips were well formed, and the cantilevers without gold
were flat and uniform. Cantilevers with metal exhibited a stress mismatch between the Au and
silicon nitride. Die yield was about 70% to about 80% on the first wafer.
Example 4Force-Distance Curve Calibration of Dual Tip Probe
 [0097]
Referring to FIGS. 14A-D, as a control, force-distance measurements of a single tip (not shown),
commercially available probe (Pacific Nanotechnology) were taken. It was observed that the knee
in the approach curve followed by a sharp rise in force is indicative of contact. The sharp dip is
associated with attractive capillary forces that cause the tip to “snap” into contact with the
substrate, and the rise indicates the increased force necessary to push the tip into the surface
after contact has been made. Generally, the flat line of the curve higher on the relative distance
axis indicates that the tip and substrate are not in contact. This curve was measured multiple
times starting from different z-piezo distances and always the same phenomenon was observed.
 [0098]
The spring constants for the silicon nitride dual tip probes tested was assumed for all
measurements to be 0.200 N/m with a sensitivity of 3.506 mV/nm. This estimated spring constant
value was used because the silicon nitride dual tip probes were formed from masks used for the
commercially available NanoInk Active Pen arrays. For the Active Pen arrays, the reported spring
constant of silicon nitride tips which were 30 μm wide and 150 μm long is 0.180 N/m; these
geometric parameters most closely resemble the dual tip probes tested, which have a maximum
width of 40 μm and maximum length of 164 μm.
 [0099]
For the dual tip probes with a reflective back layer, a photodiode was able to detect and measure
a sufficient laser signal. Thus, it was possible to make force-distance measurements and detect
distinct points of contact for both the first and second tip. The contact of the second tip is
indicated by a second knee in the force distance curve with a concomitant increase in the force
required to further extend the z-piezo. Though in the single tip force-distance curve there was a
flat line for the points where there was no tip-substrate contact, the dual tip probe does not
generate a straight line when not in contact. Without intending to be bound by theory, it is
believed that this anomalous behavior is the result of the dual tip cantilever bending beyond the
range of the z-piezo (13 μm). For bending that exceeds this amount, there may be a laser signal
detected that is not real and not representative of tip-substrate contact.
 [0100]
Referring to FIGS. 14A-14D, a comparison of how dimensions affect the dual tip probe's stiffness
was performed using force-distance curves. Referring to FIGS. 14A and 14B, two dual tip probes
having cantilever arms with different first second widths W1 were compared,
designs 110 and 130, the dimensions of which are in Table 1 above. The two probes had a
difference of 20 μm in the first section widths, with design 110 being wider than design 130. It
was difficult to observe distinct points for design 110 when the writing and synthesis tips made
contact with the substrate. For initial contact of dual tip probe design 110 between relative
distances of −0.75 and −1.00 μm, the stiffness was −19.8 nN/μm, while further contact between
the relative distances of −1.28 and −1.86 μm the stiffness was −67.4 nN/μm. For dual tip
design 130, when the first tip makes contact, the stiffness is −10.8 nN/μm, as given by the linear
slope; when both tips are in contact, the stiffness increases by more than two-fold to −28.1
nN/μm. The force distance curves show that there is a 200 nm vertical distance between where
the first (imaging) tip makes contact and the second (synthesis) tip makes contact with the
substrate. It was found that the stiffness values are higher for dual tip probe with a large first
second width (i.e. design 110).
 [0101]
In another dual tip design comparison, the effect of ribs disposed between the first and second
tips was examined by comparing Design 110 and 213. Again, it was difficult to observe distinct
contact points for Design 213 that distinguish the first tip from the second tip, and only one
stiffness value was measured, −752 nN/μm (FIG. 14B). The addition of two ribs makes the
stiffness values increase by an order of magnitude. Moreover, with such stiff tips, it was difficult to
discern the contact point of the synthesis tip.
Example 5Imaging Capabilities of the Dual Tip Probe
 [0102]
Imaging capabilities of the dual-tip probes were demonstrated in both contact mode and non-
contact mode using a calibration grid. Images were obtained on a 9.9×9.9×0.175 μm calibration
grid using the dual tip probe (FIG. 24B), and compared to images obtained by high-resolution
scanning probe tips from NanoProbes, Inc. (Yaphank, N.Y.) (FIG. 24A). The dual-tip probes were
capable of obtaining high-quality images. The high-resolution probes and the dual-tip probes
obtained the same value for the depth of the wells in the calibration grid (175 nm). There was a
discrepancy, however, in the wall-to-wall distance. While the high-resolution probes measured a
distance of 9.9 μm, a distance of 10.2 μm was obtained with the dual-tip probes. Without
intending to be bound by theory, it is believed that this discrepancy is the result of the larger tip
radius of the dual-tips. Importantly, the dual-tip probes obtained images under high load,
indicating that even when large amounts of force are applied to the cantilever, the writing tip does
not interfere with imaging capabilities.
Example 6Formation of Nanowires and Carbon Nanotubes using Field-Induced Evaporation
 [0103]
Gold was evaporated using electric field induced evaporation of gold from a conductive synthesis
tip onto a silicon dioxide surface. Low-resistivity AFM tips were used as the synthesis tip and
were coated with a 5 nm Cr adhesion layer followed by the thermal deposition of a 100 nm gold
layer. Patterns of gold on the surface were generated by electric-field induced migration of the
gold from the probe tip to the surface. A custom-built platform was used to induce the deposition
of gold onto the surface. Short electrical pulses (1-100 ms) of 20 V bias were applied to the tip.
The pulses were controlled with custom LABVIEW software (National Instruments, Austin, Tx)
that operated a pulse generator through a GPIB interface. The probe was mounted onto a
Multimode III AFM platform (Digital Instruments) with a MMTR-TUNA-CH cantilever holder that
isolates the piezo from the electric fields applied to the probe. Evaporation was induced in contact
mode, and pulses were monitored in real-time with an oscilloscope.
 [0104]
The resulting patterns of gold nanoparticles were observed by both AFM topological imaging
(FIG. 25) and SEM characterization (FIG. 26). By applying 1 ms pulses at a 20V bias and rate of
10 Hz, a square pattern of 10 nanoparticles per line was generated. The height of the gold
particles was measured to be approximately 8 nm by AFM topographical imaging. The patterns
were also seen in the SEM, and the contrast observed in the images is consistent with gold on
silicon dioxide. Further confirmation that the structures were gold was established by imaging
single particles evaporated from the probe. Larger particles were generated by applying longer
pulses to the tip. A 10 ms pulse generated gold features 250 nm in height with distinctive
hexagonal shapes and angles of 60°. Because gold (111) has a hexagonal lattice, it is believed
that if the gold forms in a crystalline fashion, the resulting nanostructures would be hexagonal.
Additionally, larger features contain terraces that are also indicative that single-crystal gold was
formed. This method of fabricating nanostructure by varying bias, pulse width and pulse period
can result in the ability to determine the position and feature size of nanostructures with precise
control.
 [0105]
FIG. 3 shows an non-contact (nc) AFM images of Au patterns on a silicon dioxide surface drawn
with this technique using a dual tip probe. A series of nc-AFM images of dot-shaped Au pattern
before and after multiple depositions are shown in FIGS. 3A-3C, clearly demonstrating the
electric field-induced evaporation onto the surface. By applying 20 μs at a 12 V bias, a square
pattern with 25 dots was generated (FIG. 3B), compared to the surface before deposition (FIG.
3A). It was determined that the first pattern contained an error, namely the omission of a square
pattern of 16 dots between the 25 dot square first pattern. This error was detected using the non-
synthesis probe, and a second square pattern with 16 dots (FIG. 3C) is drawn between the first
pattern with the same voltage pulse and tip to correct the error. This result clearly reflects the
ability of the dual tip probe to determine the site-specifically controlled protocol for patterning Au
dots. Importantly, this experiment demonstrates error-correcting ability of the dual tip probe with
nanometer precision.
Example 6Nanoreactor Well Formation
 [0106]
Referring to FIGS. 16A-16D, nanowells were fabricated on silicon dioxide wafers. Referring
to FIG. 16C, a nanowell was formed using electron beam lithography. A 120 nm layer of
poly(methyl methacrylate) (PMMA) photoresist was spin-coated onto a silicon substrate and
fabricated circular well patterns about 50 nm in diameter using electron beam lithography (EBL).
The nanowells were shown to be highly ordered and uniform by AFM imaging, but the throughput
was slow because EBL is a serial technique.
 [0107]
Referring to FIG. 16D, electrochemical methods were also used to fabricate AAO nanowells. In
one experiment, 70 nm of aluminum was evaporated and anodized for 40 seconds in 0.3M oxalic
acid with 40V applied bias. In this process, the choice of acid determines the pore diameter.
Oxalic acid made 30-50 nm diameter pores. The nanowells were characterized using scanning
electron microscopy (SEM). The bottom side of the AAO in contact with the silicon substrate
formed well-ordered pores, the top side remained rough and disordered. In order to decrease the
surface roughness and have well-aligned nanowells extended to the top side, a thicker layer of
aluminum (225 nm) was evaporated, a first anodization (0.3M oxalic acid, 80 seconds, 40V) was
completed, and the top layer was etched (phosphoric and chromic acid, 63° C., different times),
and added a second anodization step.
 [0108]
Referring to FIG. 16B, nanotemplates, such as nanowells were also formed using phase
separating polymers like polystyrene (PS) and polymethyl methacrylate (PMMA). Immiscible
polymer phase separate into well ordered domains that can be removed selectively, leaving
behind well-ordered wells. PMMA cylinders can be aligned vertically in a PS matrix such that they
are perpendicular to the substrate. By exposing the PS to ultraviolet (UV) light, the polymer
crosslinks and selective removal of PMMA can be achieved with acetic acid washing. AFM
topography images reveal that there are indeed PMMA nanowells approximately 15 nm in
diameter and height. To render the nanowell surface hydrophilic, timed oxygen plasma cleaning
experiments were performed to make sure PMMA was not destroyed or removed.
 [0109]
Following a one minute plasma clean, nanostructure precursor materials (e.g., silver nitrate,
sodium citrate, and sodium hydroxide) were dropcast onto the wells. These precursors were
exposed to UV light for about 30 minutes form nanoparticles. Nanoparticles, however, did not
form inside the wells. Without intending to be bound by theory, it is believed that this is most likely
because of the high surface tension of water. By creating nanowells using several different
approaches, nanoarchitectures with different size, shape and surface chemistry can be created.
 [0110]
While the present invention has now been described and exemplified with some specificity, those
skilled in the art will appreciate the various modifications, including variations, additions, and
omissions that may be made in what has been described. As one example, while various
embodiments have been described as including a cantilever with two tips, other embodiments are
contemplated to have more than two tips, e.g., three, four, or five tips, without limit. Accordingly, it
is intended that these modifications also be encompassed by the present invention and that the
scope of the present invention be limited solely by the broadest interpretation that lawfully can be
accorded the appended claims.
 [0111]
All patents, publications and references cited herein are hereby fully incorporated by reference. In
case of conflict between the present disclosure and incorporated patents, publications and
references, the present disclosure should control.
PATENT CITATIONS
Cited Patent Filing date Publication date Applicant Title
Btu Engineering
US4080510 * 18 Nov 1976 21 Mar 1978 Silicon carbide heater
Corporation
US5266801 * 26 Jan 1993 30 Nov 1993 Digital Instruments, Jumping probe

Inc. microscope
Mechanically coupled
International
alternatively usable
US6079255 * 13 Mar 1996 27 Jun 2000 Business Machines
cantilever structures for
Corporation
scanning a surface
Methods utilizing
scanning probe
Northwestern
US6635311 * 5 Jan 2000 21 Oct 2003 microscope tips and
University
products therefor or
products thereby
Methods utilizing
scanning probe
Northwestern
US6827979 * 24 May 2001 7 Dec 2004 microscope tips and
University
products therefor or
produced thereby
Method and apparatus

Veeco Instruments
US6862921 * 13 Jul 2001 8 Mar 2005 for manipulating a
Inc.
sample

Veeco Instruments
US7040147 * 8 Mar 2005 9 May 2006 for manipulating a
Inc.
sample
Patterning of solid state

Northwestern
US7273636 * 17 Dec 2002 25 Sep 2007 features by direct write
University
nanolithographic printing
Dual tip atomic force

Interuniversitair
microscopy probe and
US7500387 * 28 Jul 2006 10 Mar 2009 Microelektronica
method for producing
Centrum Vzw (Imec)
such a probe
The United States Of High resolution coherent

US7526949 * 21 Jul 2006 5 May 2009 America As dual-tip scanning probe
Represented By The microscope
Secretary Of The
Army
Apparatus for recording

and reproducing high-
US20020080709 * 18 Jul 2001 27 Jun 2002 Kang-Ho Park density information
using multi-functional
probe
High speed/high density

optical storage system
US20020080710 * 12 Dec 2001 27 Jun 2002 Kim Jeong Yong using one-dimensional
multi-function/multiple
probe columns
Nanolithography
methods and products
US20020122873 * 28 Jan 2002 5 Sep 2002 Mirkin Chad A.
therefor and produced
thereby
Scanning probe
microscopy probe and
US20040228962 * 16 May 2003 18 Nov 2004 Chang Liu
method for scanning
probe contact printing

US20050145021 * 8 Mar 2005 7 Jul 2005 Ami Chand for manipulating a
sample
Storage device having a

US20050201257 * 9 Mar 2004 15 Sep 2005 Champion Corbin L.
probe with plural tips
Scanning probe for data

US20060043288 * 11 Dec 2002 2 Mar 2006 Binnig Gerd K
storage and microscopy
* Cited by examiner
REFERENCED BY
Citing Patent Filing date Publication date Applicant Title
Scanning probe
US8296857 * 17 Dec 2008 23 Oct 2012 Specs Zürich GmbH microscope with
current controlled
actuator
Piezoresistor
US8393011 13 May 2009 5 Mar 2013 Nanoink, Inc. height sensing
cantilever
Nederlandse Organisatie
Voor Toegepast-
US8914910 * 21 Dec 2012 16 Dec 2014 Probe calibration
Natuurwetenschappelijk
Onderzoek Tno
System and
method for high-
speed atomic force
US9091705 * 1 May 2013 28 Jul 2015 Boise State University microscopy with
switching between
two feedback
loops
Piezoresistor
US20100100989 * 13 May 2009 22 Apr 2010 Nanoink, Inc. height sensing
cantilever
Scanning probe
US20100115672 * 13 May 2009 6 May 2010 Northwestern University
epitaxy
Scanning probe
microscope with
US20110296564 * 17 Dec 2008 1 Dec 2011 Specs Zurich Gmbh
current controlled
actuator
System and
method for high-
US20130312142 * 1 May 2013 21 Nov 2013 Boise State University
speed atomic force
microscopy
Nederlandse Organisatie
Voor Toegepast-
US20150013038 * 21 Dec 2012 8 Jan 2015 Probe calibration
Natuurwetenschappelijk
Onderzoek Tno
Multiple
Integrated Tips
US20160252545 * 26 Feb 2016 1 Sep 2016 Xallent, LLC
Scanning Probe
Microscope
Improved
WO2011130446A1 13 Apr 2011 20 Oct 2011 Nanoink, Inc. cantilevers for
deposition
Functionalizing
biosensors using a
WO2011133663A1 20 Apr 2011 27 Oct 2011 Nanoink, Inc.
multiplexed dip
pen array
Method and
apparatus for
WO2013067395A2 2 Nov 2012 10 May 2013 Nanoink, Inc.
improving ink
deposition
* Cited by examiner
CLASSIFICATIONS
U.S.
Classificat 850/21
ion
Internatio
nal
G01Q60/00
Classificat
ion
Cooperati
ve
G01Q80/00, G03F7/0002, G01Q70/10, G01Q70/16, G01Q40/00, G01Q20/04
Classificat
ion
European
B82Y10/00, B82Y35/00, B82Y40/00, G01Q70/10, G01Q80/00, G01Q70/16, G03F7/00A, G
Classificat
01Q40/00, G01Q20/04
ion
Google Home - Sitemap - USPTO Bulk Downloads - Privacy Policy - Terms of Service - About Google Patents - Send Feedback
Data provided by IFI CLAIMS Patent Services

Biological nanostructures: platforms for analytical chemistry at the sub-zeptomolar level
Graham J. Leggetta
Author affiliations
Abstract
The analysis and manipulation of molecules at the sub-zeptomolar level (i.e. from 1 to 600 molecules) remains the unconquered frontier of analytical chemistry.
While some techniques offer sensitivity to single molecules, there are no established tools for the manipulation of such small quantities of material. Scanning probe
lithography has begun to provide practicable means to manipulate biological organisation on length scales of 100 nm and less, and three promising approaches
(dip-pen nanolithography, nanoshaving and scanning near-field photolithography) are reviewed. Each offers extraordinary spatial resolution combined with the
capability for use under ambient and, in some cases, fluid conditions. These techniques offer a multitude of strategies that may at last make the manipulation of
handfuls of molecules—and perhaps single molecules—a practical possibility for the analytical chemis
A diffusive ink transport model for lipid dip-

pen nanolithography†
A. Urtizberea and M. Hirtz *
Institute of Nanotechnology (INT) and Karlsruhe Nano Micro Facility (KNMF), Karlsruhe Institute of
Technology (KIT), Hermann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany. E-
mail: michael.hirtz@kit.edu
Received 30th June 2015 , Accepted 23rd July 2015
First published on 12th August 2015
Despite diverse applications, phospholipid membrane stacks generated by dip-pen nanolithography

(DPN) still lack a thorough and systematic characterization that elucidates the whole ink transport
process from writing to surface spreading, with the aim of better controlling the resulting feature size
and resolution. We report a quantitative analysis and modeling of the dependence of lipid DPN
features (area, height and volume) on dwell time and relative humidity. The ink flow rate increases
with humidity in agreement with meniscus size growth, determining the overall feature size. The
observed time dependence indicates the existence of a balance between surface spreading and the ink
flow rate that promotes differences in concentration at the meniscus/substrate interface. Feature shape
is controlled by the substrate surface energy. The results are analyzed within a modified model for the
ink transport of diffusive inks. At any humidity the dependence of the area spread on the dwell time
shows two diffusion regimes: at short dwell times growth is controlled by meniscus diffusion while at
long dwell times surface diffusion governs the process. The critical point for the switch of regime
depends on the humidity.
1. Introduction
In dip-pen nanolithography (DPN), modeling is a key element of the nanofabrication process, required not only
to identify the critical parameters of the ink transfer and subsequent reorganization but also to actually quantify
their influence and therefore the process sensitivity to them in order to achieve accurate writing control. This is
also true for the development of DPN with lipids (L-DPN) by enabling a systematic informed choice of ideal
materials and/or a combination of materials for a given application need, as well as to improve the quality of the
outcome.1 As a striking example, a multilayer structure (height and width) decisively determines the
functionality of lipid multilayer gratings.2
The factors that govern the transport and assembly of the different ink/substrate systems patterned
with DPN mainly depend on the physicochemical properties of the ink: one main distinction that can
be made here is the difference in molecular (diffusive) inks and liquid inks. In liquid inks, DPN
proceeds in a bulk transfer of material into a droplet on the substrate over an ink meniscus that gets
snapped off upon tip retraction. Here, transport is governed by the fluid dynamics of the capillary
rupture process and the meniscus volume, so that patterned features depend strongly on the contact
angle, viscosity, retraction speed, dwell time and volume of ink at the pen.3–5 Basically, the competition
between surface energy and ink viscosity connected by a variable shaped meniscus decides the
outcome, while surface diffusion and spreading as well as tip ink solubility kinetics are expected to
play a minor role. High molecular weight polymer inks are also classified as liquid inks due to the fact
that their transport proceeds by a bulk liquid flow, better described as capillary mass transport. In this
case meniscus transport and surface diffusion are strongly dependent on the ink viscosity. The
polymer chains pertain to a molecular entanglement, absent in liquid inks, yielding a high viscosity.
Contrary to other liquid inks, whose fluidity provides a volume dependence that increases
exponentially with time,3 patterned features of polymer inks show a dependence on dwell time that
saturates at long dwell times.6 This behavior is probably due to the increased viscosity, leading to the
point where even the features themselves influence the transport during patterning. Both systems can
be described within the growth mechanism of bulk flow within mass transfer.6,7 Nevertheless, a
complete analytical model has not yet been developed.
Neither of these liquid inks forms chemical bonds between its constituents and the substrate, but
can rather be seen as a drop of liquid on a surface. The binding to and diffusion over the surface is,
however, the main characteristic of molecular inks. Here, ink molecules diffuse until they bind to the
substrate and the nature of this bond can ultimate determine the geometry of the final structure.
Strikingly, there are reports showing that in some ink/substrate systems the ink molecules do not
necessarily rearrange over the surface in an independent fashion but exhibit collective behaviour,8 nor
do they always spread homogeneously,9 up to the point of displaying anisotropic patterns instead of
the conventionally expected round features.10 Also mentionable are some reports showing that
molecular inks can grow in 3-D structures when controlling the balance between the surface spreading
and ink flow rate.11 Molecular inks – with the prominent example of the thiol ink on a gold substrate
system – have been widely analyzed and models that encompass not only the surface diffusive stage
but also the whole transport process have been developed.11,12
Lipid inks share some attributes of a liquid as well as of diffusive molecular inks. They retain the
fluidity aspect of liquid systems, due to their fluid properties.13 But they also share their assembly and
spreading behaviour with molecular ink systems, which ultimately determines the geometry of the
final structure.14 Many empirical studies have identified the experimental variables that influence
feature size in L-DPN. Some initial quantitative studies conducted previously addressed the writing
process,15,16 the thickness dependence on the tip speed and humidity,17 membrane stack
organization,18 and the accurate height characterization of multilayer thickness by precise calibration of
fluorescence microscopy.19 All reports agree that as a ‘rule of thumb’ humidity controls the phase
behaviour of the ink at the tip as well as the meniscus itself, in addition to their diffusion and
spreading.2,17 Given the amphiphilic nature of the phospholipids, this is not surprising; actually,
transport is expected to proceed by the condensed meniscus.13 Subsequently, the lipids become
physisorbed on the substrate, and the substrate properties (hydrophilic or hydrophobic; also surface
energy) influence the surface diffusion, spreading and membrane organization.14 The time the tip is in
contact with the substrate (dwell time) limits the amount of ink delivered. The lipid ink transport is
usually believed to follow the scheme presented in L-DPN in Fig. 1. Though these experimental
parameters have been experimentally known to influence L-DPN transport, yet a fundamental
question remains open: how do these parameters quantitatively influence the feature size and shape? Is
it possible to develop a comprehensive model for the morphology and dynamics of L-DPN transport?
Fig. 1 Scheme of the L-DPN process. A tip coated with a lipid ink is brought into contact with a substrate. The lipid ink tra
meniscus onto the substrate to form the desired feature.
Patterns of DOPC on glass substrates have been quantitatively analyzed as a function of dwell time
at different relative humidities (RH). DOPC is a suitable ink for DPN at room temperature (RT),17 and
it is a well-known standard lipid for unsaturated lipid bilayer membranes.20 DOPC is widely employed
as a lipid ink carrier in lipid mixtures. Being in the Lα liquid state at RT, this amphiphilic ink shows
interesting properties concerning the diffusive dependence and mobility on hydration state, thereby
allowing the control of its diffusion and therefore transport by RH under ambient conditions. All in all,
it is the key molecule to be employed within the analysis of controlled transport in L-DPN. Here, dot
features were chosen for the study to ensure an equilibrium growth regime and to be able to focus on
the basic transfer mechanisms, instead of lines that would require a more complicated description of
additional dynamic processes. In the case of lines, the establishment of meniscus is time
dependent.21 Furthermore, and even more importantly for lipids that grow in a multilayer fashion, the
large concentration gradient between the tip and the surface does not reach equilibrium in line writing
since the tip is constantly exposed to clean surface areas as it travels, thereby creating a large driving
force for ink deposition.21,22
Experimental data show that the surface spreading of lipid dots exhibits dynamics resembling
molecular ink diffusive behaviour, yet features can grow in a 3-D fashion. The height growth rate
shows a stronger dependence on humidity than the surface spread. An extensive analysis reveals that
the dependence of both quantities on relative humidity and dwell time arise from their dependence on
the ink flow rate, i.e. the amount of material delivered per unit time. The ink flow rate is shown to
depend on humidity following the reported meniscus size dependence on RH. Additionally, it is also
found to depend on dwell time. This is associated with the lipids’ diffusive transport towards the
surface driven by a difference in lipid concentration at the tip/meniscus and meniscus/substrate
interface, i.e. transport through the meniscus follow Fickian diffusion. Additionally, the concentration
at the meniscus/substrate interface is shown to depend on time. The surface diffusion and height
growth rates are therefore controlled by the RH and dwell time through the dependence of the ink
flow rate through the meniscus. The role of the substrate surface energy emerges when the wetting
angle of the structures is analyzed as a function of time and humidity. Surface energy governs the
overall feature shape.
An analytical model is proposed to describe the surface spreading. It is based on a reported model
for diffusive molecular inks,12 yet modified to include the properties of lipid ink transport. It shows that
in the case of lipid ink transport, both the surface diffusion regime and the meniscus diffusion regime
are present, in contrast to the molecular ink transport in which only one of them is followed,
depending on the ink characteristics. At short dwell times growth is controlled by meniscus diffusion
while at long dwell times surface diffusion governs the process. The critical point for the switch of the
regime depends on the humidity. This is associated with the diffusive characteristics of the lipids
whose transport is ultimately controlled by the ink transport rate.
These experimental results provide a comprehensive account of the ink transport in dip-pen
nanolithography with lipids.
2. Results
The following results are reported for one tip of the tip array, in order to analyze a coherent ink transport. Data
corresponding to three additional tips of the tip array are included in the ESI† showing that the principal
mechanisms agree from tip to tip. Effects and causes of inter-tip variance are discussed there.
Smaller features show clearly the steps corresponding to the phospholipid bilayers height (Fig. 2).
However, for larger features, imaging resolution does not allow one to see these steps, resulting in a
smooth dome shape.
Fig. 2 DOPC dot patterned on a glass at RH 30.6 ± 0.3% and dwell times of (a) 0.5 s and (b) 10 s, measured by AFM. Small
three layers height. The resolution does not allow seeing these steps for larger features presenting a dome shape as in (b
2.1 Feature growth size dependence on dwell time and humidity
The area, maximum height and volume of the features show a strong correlation with dwell time and relative
humidity as shown in Fig. 3. As a first approximation, feature growth agrees with an empirical power
law y = A + Btn, for any of the three measured dimensions (area, height and volume). As suggested in ref. 23, the
dependence shown is indicative of the ink to follow a diffusive transport mechanism. Note that the
parameter n provides information about the dynamic growth rate of the corresponding quantity.23 It decreases as
RH increases for all of them, though different dependences are followed by each: n-height dependence on
humidity is uniformly decreasing while n-area stays roughly constant up to 35% where it starts to slowly
decrease.
Fig. 3 AFM area, height and volume of DOPC dots over glass at humidities of 39.3 ± 0.2% (■), 37.8 ± 0.1% ( ), 35.9 ± 0.1%
± 0.1% ( ), 30.6 ± 0.3% ( ) and 28.9 ± 0.2% ( ). Solid lines are fits of the data to the empirical power law. The inset show
decreases with increasing humidity.
This difference in dependence of the growth rate factor n on humidity can be better seen in a
logarithmic representation (Fig. 4). The height growth rate shows a stronger dependence on humidity
than that of radial spreading, which is rather similar for the different RH, only shifted in the initial
size, probably due to differences in the initial meniscus formation. The growth rate ndependence on
humidity, shown in the insets of Fig. 3, resembles molecular ink transport dynamics with varying
temperature.11The molecular transport rate depends on the balance between the tip flow rate and the
surface diffusion rate. Radial spreading follows R ∝ tν, where the growth rate coefficientν is
a being the tip contact radius, α the ink diffusion coefficient, related to the mobility of molecules (molecular ink
transport), and Dthe surface diffusion coefficient. When ink mobility and surface mobility are increased
simultaneously, ν does not change. In lipid transport, increasing the humidity increases lipid diffusion,24 thereby
the lipid ink flow increases, and also lipid surface diffusion. In combination, the net feature area growth rate
does not change when changing from one humidity to another. The fact that nslightly decreases with increasing
humidity for RH > 36% indicates that the ink flow rate is getting more enhanced, yet weakly, over surface
diffusion.
Fig. 4 AFM radius and height of DOPC dots over glass at humidities of 39.3 ± 0.2% (■), 37.8 ± 0.1% ( ), 35.9 ± 0.1% ( ), 3
( ), 30.6 ± 0.3% ( ) and 28.9 ± 0.2% ( ). Solid lines are linear fits of data. The scale is 10-logarithmic.
In contrast, concerning height growth, increasing humidity promotes lipids to be displaced from
additional membrane stacks into the base membrane through dislocation places,25 so the equivalent
‘diffusion D’ along the height growth actually decreases with increasing humidity, relative to the flow
rate (the ratio aα/2D increases). Thereby, the net height growth rate n decreases with increasing
humidity. In contrast, for line writing the height growth rate increases upon increasing RH.17 This
difference between dot and line writing, the latter continuously exposing the bare surface upon tip
movement, indicates that an additional dynamic effect influences the flow rate, and that it is related to
the lipid concentration on the surface.
This suggests that not only the amount of material, but also the state of the lipid ink is strongly
controlled by the relative humidity. This is in agreement with the strong influence of humidity on the
mobility characteristics of the lipids themselves.26Actually, as will be shown below, the ink flow (ink
delivered per unit time) shows a time dependence in which not only the flow increases, but its rate
(dJ(t)/dt) is strongly dependent on RH. This feature resembles the dependence of the transport rate in
molecular inks, but on temperature, in the sense that RH not only enhances transport but triggers the
transport rate itself. It suggests that humidity plays a similar role in lipid transport kinetics to the role
temperature plays in molecular inks.11 This feature will be discussed below.
Yet the feature growth exponent n follows a defined dependence on humidity for area, height and
volume growth, suggesting that additionally a dynamic effect, related to RH, is influencing ink
transport. The increase of feature area as well as height with relative humidity at any dwell time is
shown in Fig. 5 (the corresponding linear scale graphs are shown in the ESI†). The feature area
growth depends quite similarly on humidity for different dwell times, but is merely shifted in the
initial size. As previously discussed, this indicates that surface diffusion compensates for increased ink
flow rate. This relationship between delivery (aα) and spreading rates (D) suggests that a higher ink
flow is also promoting a higher diffusion over the surface, i.e. that the speed of diffusion over the
surface is correlated with the lipid ink flow rate. This feature can be explained by a concentration
driven diffusion and will be discussed below. The effect is expected to be strongly dependent on the
substrate surface energy, i.e. the strength of interaction between the lipids and the surface.27
Fig. 5 Surface spread and height dependence on humidity at dwell times of 10 s (■), 8 s ( ), 4 s ( ), 2 s ( ), 1 s ( ), and 0
linear fits to the data.
In contrast, the larger the dwell time the weaker the height dependence on humidity. A parallelism
with the former situation suggests that dwell time is affecting ink diffusivity along the height growth
in the sense that with longer dwell time, the ink becomes less diffusive. This indicates that the net flow
acquires a stationary situation with time, which is also in agreement with a concentration driven flow.
The growth of the feature area S, shown in Fig. 3, does not follow the diffusive S ∝ t spreading
dynamics found in some reports on alkanethiol inks,28,29 which matches the continuum diffusion theory
in which a tip is treated as a point source of constant ink flow.28 A similar behavior has been shown in
some reports for alkanethiol inks when different humidities were analyzed.11,30 It was assigned to two
dynamic regimes that occur with varying dwell time.11,30 The regime change was ascribed in ref. 30 to
the time dependence of thiol concentration at the tip/meniscus interface. The thiol transport from the
tip to the surface starts in a dissolution-dominated regime (controlled by ink solubility at the tip) that
subsequently evolves towards the diffusion-dominated regime (controlled by diffusion over the
meniscus). Changes in the spreading kinetics of S(t) were related to time evolution of the
concentration at the tip/meniscus in the case of molecular inks. As shown below, in the case of L-DPN
it is the concentration at the meniscus/substrate that changes with time; yet in both systems it is due to
a time evolution of the ink concentration. However, there is a difference between both systems: in ref.
30 the effect of humidity appears as a multiplicative factor in regard to dot radius, in both regimes, and
so the influence of RH can be just scaled, i.e. humidity affects both regimes equally (see Fig. 1 in ref.
30). This suggests that in the case of molecular inks, RH just increases the amount of ink delivered.
Instead for L-DPN, as shown in the inset of Fig. 3, the growth factor n depends on humidity,
following a coherent dependence. This feature suggests, as discussed above, that in L-DPN humidity
plays a major role, enhancing lipids mobility and increasing lipid concentration at the
meniscus/substrate interface; as shown below, flow not only increases with RH, but RH controls its
rate (dJ(t)/dt). Thus humidity cannot be included in the present case just as a multiplicative factor, but
is rather, as previously mentioned, similar to the effect of temperature on the dissolution kinetics
reported by Cho et al.11: the factor that triggers the transport rate.
2.2 The flow rate
Both the feature area and height depend on humidity, i.e. ink diffusivity, as well as on the kinetics of the ink
transport itself. This indicates that they are thus governed by the ink flow rate J = ρV/t. This quantity is the
amount of delivered material by the tip through the meniscus towards the surface per unit time. It should be
emphasized that this is not a bulk liquid ink transfer to the surface. Instead, as was previously introduced, it is
likely that lipids diffuse through the meniscus to the surface. The flow is the amount of lipids diffusing per unit
time, but with a diffusivity that changes with the liquid state of the ink due to the RH.
Its magnitude in L-DPN (Fig. 6) agrees with the reported flow rates for alkanethiols.29 It is of the
same order of magnitude as the value expected considering the DOPC diffusion coefficient24 and
DOPC molecular size,31 indicating there is not a very large concentration gradient within the meniscus.
It is strongly dependent on the humidity and dwell time, though: while RH ranges from 29% to 39%,
dwell time ranges nearly in two orders of magnitude.
Fig. 6 Flow rate as a function of (a) relative humidity and (b) dwell time, showing that J depends on the fluidity provided
dynamic parameter modifies its magnitude. Solid lines are a guide to the eye.
Fig. 6(a) shows that the flow rate is only weakly dependent on humidity up to a threshold value
above which it increases strongly with RH, in agreement with molecular ink transport.32 This
observation is consistent with the reported water meniscus size dependence on relative
humidity.33 These features suggest that the lipid ink transport flow rate is governed by the meniscus
itself, being ultimately controlled by the RH. As the lipid diffusivity is strongly dependent on the
available water24 the condensed meniscus governs lipid diffusivity, therefore controlling the process of
ink transport. These results highlight the influence and importance of the meniscus for lipid ink
transport.
A difference between the diffusive transport of L-DPN and the liquid ink transport should be noted
here: at low humidities only a small meniscus will condense, up to a critical water pressure
(temperature) condition at which the condensation starts building a larger meniscus.
Changes in meniscus size influence not only its geometry and available fluid but also the
magnitude of the capillary force. A smaller meniscus is expected to have a larger curvature and
therefore higher capillary forces. Thereby, in a transport due to capillary forces as in liquid inks, one
would expect that the higher the transport rate the lower the humidity. However, in L-DPN, humidity
facilitates lipid diffusion, and the flow rate increases with larger RH. This effect will become even
more pronounced the larger the RH is.
Within this picture, in which transport is completely dominated by the meniscus, one striking
feature arises from Fig. 6(a): at higher dwell times the flow rate is larger. This indicates that the flow
rate is not only governed by the meniscus but an additional dynamic effect must be involved.
The dependence of the flow rate on the dwell time at different humidities, shown in Fig. 6(b),
resembles the dependence of the water meniscus pull-off force on the duration of the contact time, at
different humidities.34,35 Therefore, one might assign this feature to the fact that menisci continue
growing from vapor condensation,36 which at small distances proceeds viaKnudsen diffusion and from
the flow of the liquid substrate film over the surface.37,38 However, water meniscus J(t) decreases with
time,39 contrary to Fig. 6(b). Water meniscus growth rates do not account for the lipid inks’ flow rate
dependence on time. Considering that phospholipids would arrange at the meniscus/air interface with
their non-polar groups facing air, water vapor is hindered from condensing. Also, when the lipids start
spreading over the substrate water is no longer available from the surface to increase the meniscus
size.
Ultimately, this influence of the water meniscus growth, sometimes called meniscus instability, on
the ink transport rate can be analyzed following experiments reported in ref. 40. In Fig. 3 of ref.
40 error bars of dot area ratios are larger at the beginning of the patterning sequence, showing that
initially the condensed meniscus is unstable. In our experiment larger dwell times were printed first.
Yet, as shown in Fig. 7, the fluctuation in the volume of the dots is larger at small dwell times, i.e. for
the dots patterned at the end of the patterning sequence.
Fig. 7 The relative fluctuation in the volume of the patterned dots at each dwell time is not larger at the beginning of the
any humidity. This indicates that the flow is not more unstable at the beginning of the patterning sequence.
Both arguments, J(t) not resembling water meniscus growth and the volume fluctuation being
smaller at the end of the patterning sequence, indicate that flow changes with time are not arising from
meniscus size changes with time; instead they may arise from a dynamic effect of the flow itself. This
effect is in agreement with the time dependence discussed in Fig. 3–5, indicating that the net flow
reaches a stationary state with time, more likely due to an equilibrium in concentration differences.
This notion is ultimately in agreement with Fig. 6(b), showing an increase in flow with dwell time,
reaching equilibrium at large RH, i.e. the larger the flow the sooner the equilibrium is achieved. This
indicates that the dynamic parameter affecting feature size (Fig. 3) and flow rate dependence on time
(Fig. 6) may be an increase of lipid concentration at the meniscus/substrate with time until a stationary
equilibrium situation is achieved. The higher the RH is (and therefore the ink flow), the faster this
equilibrium is reached.
It should be mentioned that molecular inks display also a flow rate dependent on time.39 Yet in this
case, J(t) follows the water meniscus growth rates, so that in molecular ink changes in J(t) are
attributed to meniscus dynamics.39 However, lipid inks show a striking difference with regard to J(t):
while in molecular inks the flow decreases with time, in lipid inks the flow increases with time (Fig.
6(b)). These differences in the flow rate dependence on time between molecular and lipid inks likely
arise from the fact that the former are in a solid state on the tip, while the latter are in a more liquid- or
gel-like state. Due to their aggregation state molecular ink transport depends strongly on the ink
dissolution (molecular detachment) from the tip.11,29 The transport or deposition rate is actually
attributed more to the rate of dissolution than to phenomena within the meniscus itself.29,30 In these inks,
the relative humidity influence is reduced to the meniscus height dependence on RH, so that the
transition from one regime to the other is independent of humidity, i.e. the functional form of the
molecular ink dots area with humidity is the same for all dwell times.30 This explains why the
patterning flow rate in molecular inks agrees with the meniscus growth rate.39However, lipid inks are
fluid-like (DOPC is in the Lα liquid phase at RT), and their viscosity and therefore diffusion depend
strongly on the amount of available water. This implies a Fickian diffusion mechanism for lipid ink
transport through the meniscus in which the flow rate, due to differences in concentration between the
tip/meniscus and the meniscus/substrate interface, will increase until concentrations become
equilibrated. At large humidities, fluidity is enhanced, and the concentration at the meniscus/substrate
interface can reach equilibrium fast, as shown in Fig. 6(b). Lower humidities lead to slower diffusion,
thereby the flow rate is not able to bring the concentration into equilibrium: the concentration at the
meniscus/substrate interface is increased by the tip flow rate, but surface spreading decreases this
concentration again, keeping a net flow that slowly increases with time, while being unable to reach
equilibrium within the dwell time.
Ultimately, this indicates that the flow rate is changing not just because of a larger meniscus size:
the lipids fluid behavior itself is changed, suggesting that the ink phase is changing.
These results suggest that lipid ink transport will require a model in which ink flow will depend on
the concentration at the meniscus/substrate interface, controlled by the diffusion over the surface.
Also, it should include the balance between ink flow and surface diffusion rates for the description of
the feature growth. Finally, concerning ink supply, in lipid inks the key parameter will be the
hydration diffusion kinetics, rather than the chemical dissolution kinetics affecting molecular inks.
2.3 Feature shape: role of the substrate surface energy
As discussed above, radial spreading and height growth of the L-DPN features are fully governed by the ink
flow, which can be tuned by the humidity and whose magnitude varies with the dwell time, due to changes in
concentration at the meniscus/substrate interface. This is supported by Fig. 8 showing that the feature diameter
and height are directly related to each other, even for different humidities and different dwell times, despite the
membrane stack organization of the lipids on the surface. This underlines the fluid-like behavior of the lipid inks
during the L-DPN process.
Fig. 8 Dot height is closely related to dot diameter even at different dwell times and RH following a linear relationship. In
diameter derived from the linear fit intercept for zero dot height.
A linear fit provides the extrapolation for zero dot height, i.e. the minimum dot diameter, of about
215 nm, in agreement with the expected meniscus size33,39,41,42 (the absolute magnitudes are difficult to
compare with the literature as these references are for pure water menisci, i.e. no ink is present) and
the lateral resolution of L-DPN.17 It is about four times the magnitude of the tip size (see the ESI†).
The fact that the meniscus width seems not to depend on RH indicates that the tip–substrate
distance is small. At small tip–substrate distances meniscus size increases only slowly with
humidity43 and our measurements are not sensitive to these changes. Though the lipid features consist
of stacks of membranes, their overall shape can be still described as droplet-shaped for these very
small dot features. The droplet contact angle θ arises from the balance between the three interfacial
energies involved, substrate/air γsa, substrate/lipid γsl and air/lipid γal, as depicted in Fig. 9, where cos
θ = (γsa − γsl)/γal.
Fig. 9 Scheme and definitions of the contact angle of a lipid droplet on a hydrophilic surface.
The equilibrium contact angle calculated from h = (w/2)tan(θ/2) is shown in Fig. 10. The angle
follows a similar time dependence for all the humidities, leveling off at longer dwell times to θ ≈ 25°.
Considering the definition of equilibrium angle, this indicates that the dominant force influencing
shape for dot features in L-DPN is the interfacial energy between the ink and the substrate, i.e. when
flow and substrate energies become equilibrated, the equilibrium angle does not depend on RH.
Differences in spreading and membrane organization due to the relative strength of intermolecular
interactions as compared to the molecular interactions with the substrate were also observed when
writing on substrates with large differences in surface energy, e.g. graphene and silicon oxide.14
Fig. 10 Wetting angle at humidities of 39.3 ± 0.2% (■), 37.8 ± 0.1% ( ), 35.9 ± 0.1% ( ), 34.5 ± 0.2% ( ), 32.6 ± 0.1% ( ),
0.2% ( ), showing that, though equilibrium angle is RH independent (at any RH they extrapolate to the same value), its d
ink flow J, i.e. they are controlled by the balance between surface and flow rates. Solid lines are a guide to the eye.
Note in Fig. 10(b), however, that actually feature growth depends on RH due to the changing
fluidity of lipids in L-DPN. Humidity at 34.5% entails a large error and so the tendency with RH is
slightly masked.
Overall, this confirms that growth is controlled by the balance between surface diffusion and flow
rates, but when equilibrium between both rates is achieved the feature shape (equilibrium angle)
depends on the surface energy.
3. Discussion: qualitative picture of the ink transport
In order to develop a model for the growth of lipid structures during L-DPN, one has first to understand the
whole process from ink wetting to the spreading of lipids on the surface. In this section we propose a qualitative
description of the transport process for L-DPN, based on the knowledge gained and discussed in the previous
sections. Most of the analytical studies, modeling and simulations have been carried out on molecular inks,
especially alkanethiol inks on gold. Model alkanethiol inks such as 16-mercaptohexadecanoic acid (MHA) and
1-octadecanethiol (ODT) and their transport properties have been widely studied.11,12,22,29,40,44–46 Here, we present L-
DPN transport in analogy with the reported transport analysis on these molecular inks. We shortly review the
molecular transport characteristics, outlining the differences between both systems.
Within this framework and combined with previous experimental results an L-DPN transport
model is proposed in the following section.
We hypothesize that L-DPN transport proceeds through three stages (cf.ref. 47 for molecular ink
systems): (I) ink ‘dissolution’ from the tip at the meniscus/tip surface, (II) transport to the
meniscus/substrate surface interface and (III) surface diffusion and/or spreading. In some molecular
systems, since the slowest stage dominates the transport, one or two of these stages can be
neglected, i.e. that transport is mainly controlled by e.g. dissolution at the tip,29,30 surface diffusion,28,44 or
transport and subsequent diffusion.11 In L-DPN none of the transport stages can be completely
neglected, but actually they become coupled. This fact arises from the influence of one of the key
parameters that controls L-DPN transport: the fluidity of the lipid ink. Water in DOPC membrane
structures is localized only in small amounts in the hydrocarbon core48 and in large amounts around the
phosphocholine groups as well as in the interbilayer spaces.48,49 This inter and intramembrane water
provides lipid membranes fluidity,48,49 therefore the membrane diffusion coefficient strongly increases
with the amount of water available.24 Above a threshold value, membranes become fully hydrated and
from then on, one-dimensional swelling of bilayers takes place.24,49 This water-content dependence of
the lipid fluidity influences every one of the three transport stages.
3.1 Step I: ink dissolution at the tip/meniscus interface
Lipids at the tip will become hydrated at ambient temperature and with raising humidity, creating a hydrated
lipid ink on the tip. Also, a thin layer of water is always present on the substrate under ambient conditions. As
the tip approaches the substrate a meniscus condenses between the tip and the surface, creating a water vessel in
which the hydrated lipid ink is suddenly immersed. Then, more water is available, and the lipids become even
more hydrated, mainly around the polar groups, while the hydrocarbon tails will avoid the water. Hydration will
be inhomogeneous, since those lipids in direct contact with the meniscus surface have more water available.
Those fully hydrated lipids start diffusing outwards from the tip, letting water propagate inwards, thereby
increasing the concentration of lipids that become fully hydrated at the tip/meniscus interface. This mechanism
is not unlike the spreading of lipid membranes in water environments.50,51 In L-DPN the ‘source membrane’
would be the hydrated lipid ink on the tip, where lipids become progressively hydrated and start diffusing.
This picture is supported by the experiments of Lenhert et al.13 in which L-DPN under water
requires the meniscus to be formed in air, prior to immersion into water for subsequent writing. When
meniscus formation is attempted upon tip–surface contact under water, writing was not achieved. This
fact indicates that a particular hydrated lipid ink phase of the lipids within the meniscus ‘water vessel’
is formed, under the particular thermodynamic conditions. This phase is stable such that L-DPN
patterning is achieved when the tip is lifted and moved to fresh areas,13 even under water. This phase
state formation of the ink in the meniscus will strongly depend on the geometry of the meniscus,
allowing for different amounts of water. Since phases with different fluidities will be obtained for the
ink, subsequently different transport rates will be obtained, as suggested in ref. 13 and shown in the
previous section.
The phase state of the lipid species on the tip is different from that usually found in molecular ink
transport systems.11Contrary to lipids, molecular species have a melting temperature higher than room
temperature, so most of the molecular layer on the tip can be assumed to be solid under ambient
conditions. Therefore, the kinetics of molecular ink dissolution greatly influences the supply of
molecules from the tip and plays a critical role in their transfer rate.30 Parameters affecting the
chemical kinetics of the ink dissolution at the tip, as the thermal energy, greatly influence the
molecular ink supply and thus the molecular ink transport.11,22,29
However, DOPC is in the Lα liquid state at room temperature. Therefore, the lipid molecules do not
need to chemically dissolve in the same sense as molecular inks, but rather undergo a diffusive
hydration process. Therefore, in lipid inks the key parameter affecting the ink supply rate kinetics will
be hydration diffusion kinetics and not the chemical dissolution kinetics. The hydration will rely in
particular on the water provided for lipid diffusion, i.e. the meniscus size. Prior to meniscus formation,
increasing the humidity already condenses water on the tip, increasing the lipid mobility even before
the patterning process begins, i.e. the ink is rendered already more liquid-like prior to meniscus
condensation. However, only after the meniscus is condensed full hydration of lipids takes place,
mostly close to the tip/meniscus interface. Due to this it is expected that the ink delivery would follow
more likely a hydration diffusing process (see e.g. Ch. 7 in ref. 52) than an energetically dissolution
activated process.30
3.2 Step II: transport through the meniscus
The role of the meniscus in molecular ink transport during DPN has been a controversial subject. Just two years
after the first molecular transport report by Jaschke and Butt,53 Piner and Mirkin proposed that water adlayers
were transported onto the substrate surface, mediated by the meniscus that condenses as the AFM tip is held
stationary over a substrate.21 In their subsequent report they proposed that molecular inks actually flow from the
tip to the substrate by capillary action.54 From then on, different reports have shown meniscus existence,33,55 and
demonstrated a dependence of meniscus width on the tip–substrate distance and ambient humidity,56–58 observed
a minimum distance to grow a stable meniscus,43,59 and analyzed the influence of surfaces wettability60,61 and
roughness.62 Ultimately, all these parameters influence the meniscus size.
However, and especially true in the case of lipid transport, more important than the size of the
meniscus itself is how the chemical structure of the ink bonds and interacts with water, i.e. which is
the ink state in the ‘water vessel’ formed by the meniscus, and how do the ink molecules align at the
air interface at the meniscus surface. These properties finally decide the ink transport dependence on
parameters influencing the meniscus. For example, slight differences in the carboxylic acid group of
MHA and ODT make the first more soluble in water and make its transport properties RH
dependent,22,32 contrary to ODT.63However, the fact that the ink is ‘water compatible’ (i.e. soluble or
hydratable) does not necessarily imply that it does not transfer to the substrate surface unless a water
meniscus is present. For example, MHA transfers to a surface even under very low humidity
conditions, but its transport is enhanced when a water meniscus is able to condense.32 Therefore, water
soluble inks fit well with a meniscus-based transport model in the humidity regime in which the
meniscus enhances its transport, showing a rate that depends on the meniscus parameters, while
different transport modes are followed by non-water-compatible inks.44,64
Phospholipids are amphiphiles and thus their characteristics with respect to polarity fall between
those of nonpolar complexes and salts. Also, they do not dissolve in water like a salt, but rather
incorporate the water inside lipid structures. The degree of hydration determines amphiphile fluidity
and therefore diffusion, which in DPN essentially means transport to the surface. As more lipids
become hydrated in the lipid ink on the tip, lipids diffuse towards the surface driven by a difference in
lipid concentration at the tip/meniscus and meniscus/substrate interface (Fickian diffusion). As the
lipid concentration starts to increase at the meniscus/substrate interface, lipid will spread over the
substrate. When the incoming flow from the tip is faster than the spreading, a 3-D structure grows.
The lipid concentration at the meniscus/substrate interface increases until the surface spreading rate
and the flow rate gets into equilibrium. Increasing the humidity leads to a larger meniscus and a larger
amount of water available, so that the lipid ink will hydrate faster and the meniscus flow rate will
become stable earlier at high humidities, with an initial faster flow rate (see Fig. 6(b)). This would
indicate that lipid ink viscosity and/or fluidity is changing with the humidity, i.e. a liquid phase
modification with changing humidity. Within this dynamic picture, the transport entails a time
dependence based on a time-dependent difference in concentration between the tip and the
substrate, i.e. a transient regime is expected until the system, though lipids on the surface are still
spreading, gets into a dynamic equilibrium.
An interesting conclusion on this is that in order to achieve a controlled transport flow for lipid
inks, conditions that lead to a stable meniscus should be implemented. Also smaller sized menisci will
influence the ink transport less, allowing for better transport control (see e.g. the different sized error
bars of Fig. 3 in ref. 40 for the two humidities). This includes making a tip more
hydrophobic,65,66 decreasing relative humidity, or setting the tip further away from the substrate surface.
There are some reports that show that the meniscus is unstable till some time has elapsed.34 Once
condensed, menisci continue growing from vapor condensation,36 which at small distances
proceeds via Knudsen diffusion, and from the flow of the liquid film on the substrate surface;37,38 the
latter mechanism leads to the longest times of instability. The time required for the water meniscus to
become stable depends on meniscus geometry and RH (see e.g. eqn (32) in ref. 67) and therefore is
influenced by the tip size,59 tip–substrate distance43 and wettability of the tip and the substrate.60,61 In
general the smaller the meniscus size the faster is the equilibrium achieved. At large relative
humidities, condensation takes more time to create a stable meniscus.33
However, as discussed above, the main influence is not only the change in the meniscus geometry
but mainly the change in the ink state in this water meniscus governing transport. Since phospholipids
have nonpolar chains and polar headgroups, they rearrange in the meniscus keeping the nonpolar
chains away from water, maybe in the shape of little micelles, a laminar phase state68 within the
meniscus, and as a fluid (single layer) membrane exposing nonpolar groups to the air at the
air/meniscus interface. Therefore, water vapor will be hindered to enter this sheet of nonpolar tails,
suppressing meniscus growth by condensation. Additionally, when lipids start rearranging into
membrane stacks at the surface the meniscus has no longer surface water available to grow from the
substrate water layer. Therefore it is hard to picture that in L-DPN meniscus, instability arises from the
condensation of the meniscus itself. Instead, due to the lipid transport mechanism discussed above, it
is more likely that differences in lipid concentration promote differences in the transport flow until a
dynamic equilibrium regime is reached.
3.3 Step III: surface spreading
Upon transition to the substrate surface, lipid rearrangement14 and membrane self-healing50,69 are the mechanisms
for the lipid membrane assembly; additionally, lipids can transfer from multilayer membranes through
dislocation places, feeding the layers below, while the base layer that covers the surface progresses outwards
keeping a circular shape.25
In L-DPN, membrane formation is expected to follow the common phospholipid bilayer

spontaneous self-assembly in which single amphiphiles self-organize excluding some water from their
hydrophobic molecule parts leading to a decrease in water molecules order (entropy increase driving
force).70 The mechanism of membrane formation from a water containing vessel (meniscus) suggests
that in membrane formation during L-DPN some of the meniscus water can be kept, mainly
intralayers, that will provide the membrane stacks fluidity.17 In this stage the freshly formed membrane
is rather fluid, enabling the incorporation of new amphiphiles while the membrane front spreads out
over the substrate surface below the tip at the meniscus/substrate contact area. The mechanism of
bottom layer feeding from the upper bilayers has been reported by Mohamad et al.,25 showing that
lipids are transported to the bottom layer through dislocation places when humidity is increased up to
the point at which the membrane transits into the fluid phase. This accounts for the spreading behavior
reported by Lenhert et al.2 as humidity is increased above a threshold value in which membrane
fluidity would allow lipids from upper membranes to submerge into the lower layers. It resembles the
spreading of a strongly stratified smectic liquid.71
As more lipids arrive through the meniscus the concentration at the meniscus/substrate interface
increases while the surface spreads outwards (thereby reducing the concentration again). Eventually,
the net flow at the meniscus/substrate interface gets into a dynamic equilibrium, and surface spreading
follows a steady-state like kinetics. Finally, when the tip is removed the concentration is no longer
increased at the tip/surface. The growth and spreading of the membrane then stop.
Previous studies show that dry lipid membranes spread on immersion into water environments by
either bilayer sliding, monolayer–monolayer rolling or rolling of a double bilayer lope.50,51 However, in
L-DPN there is no already formed dry lipid reservoir that is immersed in water and spreads, but
instead, the membrane structure is emerging while the ink flow from the tip incorporates new
amphiphiles and surface spread. This makes a crucial difference. When a membrane is spreading
under a water environment (i.e. a system setup of water/membrane/hydrophilic substrate) the driving
energy is the formation of a bilayer–substrate contact that at stationary equilibrium equals the energy
dissipated by friction. This leads to the well-known dependence of the spreading speed on time
as t−1/2,51 even in monolayer spreading.72 More recently studies of Verma et al.include an elastic
membrane term that competes with the surface roughening upon spreading, thereby changing the
spreading kinetics.73 In L-DPN, the concentration is dynamically increased below the tip at the
meniscus/substrate interface, lipids rearrange and spread over the surface keeping fluid state. Here, the
spreading environment is the air/lipid/hydrophilic substrate. The driving force is now the flow of lipids
arriving from the tip vs. the surface spreading and so this balance governs the kinetics. Therefore, we
do not expect the spread area to depend linearly on time as in membrane spreading studies.72 Instead,
depending on the ink flow over the meniscus, different spreading kinetics will arise, either a meniscus
diffusion one or a surface diffusion controlled one. This is in agreement with the spreading in
molecular systems reported by Cho et al. showing that when tip–surface flow rates are increased in
relation to surface spreading, molecular diffusive inks can also form 3-D structures.11 In liquid inks the
3-D structures are actually far more common, because tip/surface flow rates are often much faster than
in surface spreading. For molecular inks, diffusion over gold is usually fast compared to the tip/surface
transfer rate, and only when this rate is considerably increased and surface spreading is decreased, a 3-
D growth mode is observable.11
Within this growth-spreading mechanism one may expect irregular-shaped circumferences on
circular patterns as reported by Manandhar et al.10 However, since lipid spreading and rearranging are
not anisotropic,25i.e. no preferred crystalline direction for growth is present, anisotropic cluster growth
is not expected,8 nor dendritic structures. Here, the ‘pushing’ models of the anomalous surface
diffusion9,74,75 would yield better agreement, though still not completely. Considering the intrinsic
elasticity and fluidity of the L-DPN membranes, lipid molecules freshly arriving from the tip are not
expected to enter the membrane structure and push the next neighbors towards the structure front, but
rather diffuse independently from the other molecules through the membrane. This is shown in the
membrane spreading studies of Mohamad et al.25 in which the spreading monolayer is fed without a
radial symmetry but the monolayer that covers the surface progresses outwards in a 2-D spreading
while keeping a circular shape.
While surface roughness will intrinsically be accounted for in our modelling by its influence on
surface diffusion, larger surface defects leading to uneven flow and pinning effects50,69 will not be
regarded, as L-DPN is performed typically on extensively cleaned and homogeneous substrates where
defective surfaces are usually discarded.
4. Model
As seen above, lipid inks share some of the fluidity characteristics of liquid inks and the diffusive-like behaviour
of molecular diffusive inks. Coincidently, following the classification of ref. 76 concerning their molecular
weights (786.11 Da for DOPC) lipid inks would also settle between liquid inks and diffusive inks.
Due to lipids’ diffusive properties, the ink dissolution process at the tip is expected to follow a
hydration–diffusion process, rather than a dissolution activated one. Ink flow over the meniscus
depends on the meniscus volume but also on ink diffusivity; additionally, it changes with time due to
the balance between ink flow and surface spreading until a dynamic equilibrium between both is
achieved. When lipids arrive at the substrate surface they spread and assemble. They form 3-D piled
structures as in liquid inks or in molecular systems in which the tip–substrate flow rate is considerably
faster than the surface spreading.11Due to the lipids’ fluidity, that allows their spreading along the
surface,50,72 a non-correlated circular spreading as shown in ref. 25 is expected. Surface spreading will –
after a certain RH dependent dwell time – come into dynamic equilibrium with the tip–substrate flow
rate, with a time dependence tα where the growth rate coefficient will be α < 1.11 In this context two
models may describe the ink transport. The first is based on transport of diffusive inks, but has to be
modified to incorporate the fluid character of lipid inks: the ink hydration and diffusion at the tip, the
3-D growth character, and the fact that tip–substrate flow is influenced by differences in
concentration, leading to the emergence of two spreading regimes with time. The second would be
based on the spreading of a liquid droplet over a surface.
Here, modification would be needed to include the diffusive character of the lipid ink: the kinetic
energy arising from the ink flow from the tip as well as its time dependence. In the present work we
will deal with and develop the first one.
Saha and Culpepper developed a general model for the molecular ink transport in line-writing that
encompass a description of each transport stage described above. The advantage over other models for
diffusive inks is that – as it was developed for line writing – it incorporates the influence of changes in
concentration at the meniscus/substrate interface on the ink flow. We will therefore use this model as a
starting point for L-DPN, modifying some of the stages to reflect the unique lipid ink characteristics.
A representative scheme of this model is shown in Fig. 11.
Fig. 11 Scheme of the ink transport showing three stages: (I) ink-dissolution into the meniscus at the tip/meniscus interfa
forward flow D0c0, where D0 accounts for the diffusivity of the lipids with a concentration at the tip surface c0, and a back
where Ctm = Ctip/meniscus is the concentration at the tip/meniscus interface and β represents the impingement and attachme
flow J transport via a meniscus of size 2R and height L to the meniscus/substrate with an ink diffusivity D1, as a 1-D Fickia
differences in concentrations Ctip/meniscus and the concentration at the meniscus/substrate interface Cmeniscus/substrate; (III) spre
creates a 3-D circular feature with density ρ, width w and height h.
With the definitions of Fig. 11 and based on the previous experimental results, L-DPN transport
encompasses the following stages.
4.1 Step I: ink dissolution at the tip/meniscus interface
In molecular inks, molecular species dissolve from a solid state at the tip through thermal activated detachment
(first-order chemical reaction), creating a forward rate Nα, where α is the ink solubility,
α = γe−Ed/kBT,
and N is the number of ink molecules at the tip.29,30 For lipid inks, the hydrated liquid ink at the tip is immersed in
the condensed meniscus and lipids become progressively fully hydrated and start diffusing outwards from the
tip. As previously discussed, it is expected that ink delivery would therefore follow a hydration diffusing process
(rather than an energetically dissolution activated process), providing a forward rate D0c0, where c0 is the
concentration of lipid ink molecules at the tip (areal ink concentration‡), and D0 is a diffusion coefficient of the
hydrated lipid ink at the tip into the meniscus. Approximate expressions can be found for the diffusion
coefficient,
assuming a Stokes–Einstein diffusion approximation, where vH2O is the partial molar volume of water (in the
water/lipid mixture),C0 is the concentration of lipid ink molecules at the tip (volumetric ink concentration), μ is
the water viscosity, r0 is the lipid radius of gyration, and n is the number of water molecules bound to the single
lipid amphiphile.52 However, since the exact diffusion kinetics of lipid inks are not known we will use a simple
forward rate D0c0. Note that due to the progressive hydration of lipids at the tip, it cannot be discarded
that D0c0 may slightly increase with time, becoming faster hydrated the larger the RH is. However, as a first
approximation, we assume it to be constant. Lipid molecules return to the tip due to impingement and
attachment at a rate Ctip/meniscusβ, where β is the detachment rate,
assuming a gas kinetic expression for the flux, following Weeks et al.30 It should be noted that the rate of
forward reaction depends on the volumetric ink concentration at the tip/meniscus interface Ctip/meniscus. The net ink
flow from the tip is then
J = D0c0 − Ctip/meniscusβ.
4.2 Step II: transport through the meniscus
As we have discussed in previous sections, the ink flow is a concentration-driven diffusion of the ink from the
tip towards the substrate surface, mediated by the meniscus. We assume that diffusion is one-dimensional, with
a concentration gradient along the cone. Then a Fickian transport,

follows, where D1 represents the ink diffusivity in the meniscus transport. The term L/(D1πRr0) represents the
ink resistivity to flow, similar to a current that would flow through an electrical resistance,§ with the shape of a
truncated cone geometry of sections πR2 and πr02, length L and resistivity 1/D1.
4.3 Step III: surface spreading
The flow J of molecules will first start spreading as a bottom layer and at some point additional layers will start
growing over the first. Amphiphiles can be swallowed by layers below, finally feeding the bottom layer. This is
more likely to take place just below the tip, where higher humidity enhances lipid fluidity and creates
dislocation sites. Accordingly, the flow J will be distributed among the different layers. It is difficult to
quantitatively account for how J is distributed into the different layers, how much of the flow is dislocated from
upper to lower layers, and how much is employed by each layer for radially spreading over the bottom layer.
However, it can be concluded, by considering conservation of mass that after a time t has elapsed and a
flow J has been provided, that a feature of volume V ≈ ρSh/2 has grown,¶ where the flow
J = ρV/t ≈ ρSh/2t.
At the same time, an area of size S has been created, due to the diffusive spread of lipids over the
surface. Keeping in mind the constant volume, the creation of a higher dot feature would correspond
to a smaller area. Inversely, a larger surface spread Scorrelates with lower features. Following the
same idea, for a bottom layer of volume Vbottom ≈ ρSzt/2, where zt is the layer thickness, a
flow Jbottom = Vbottom/t ≈ ρSzt/2t is needed. A flow creating a two-dimensional spread in a symmetrically
radial growth generates an area of size S = πw2/4, where
for dot writing,77 where DS is related with the surface diffusion coefficient, and R is the radius of the source that
creates the spreading with a concentration denoted as Cmeniscus/substrate. As mentioned above, it is assumed that inter-
membrane dislocation of the lipids proceeds mainly under the tip where the meniscus enhances fluidity, so
that R can be assumed to be approximately the meniscus size, 2R ∼ 200 nm (see Fig. 8). After trivial
manipulation follows
Two approximations are usually employed in the analysis of surface spreading in dot writing with
molecular inks: (I) a constant flow approach28 or (II) a constant concentration approach.44,77 At short
dwell times, as a first approximation, a constant flow can be assumed.28,78 Following this regime,
behaviour as either by molecular diffusive inks with a substrate spreading rate large compared to the
tip–surface ink flow rate11 or by membrane spreading in water environments in which the driving
energy is the bilayer–substrate interaction50,51,72 is assumed, i.e. a system in which the surface spread rate
is faster than the ink flow rate. In these systems, at short dwell times, feature radius increases as t1/2,
with a slope dependent on the balance between the flow rate and the surface diffusion rate.28,30 On a
larger time scale, ink transport is better described by the approximation of constant concentration.
Now the balance between the ink flow and diffusion rates over time11,30 leads to the radius increasing
as tα, where α < 0.5.47,77 For L-DPN, as shown in Fig. 6, a constant flow rate cannot be assumed.
Nevertheless, the conditions of the constant concentration approach are not completely fulfilled. Using
the expression introduced previously, the concentration at the meniscus/substrate can be calculated as
Experimental data calculated with this equation are shown in Fig. 12. Cmeniscus/substrate increases
strongly with dwell time up to a critical value above which the concentration stays roughly constant
(at larger humidities with a stable J), or at least increases much more slowly (at smaller humidities,
where the flow rate does not achieve a dynamic equilibrium). Comparing Fig. 6 and 12reveals that
changing the humidity strongly influences flow while the concentration at the meniscus/substrate
interface is not that strongly influenced. This is in agreement with the fact that humidity enhances lipid
fluidity, and therefore the flow rate, surface diffusion and spreading, as previously discussed (section
2.1).
Fig. 12 The concentration at the meniscus/substrate interface, at humidities of 39.3 ± 0.2% (■), 37.8 ± 0.1% ( ), 35.9 ± 0.
32.6 ± 0.1% ( ), 30.6 ± 0.3% ( ) and 28.9 ± 0.2% ( ), increases with dwell time for all humidities up to 4 s; for lower hum
reached. Solid lines are a guide to the eye.
An elegant way of incorporating this meniscus/substrate concentration dependence on flow was

proposed in the model of Saha and Culpepper,12 relating the meniscus flow rate J with Cmeniscus/substrate as
in section 4.2. In regard to the lipid spreading, a central source is assumed from the fact that lipids
spread over the surface with a flow that depends on the difference in concentration between
tip/meniscus and meniscus/substrate interfaces.
Combination and rearrangement of these expressions lead to eqn (11) that can be used to fit the dot
feature area in L-DPN.
The square bracket of the first term on the right side of eqn (11) describes the spreading on the
surface relative to the backward ink flow at the tip, representing the surface spreading contribution.
The square bracket of the second term is the meniscus diffusion relative to the backward ink flow,
which represents the meniscus transport. When surface diffusivity is high compared to the meniscus
difussion D1πRr0/L, the second term dominates over the first and therefore flow is limited by meniscus
diffusion. At large meniscus diffusion, the first term dominates over the second, and the flow is
limited by surface diffusion. The overall transport is limited by the slowest mechanism.
The flow involved in each diffusion mechanism can be calculated as
excluding proportional factors. They are shown in Fig. 13.
Fig. 13 Flow contributions by (a) surface diffusion and (b) meniscus diffusion, at humidities of 39.3 ± 0.2% (■), 37.8 ± 0.1%
± 0.2% ( ), 32.6 ± 0.1% ( ), 30.6 ± 0.3% ( ) and 28.9 ± 0.2% ( ), show the same dynamic dependence as the total ink flo
Solid lines are a guide to the eye.
Interestingly, both contributions to the total flow show the same dynamics, and are in agreement
with the time dependence of the total flow, shown in Fig. 6(b). This indicates that the same factor is
controlling surface spreading, governed by diffusion over the surface, and meniscus dynamics, driven
by the liquid ink tip–substrate flow: this factor being the ink fluidity. Though one may expect that
meniscus flow would be controlled by the ink fluidity it is surprising that both meniscus flow and
surface spreading are, within their respective contributions, dynamically evolving in similar functional
shape. Yet these results confirm the notions of section 2 that ink mobility and surface mobility were
similarly affected by humidity, thereby keeping a surface spreading growth rate that depends only
weakly on humidity, as discussed in Fig. 4. This confirms that the ink fluidity properties, modulated
by RH, completely control the transport and therefore the feature growth. This further indicates that,
as suggested in section 2.2, RH controls the viscosity or phase state of the lipid ink.
A fit of the experimental data for dot feature area versus dwell time with eqn (11) is shown in Fig.
14. For the fit, the meniscus size at the substrate surface has been fixed at the value determined in Fig.
8 (2R = 200 nm) and at the tip surface to tip size (r0 = 40 nm); monolayer height is assumed to be zt =
1.3 nm.18 The fit shown in Fig. 14 assumes the following parameters to be RH independent: setting β =
10−6 μm3 s−1, L = 160 nm,|| and diffusion coefficients for the surface DS = 11 μm2 s−1 and ink D1 = 11
μm2s−1, as reported for DOPC lipid diffusion.24 The dependence of the ink flow D0c0/ρ on RH
resembles the observed flow dependence on humidity shown in Fig. 6(a). It should be emphasized that
actually this parameter, related to ink diffusivity, depends strongly on RH, again suggesting that
feature growth dynamics is controlled by the ink fluidity, as previously discussed. While the fit
describes the experimental data well for longer dwell times, deviations occur at short dwell times.
Letting D0c0increase with time improves the fit (see the ESI†). However, this introduces an additional
parameter into the fit. This indicates that this difference between the model and the experimental data
at short dwell times can be associated with a progressive hydration of the lipids at the tip/meniscus, in
agreement with the qualitative picture of the ink supply drawn in section 3.1.
Fig. 14 Dot area as a function of the dwell time at humidities of 39.3 ± 0.2% (■), 37.8 ± 0.1% ( ), 35.9 ± 0.1% ( ), 34.5 ± 0
30.6 ± 0.3% ( ) and 28.9 ± 0.2% ( ). Solid lines are a fit to eqn (11). Inset: D0c0/ρ depends on RH resembling the behavio
The existence of the two regimes, surface controlled and meniscus controlled, can be seen as
follows. eqn (11) describes the growth of a 3D feature, including height and surface growth. It can be
rearranged to show only surface spreading as
The left side of the equation is the measured height divided by the related dwell time h/t multiplied
by time independent factors. The right side of the equation now shows the spreading related to the
surface diffusion (first term) and to the meniscus diffusion (second term). Meniscus diffusion is
directly proportional to w2, while surface diffusion grows as w2 ln(w/2R); the surface area growth is a
combination of both. Fig. 15 shows the dependence of the left side as a function of w2. For high
humidities a dependence proportional to w2 ln(w/2R) is visible in the experimental data (solid lines
represent a dependence of the type w2 ln(w/2R) and dashed lines a dependence of the type w2). This
indicates that meniscus diffusion is so large that the total area growth is controlled by the slower
mechanism: surface diffusion. However, as humidity decreases the meniscus flow becomes
comparable to surface diffusion. At 34.5% ± 0.2% and short dwell times, the dependence of the
spread on w2 is observable: here, area growth is controlled by the slower meniscus flow. Due to the
time dependence of the flow (see Fig. 6(b)), the smaller the humidity, the longer the dwell time for
which this regime change is visible.
Fig. 15 t/h as a function of the area at humidities of 39.3 ± 0.2% (■), 37.8 ± 0.1% ( ), 35.9 ± 0.1% ( ), 34.5 ± 0.2% ( ), 32
) and 28.9 ± 0.2% ( ). Solid lines correspond to a growth ∝ w2 ln(w/2R) and dashed lines to ∝ w2.
5. Conclusions
Our results provide a deeper understanding of ink transport in L-DPN, thereby helping in making an informed
choice of experiment conditions to control L-DPN features. The procedure employed can be used in the analysis
of any DPN ink transport as it provides useful information concerning the ink state within the transport, its
transport mechanism, the influence of substrate surfaces and so on. Ultimately, this can explain quantitatively
how experimental parameters influence ink transport, indicating those parameters that govern the feature size
and shape.
For L-DPN we showed that the spreading rate depends on the balance between the substrate
spreading and the meniscus flow rate. In molecular inks, as well as in membrane spreading, the
surface area spread is S ∝ t due to a high surface spreading in comparison with the meniscus flow rate.
In L-DPN, the ink transport rate is controlled by relative humidity and by differences in lipid
concentration at the tip/meniscus and meniscus/substrate interfaces. Both properties are relevant and a
balance between them determines which regime is followed: at low dwell times and humidities L-
DPN is controlled by meniscus diffusion, while at long dwell times and humidities, i.e. at large ink
flows, surface diffusion controls the transport.
Control of the L-DPN feature size requires control of the ink flow. This can be achieved by
controlling the meniscus, either with RH (altering ink diffusion) or the meniscus geometry itself. For
the latter possibility, the smaller the meniscus the smaller the ink flow will be and hence the better the
control achieved. This suggests the use of more hydrophobic tips or larger tip–substrate distances, as
control of meniscus height is also critical.
Feature shape is mainly determined by the surface. In order to control the spreading of the
feature, i.e. the height of the feature as compared to the feature area, the surface has to be properly
selected or modified to achieve the desired hydrophilicity.
An adjusted model based on the transport of diffusive inks reproduces the L-DPN ink transport
properties. It takes into account the fluidity of the system, its 3-D growth and influences from the
surface. The model shows the existence of a surface kinetics and a meniscus diffusion regime, which
is observed in the experiment.
6. Experimental
6.1 Materials
All the phospholipids employed in our experiments were obtained dissolved in chloroform from Avanti Polar
Lipids, USA and used as received. A 20 mg mL−1 (25.4 mM) solution of 1,2-dioleoyl-sn-glycero-3-
phosphocholine (DOPC) was admixed with 1 mol% of the fluorescently labelled phospholipid 1,2-dioleoyl-sn-
glycero-3-phosphoethanolamine-N-(lissamine rhodamine B sulfonyl) (Liss Rhod PE).
6.2 Methods
DPN tip coating and patterning were carried out by using a commercial DPN system equipped with an
environmental chamber (DPN 5000, NanoInk Inc.; relative humidity tolerance ±0.5%78). Relative humidity was
additionally monitored with an external digital hygrometer positioned close to the sample stage, and showed
fluctuations in humidity that did not exceed 0.3%. Tip arrays of type F and microfluidic ink chips of type F1-2
(Advanced Creative Solutions Technology, USA) were used. Tips were coated in a one-dip process, thus
avoiding multi-dip processes that may load the tips with different amounts of lipids. The reader tips (most left
and right tips in the 1D array) were not inked, so that they can be used for sensing cantilever deflection with the
system's laser beam during the approach of the array to the substrate surface prior to patterning. More
information about the inking procedure is included in the ESI.†
Before each experiment, glass substrates were carefully cleaned by ultrasonication, and submerged
for 10 min in chloroform, then for an additional 10 min in isopropanol and finally for 10 min in
ultrapure water. After sonication, the substrates were dried under a nitrogen flow.
The tip array was first leveled with respect to the substrate to align all pens at the same distance to
the surface. It should be noted, however, that the tips’ spring constants vary with a standard deviation
as large as 2% and thus there is some variability in contact force exerted by each tip. The array was
aligned placing the tips in contact with the substrate and then tilting the array slightly until pen
deflection was uniform all along the length of the array. By adjusting the plane of the tip array with
that of the substrate, the two planes subsequently can be aligned to better than 0.1° (the length of one
tip array is about 1.5 mm and the systems’ optical microscope offers an accuracy of about ±5 μm in
detecting the tip contact with the surface).79
In order to drive the tip array close to the substrate surface prior to patterning, two procedures can
be followed. Usually, the systems’ optical microscope is employed in DPN experiments with tip
arrays to observe the change in tip color that accompanies tip bending. This acts as an indicator of the
amount of pressure applied to each tip upon surface contact. In the second procedure, the built-in laser
of the atomic force microscopy (AFM) setup is used to control the approach to the substrate surface.
Striking differences have been found in lipid transport depending on the tip–substrate approach, which
can be explained by the fact that in the laser deflection procedure, the tip–surface distance is much
more carefully controlled ensuring that the tip is further away from the substrate (see the ESI† to find
more details about the approach). As explained in the ESI,†using the optical approach, the tip gets
usually closer to the substrate. Here, often a slight slash is seen attached to the written dot features,
while dots created using the laser deflection are completely circular in shape. In the present work we
employed the approach based on the built-in laser AFM setup, ensuring a highly controlled surface–
tip distance, thus obtaining reproducible data to be modeled.
6.3 Patterning
The patterning was performed at 26 °C and at relative humidities of 39.3 ± 0.2%, 37.8 ± 0.1%, 35.9 ± 0.1%,
34.5 ± 0.2%, 32.6 ± 0.1%, 30.6 ± 0.3% and 28.9 ± 0.2%. Writing was done starting with high humidity and then
later decreasing it step by step, which can influence patterning, as the meniscus size dependence on relative
humidity shows hysteresis. Decreasing RH was selected for our experiments since as the humidity decreases,
the water meniscus size decreases linearly while, in contrast, size increases exponentially with increasing
humidity.33 Thus it is expected that meniscus size would be more stable and the dependence will be more
homogeneous with decreasing RH. Additionally, large humidities establish a more uniform coating on the AFM
tip,32prior to meniscus formation, so that surface–area ink coverage on the different tips is expected to be more
similar, thereby minimizing the influence of different tip ink arrangements on the transport rate.29 After changing
the RH inside the environmental chamber, the system was allowed to equilibrate for three minutes before
starting the patterning.
Arrays of dots were fabricated by bringing the inked cantilever into contact with the freshly
cleaned glass surface for defined dwell times. To ensure an accurate dwell time control, the tip array
approach and the retraction speed was set to the highest available speed (20 μm s−1). This avoids the
formation of a meniscus already during the approach and the corresponding transfer of ink before the
accounted dwell time has even started. Starting with higher contact times, dot features with 10 s, 8 s, 4
s, 2 s, 1 s, and 0.5 s dwell times were written. In order to obtain a good statistical representation four
features were written for each dwell time. The overall sample pattern consists therefore of an array of
rows with 4 dots written at 6 dwell times, and repeated for each of the 7 humidities.
The data represented in the main article were taken from one tip, as inter-tip variance in L-DPN
still makes it hard to directly model all tips with the same set of parameters. This is mainly due to the
dynamic dependence of the flow itself, as discussed in the text, which is not just a simple
proportionality factor. J(t) depends additionally on the differences in concentration between
tip/meniscus and meniscus/substrate. Therefore, only one tip can be studied in all the range of RH and
dwell time in order to analyze a coherent ink transport with one set of parameters. Key figures
corresponding to three additional tips of the same tip array are shown in the ESI.†
6.4 Structural characterization
For quality control, DPN patterns were first checked by using fluorescence microscopy (Eclipse 80i, Nikon)
with a 50× objective (NA = 0.8). For this, a fluorescent probe (Liss Rhod PE) was admixed to the DOPC, as
described above, so that fluorescent intensity level analysis could provide information about the structures area
and height.19 The areas and fluorescent intensity levels of the dot features were analyzed by using the software
package ImageJ.80
After obtaining the fluorescence microscopy images, the features were measured by AFM. Images
were obtained on a Dimension Icon (Bruker, Germany) and on an Asylum Research MFP-3D-BIO
AFM system. Measurements were carried out in air under ambient conditions in tapping mode with
NSC15/AlBS silicon cantilevers (MikroMasch) with 325 kHz nominal resonance frequency. Image
processing was carried out with the software WSxM81 for the Dimension Icon AFM data, and with the
MFP3D built in software for its corresponding data. AFM data of one set of features were analyzed by
using both software processing packages (WSxM and MFP3D), in order to verify that no artifact is
introduced by the data treatment software.
While fluorescence microscopy is much less time consuming than AFM, the data extracted from
the latter is of higher resolution and more reliable (see the ESI†). The modelling and quantitative
analysis in the following is therefore based on the AFM data. The area, maximum height and volume
were measured independently by the AFM, meaning that volume was not calculated using area and
height but directly extracted from the measurements by means of the respective methods offered by
the analysis software.
Acknowledgements
A.U. acknowledges the People Programme (Marie Curie Actions) of the European Union's Seventh Framework
Programme FP7/2007–2013 under REA grant agreement no. 328163. This work was carried out with the
support of the Karlsruhe Nano Micro Facility (KNMF, http://www.knmf.kit.edu), a Helmholtz Research
Infrastructure at the Karlsruhe Institute of Technology (KIT, http://www.kit.edu). The authors thank Anna
Ovvyan for SEM imaging of the cantilever arrays and Prof. Harald Fuchs for stimulating discussions and a
critical reading of the manuscript.

Thermal Dip Pen Nanolithography For Direct Writing of Polymer

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Thermal Dip Pen Nanolithography For Direct Writing of Polymer

Caricato da

Copyright:

Formati disponibili

Thermal Dip Pen Nanolithography for Direct Writing of Polymer-NanoParticle Composites

A polymer-nanoparticle composite is deposited from a heated cantilever tip. The

without solution processing.

Classic, liquid, and matrix-assisted dip-pen nanolithography for materials research

Jian Zhong,*a Gang Sun*b and Dannong Heac

Discuss this application

Publication number US20100071098 A1

Publication type Application

Application number US 12/465,616

Publication date 18 Mar 2010

Filing date 13 May 2009

Priority date 13 May 2008

Also published as EP2291849A2, 20 More »

Inventors Chad A. Mirkin, 9 More »

Original Assignee Northwestern University, Nanolnk, Inc.

Export Citation BiBTeX, EndNote, RefMan

External Links: USPTO, USPTO Assignment, Espacenet

Scanning probe epitaxy

1. A dual tip micro probe, comprising:

a micro cantilever arm comprising first and second arm ends;

2. A dual tip micro probe, comprising:

a micro cantilever arm comprising first and second arm ends;

a strain gauge disposed along a length of the cantilever arm.

CROSS-REFERENCE TO RELATED APPLICATION

STATEMENT OF GOVERNMENTAL INTEREST

1. Field of the Disclosure

2. Brief Description of Related Technology

BRIEF DESCRIPTION OF THE DRAWINGS

Dual Tip Probe Structure

Typical Synthesis Conditions for Nanostructures

Evaporator Synthesis Tip

Thermal Catalyst-Tipped Synthesis Tip

wherein R(T) is the resistant at temperature T, R0 is the resistance at a reference temperature

Nanostructures formed in Nanoreactor Wells

Method of Making The Dual Tip Probe

Software Platform for Pulsed Evaporation Synthesis

Flow Description (Between Two Successive Feedback-Ons)

feedback loop maintains a constant set point value.

where it waits for an OK signal form process 1.

position (go there).

Example 1Method of Forming a Dual Tip Probe

Example 2Method of Forming a Dual Tip Probe

Dimension of Dual Tip Probes Formed Using

the Method In Accordance with the

Example 3Method of Making a Thermal Probe

Example 4Force-Distance Curve Calibration of Dual Tip Probe

Example 5Imaging Capabilities of the Dual Tip Probe

Example 6Formation of Nanowires and Carbon Nanotubes using Field-Induced Evaporation

Example 6Nanoreactor Well Formation

US5266801 * 26 Jan 1993 30 Nov 1993 Digital Instruments, Jumping probe

Method and apparatus

Method and apparatus

Patterning of solid state

Dual tip atomic force

The United States Of High resolution coherent

Apparatus for recording

High speed/high density

Method and apparatus

Storage device having a

Scanning probe for data

Citing Patent Filing date Publication date Applicant Title

Data provided by IFI CLAIMS Patent Services

A diffusive ink transport model for lipid dip-

Jian Zhong,a Gang Sunb and Dannong Heac