AFM Bias Induced Electrochemistry: Redox Processes at the Solid Liquid Interface by Jim Mara B.Eng, CA, CTA, PDA The thesis is submitted to University College Dublin in part fulfilment of the requirements for the degree of Master’s (MSc) NanoBio Science School of Physics UCD Conway Institute of Biomolecular and Biomedical Research Head of School: Prof. Padraig Dunne Principal Supervisor: Dr. Brian Rodriguez
61
Embed
Jim Mara MSc NanoBio Science Thesis AFM Bias Induced Electrochemistry Redox Processes at the Solid Liquid Interface
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
AFM Bias Induced Electrochemistry:
Redox Processes at the Solid Liquid Interface
by
Jim Mara B.Eng, CA, CTA, PDA
The thesis is submitted to University College Dublin
in part fulfilment of the requirements for the degree of
Master’s (MSc) NanoBio Science
School of Physics
UCD Conway Institute of
Biomolecular and Biomedical Research
Head of School: Prof. Padraig Dunne
Principal Supervisor: Dr. Brian Rodriguez
ii
Table of Contents 1 Abstract ......................................................................... 3
Wicking Paper- Filter paper ‘Whatman’ Circular (90mm)
5.1 Highly Oriented Pyrolytic Graphite (HOPG)
We conducted the vast majority (81%) of our AFM electrochemical deposition
experiments on HOPG. HOPG is an allotrope of carbon and part of the graphite
family.
"Usual" graphite, especially natural one, exhibits quite imperfect structure due to an
abundance of defects and inclusions. A number of technologies have developed for
the preparation of perfect graphite samples to take advantage of its unique structure.
Of these, pyrolysis of organic compounds is the most common and effective. Pyrolytic
graphite is a graphite material with a high degree of preferred crystallographic
orientation of the c-axes perpendicular to the surface of the substrate.
It is manufactured by graphitization heat treatment of pyrolytic carbon or by
chemical vapour deposition at temperatures above 2500K.16
Page 18 of 61
Hot working of pyrolytic graphite by annealing under compressive stress at
approximately 3300K results in HOPG. Thus HOPG is a highly-ordered form of high-
purity pyrolytic graphite (impurity level is of the order of 10 ppm ash or better).16
HOPG is characterized by the highest degree of three-dimensional ordering. The
density, parameters of the crystal lattice, preferable orientation in a plane (0001) and
anisotropy of the physical properties of the HOPG are close to those for natural
graphite mineral. In particular, like mica, HOPG belongs to lamellar materials
because its crystal structure is characterized by an arrangement of carbon atoms in
stacked parallel layers – this two-dimensional and single-atom thick form of carbon
is called graphene. Graphite structure can be described as an alternate succession of
these identical staked planes. Carbon atoms within a single plane interact far more
strongly than with those from adjacent planes- this explains characteristic cleaving
behaviour of graphite.16
Furthermore, graphene is a planar, hexagonal arrangement of carbon atoms. The
lattice of graphene consists of two equivalent interpenetrating triangular carbon sub-
lattices A and B, see Figure 5.b (below). Each one contains one half of the carbon
atoms. Each atom within a single plane has three nearest neighbours: the sites of one
sub-lattice (A – marked by red) are at the centres of triangles defined by three nearest
neighbours of the other one (B – marked by blue).16
The lattice of graphene thus has two carbon atoms, designated A and B, per unit cell,
and is invariant under 120° rotation around any lattice site. The network of carbon
atoms is connected by the shortest bonds in a honeycomb like shape. However in
bulk HOPG, even in bi-layer graphene, A- and B-sites carbon atoms become in-
equivalent (including those on the surface): two coupled hexagonal lattices on the
neighbour graphene sheets are arranged according to Bernal ABAB stacking, where
every A-type atom in the upper (surface) layer is located directly above an A-type
atom in the adjacent lower layer, whereas B-type atoms do not lie directly below or
above an atom in the other layer, but sit over a void – a centre of a hexagon. Figure
5.b illustrates the assumed non-equivalent types of carbon atoms.16
Page 19 of 61
Thus in each layer the atoms form a grid of correct hexagons with distances between
atoms equal 0.1415 nm. The distance between layers is equal 0.3354 nm which results
in a theoretical calculated value of density ρ = 2.265 g/cm3.16
HOPG terminated with graphene layer is an excellent tool for using in scanning probe
microscopy as a substrate- this is an easily renewable material with an extremely
smooth surface. This is vital for any SPM measurements that require uniform, flat,
and clean substrates, for samples where elemental analysis is to be done.
Figure 5.b
Schematic representation of the structure of the bulk hexagonal graphite crystal. The
dashed lines show the axes of bulk unit cell. Side insets: top view of the basal plane
of graphite and schematic representation of the surface structure (carbon atoms) of
graphite most viewed by SPM, where every other atom is enhanced (right-side inset)
and viewed under ideal conditions, where every single atom is seen (left-side inset).16
Page 20 of 61
Similar to mica, HOPG specimens are layered polycrystals. Each bulk polycrystal
looks like mosaic of microscopic mono-crystal grains of different sizes. The structure
is columnar, the columns run vertically within the flat slab of the material, and the
grain boundaries can be seen on the lateral surfaces. The grains are slightly
disoriented with respect to each other. An angular spread of the c-axes of the
crystallites is of the order of 1 degree. The surface of specimen consists of many
randomly placed steps – result of the cleaving process: single atomic steps and steps
of several or dozens of atomic layers.16
Although the heights of multilayer hills and valleys are not calibrated, single steps
have the well defined height of 0.34 nm. To characterize the angle of deviation of the
grain's boundaries from the perpendicular axis of the columnar structure, a measure
of the parallelism of grains – perfectness of HOPG samples, a "mosaic spread" term is
used. The lower the mosaic spread, the more highly ordered the HOPG. The term
originates from X-ray crystallography.16
The disordering results in broadening of the (002) diffraction peak: the more
disordering, the wider the peak. Therefore, ‘perfectness’ of HOPG can be easily
related to a Full Width at Half Maximum (FWHM) of the Cu-Ka rocking curve
(radiation peak) measured in degrees – "mosaic spread angle". Thus, the smaller this
angle, the higher the quality of HOPG. The size of grains also varies with the mosaic
spread. The lower mosaic spread results in a freshly cleaved surface that exhibits the
smaller number of the steps due to the bigger size of grains. The higher the quality is-
the less the roughness of the surface. The lower level grade material is also more
"cleavable" allowing the bigger number of cleavings per sample.
All the other physical characteristics of graphite, including atom-to-atom distance,
that is an atomic property of carbon, are independent of its grade and remain the
same for all types of HOPG. Due to the anisotropic nature of HOPG such
characteristics as thermal conductivity and electrical resistivity are different in
different directions: along the basal plane and along the principal axis c
(perpendicular to the basal plane).16
Page 21 of 61
HOPG is a highly stable material. It remains stable at the temperatures up to 500°C
in air and up to two-three thousand degrees Celsius in a vacuum or inert
environment. It exhibits high chemical inertness to just about everything.16
Zoval, Jim V., et al also report that the graphite basal plane surface is
electrochemically ‘very inert’17. Furthermore the same group go on to outline how
silver nanocrystallites interact weakly with the graphite surface and are removed by
the sweeping action of the AFM probe tip from the imaging area.17 This effect has
been previously documented for gold particles on graphite by Schaefer et al.18
Silver micro and nanocrystallites which nucleate at defect sites are observed by STM
and AFM, and, it has been concluded that silver overpotential Deposition (OPD) on
graphite is initiated by nucleation exclusively at defects, such as step edges, on the
graphite surface. Further they report that the NC-AFM data presented for low-defect
density surfaces such as the graphite basal plane, STM and repulsive-mode AFM data
can provide a misleading view of nucleation by “ignoring” the presence of weakly
adsorbed metal nanocrystallites which are not associated with defects.17
It is worth bearing in mind that these researchers experiments were conducted within
the following parameters: voltage pulses having amplitudes of 100, 250, and 500 mV
vs. Ag0 and durations of 10 or 50 ms were applied to graphite surfaces immersed in
dilute (≈1.0 mM) aqueous silver nitrate. Also, the potentiostatic deposition of silver
was accomplished by using a silver wire reference electrode immersed directly in the
silver plating solution. In contrast our experiments were conducted in a meniscus of
AgNO3 using conductive AFM tips at higher voltage and longer durations (see Results
7 below for details).17
Page 22 of 61
Figure 5.c (below) demonstrates Zoval, Jim V., et al’s experimental set up.
Figure 5.c
Schematic diagram of the instrument employed by Zoval, Jim V., et al for pulsed
potentiostatic deposition of silver nanocrystallites.17
Zoval, Jim V., et al go on to note that the capacitance of a graphite basal plane surface
was 1.70 µF cm-2, which is in the normal range. Further in successive silver deposition
trials in which the graphite surface was cleaved prior to each experiment, the
apparent capacitance of the surface fluctuated by 10-20%, presumably due to
fluctuations in the defectiveness of the graphite surface which is exposed during
cleavage.17 This observation can have important implications for the consistency of
results in our experiments.
Further they highlight ‘isolated silver nuclei were never observed on atomically
smooth regions of the graphite surface in any of these experiments’.
Page 23 of 61
5.2 Silver Nitrate
It is worth remembering that humans have been experimenting with depositing silver
ions for generations and it is thus an ideal system to be studied on the nanoscale
using AFM electrochemical techniques. Indeed any macroscopic deposition must, at
least transiently, by obtained by nanoscopic deposition en route to the macro scale.17
Digital photography now replaces most of the chemical and physical applications that
chemical photography employed over its relatively short history of about 170 years.
The colour and form of a digital image after capture with a camera can be
manipulated through Photoshop, allowing a photographer an exacting control over
the finished image, with little if nothing left to chance and with the ability to
reproduce numerous identical copies.
In the early development of chemical photography, individual images were more
unique in their nature, as a silver image produces something that is far more difficult
to control and exhibits a more random nature from chemical reaction. This is
particularly the case when the print maker is controlling the chemistry and exposure.
Early photographic chemistry can exhibit enormous variations in reaction to colour,
sensitivity and stability, making the visual outcome unique and often unrepeatable.
By experimenting with some of these more antiquated processes, it is possible to
produce these on paper without the need for a suspension medium such as gelatine.
Thus, early efforts revolved around the use of chemistry as a printmaking medium
and together with others, such as ink and graphite. The development of digital
photography has therefore helped establish this chemical art form in its rightful place
within the context of printmaking; see Figure 5.d (below) for an early example of
early silver ion chemical photography.
Page 24 of 61
Figure 5.d
Early print making using silver ions.
To explore the fascinating possibilities, it is necessary to look back at the dawn of
photography and some of the experiments made by practitioners working at that
time. Robert Hunt who experimented with light sensitive substances, in his
“Researches on Light in its chemical reactions” published in 1844, points out many
interesting reactions of substances with light, including charcoal:
“If a stick of charcoal is placed in a bottle in which is some solution of nitrate of
silver, so that one half of the charcoal is in the solution, and the other half above it,
there will in a short time appear little spangles of silver upon the upper portion of
the charcoal, if it is exposed to diffused light. In full sunshine the effect is greatly
retarded. If the bottle is placed in a dimly illuminated place, there will in the course
of a few weeks, form in the solution around the charcoal, a series of the most
delicate thread-like crystallisations of the silver. After these have formed, if the
bottle is exposed to sunshine they are gradually re-dissolved into the fluid.”
Page 25 of 61
There is still much to be learnt by looking in depth at the work of these experimenters
and reactions they dismissed at that time, which did not follow the objective of
achieving a practical working photographic process. As these may now prove useful
when employed in combination with 21st century technology, such as scanning and
flash photography, to record precise moments of chemical reaction and
decomposition. Let us bear in mind that technological development and in this case
chemical discoveries, are not always in step with chronological progression.
Fox Talbot who developed the first real practical form of photography, employing a
negative and positive, used the word Calotype to describe his process, which is
derived from the Greek word Kalos, meaning beautiful. Looking at some of the early
examples, beautiful is surely a very apt word.
Page 26 of 61
6 Method
1. We fixed the HOPG substrate to a copper supporting plate using conductive silver
paint, and further soldered a conductive wire to the copper plate to facilitate
attaching of a crocodile clip. The crocodile clip was then routed through the AFM
BNC to ground the circuit.
2. We cleaved the HOPG with adhesive tape, and then pipette 100μL AgNO3 onto the
HOPG.
3. We then wicked away the AgNO3 leaving only a meniscus of AgNO3 on the
substrate, taking care not to touch the substrate surface with the wicking paper if
at all possible.
4. We absolutely minimized the time between pipetteing the AgNO3 onto the
substrate and eventual bias application. This is because the AgNO3 is
photosensitive (see materials 5.2 above). Similarly the AFM light was switched off
during bias and image acquisition. This is to reduce any static noise interference
from the AFM light.
5. All AFM tips were UV irradiated (using a UV/Ozone ProCleaner) for 10 minutes
prior to loading into the AFM cantilever holder- this was to clean the tips.
6. We first AC mode imaged the substrate surface (for subsequent juxtaposition
with post bias images).
7. In order to determine if there was any deposition, we AC mode imaged the
substrate (per and post bias).We applied small tip-substrate forces when imaging
and applying bias in order to avoid unnecessary agitation of any nucleated
particles. We did this by using small amplitudes, and small set points. It is
important to only use topographic images of the substrate for confirmation
purposes, as other images such as ‘phase’ are not a true reflection of any
deposition. See Results 7 (below) for details of set points used.
Page 27 of 61
8. All biases were applied using custom code for the MFP3D. The biases were
applied in a grid fashion (see Figure 7.dd below) at predetermined sites, for set
durations at specified voltages. The voltage was applied so that the AFM tip was
at a negative bias with respect to the substrate which was grounded. In order to
apply a negative bias to the AFM tip we attached a conductive wire underneath
the cantilever holder and this wire was then routed through a BNC output
channel, thus enabling us to apply a specified voltage to the tip. We applied all
biases in contact mode with the surface. The experiment was carried out in
ambient air with temperature of ≈ 298 +/- 2K.
9. For any experiments that we conducted using a gold substrate (Arrandee gold
11 x 11 mm), we prepared the surface as follows- Ozone irradiate for 10 minutes,
rinse in isopropanol, ethanol and then polish with lens paper (Fisherbrand lens
cleaning tissue range 0960F00021, not with Kim wipes- as these scratch the
substrate) , Milli-Q water and then desiccate using a dry nitrogen gun.
10. We also calibrated the cantilever using the Thermal method (in order to ascertain
the spring constant) and by performing a force distance curve on glass to get the
Invols (inverse optical lever sensitivity) of the cantilever.
11. All experiments were performed in air at room temperature. See Results 7
(below) for details of voltages applied, bias duration and set points used in the
different experiments.
12. AFM topography images were processed using Gwyddion (64 bit) freeware,
Nanotec Electrónica WSxM freeware, and Igor Pro software.
Page 28 of 61
7 Results
The nano-deposits generated during the contact bias application are shown below.
Figure 7.a (below) shows Ag deposition (bright silver feature in upper left) on the
HOPG surface following an applied bias of -3.75V for 5.5 seconds using a set point of
0.6V which corresponds to a tip-surface force of 6.8nN. The volume of this Ag
deposition is 1,475,104.0 nm3.
Figure 7.b (below) shows the cross section profile of line 1 from Figure 7.a.
Figure 7.c (below) shows the same surface prior to any bias in a meniscus of AgNO3
for comparison purposes. The step edges and basal planes of the freshly cleaved
HOPG are clearly visible.
There is no evidence of any pit associated with this deposition.
Figure 7.a
Ag deposition on HOPG after an applied bias of -3.75V for 5.5 seconds in contact
mode.
The deposition is non symmetric, and is not associated with a step edge, and is
generally conical in shape being wider at the base.
Page 29 of 61
Figure 7.b
Cross-sectional profile of deposition for an applied bias of -3.75V for
5.5 seconds.
Figure 7.c
Surface of HOPG prior to any bias being applied. Image acquired in AM-AFM mode
Figure 7.d below shows Ag deposition on HOPG in a meniscus of AgNO3. The applied
voltage was -3.85V for 3.5seconds in contact mode with a set point of 0.8V (9.1nN
tip-HOPG contact force). The larger deposition (bottom middle) volume is
199,556.3nm3., the smaller deposition volume is 20,194.0nm3, total deposition of
219,750.3nm3;.
Figure 7.f shows the same surface of HOPG prior to any bias being applied. Image
acquired in AM-AFM mode, set point 0.8V
Page 30 of 61
Figure 7.d
Ag deposition on HOPG, bias applied -3.85V for 3.5 seconds in contact mode. Image
acquired in AM-AFM
Figure 7.e
Cross-sectional profile of deposition for an applied bias of -3.85V for 3.5 seconds.
Again the deposition is conical in shape and not closely associated with a step edge.
The deposition volume is markedly reduced from that obtained in Figure 7.a above.
The deposition in Figure 7.a was done at a higher voltage and for a longer duration.
Page 31 of 61
Figure 7.f
Surface of HOPG prior to any bias being applied. Image acquired in AM-AFM mode
Figure 7.g (below) shows Ag deposition on HOPG in a meniscus of AgNO3. The
applied voltage was -3.90V for 2.5 seconds in contact mode with a set point of 0.8V
(9.1nN tip-HOPG contact force).The deposition volume is 687,901.2 nm3; the smaller
deposition volume is 1.2 nm3.
Figure 7.g
Ag deposition on HOPG, bias applied -3.90V for 2.5 seconds in contact mode. Image
acquired in AM-AFM mode.
The bias was applied in contact mode at a single location (just below the beginning of
the finger of deposition). This finger of deposition was in contrast to the localised
deposition of Figure 7.a & Figure 7.b (above). It is possible the finger of deposition
was originally localised deposition that was spread out by the action of the AM-AFM
tip motion.
Page 32 of 61
Figure 7.h
Cross-sectional profile of deposition from Figure 7.g for an applied bias of -3.90V for
2.5 seconds.
Figure 7.i (below) shows Ag deposition on HOPG in a meniscus of AgNO3. The
applied voltage was -3.85V for 3.0 seconds in contact mode with a set point of 0.4V
(4.5nN tip-HOPG contact force).The deposition volume is 113,187.1 nm3.
Figure 7.i
Ag deposition on HOPG, bias applied -3.85V for 3.0 seconds in contact mode. Image
acquired in AM-AFM mode, set point 0.4V, free air amplitude 0.50V.
Again a finger like deposition is observed in contrast to localised deposition, even
though the bias was applied at a single spot.
Page 33 of 61
Figure 7.j
Cross-sectional profile of deposition for an applied bias of -3.85V for 3.0
seconds.
Figure 7.k
Surface of HOPG shown in Figure 7.i prior to any bias being applied. Image
acquired in AM-AFM mode, set point 0.35V, free air amplitude 0.44613V.
Figure 7.l (below) shows Ag deposition on HOPG in a meniscus of AgNO3. The
applied voltage was -3.80V for 3.0 seconds in contact mode with a set point of 0.4V
(4.5nN tip-HOPG contact force).The deposition volume is 72,617.3 nm3; the volume
of the pit is 200,864.4nm3.
Figure 7.l
Ag deposition on HOPG Bias applied -3.80V for 3.0 seconds in contact mode. Image
acquired in AM-AFM mode, set point 0.4V.
Page 34 of 61
It is clear from Figure 7.l above and Figure 7.m below that any deposition is now
accompanied by the formation of a pit. The deposition volume is 36.2% of the pit
volume.
Figure 7.m
Cross-sectional profile of deposition shown in Figure 7.l for an applied bias of -
3.80V for 3.0 seconds.
Figure 7.n
Surface of HOPG shown in Figure 7.l prior to any bias being applied. Image
acquired in AM-AFM mode, set point 0.39V, free air amplitude 0.50V.
Page 35 of 61
Figure 7.o (below) shows another site of Ag deposition on HOPG in a meniscus of
AgNO3. The applied voltage was -3.80V for 3.0 seconds in contact mode with a set
point of 0.4V (4.5nN tip-HOPG contact force).The deposition volume is 69,235.7nm3;
the volume of the pit is 283,636.6 nm3. The deposition volume is 24.4% of the pit
volume.
Figure 7.o
Ag deposition on HOPG Bias applied -3.80V for 3.0 seconds in contact mode. Image
acquired in AM-AFM mode, set point 0.4V.
Figure 7.p
Cross-sectional profile of deposition and pit shown in Figure 7.o for an applied bias
of -3.80V for 3.0 seconds.
Page 36 of 61
Figure 7.q (below) shows another site of Ag deposition on HOPG in a meniscus of
AgNO3. The applied voltage was -3.80V for 3.0 seconds in contact mode with a set
point of 0.4V (4.5nN tip-HOPG contact force). The deposition volume is
44,2983.4nm3; there is no associated pit, and the deposition is not along a step edge.
Figure 7.q
Ag deposition on HOPG, bias applied -3.80V for 3.0 seconds in contact mode. Image
acquired in AM-AFM mode, set point 0.4V.
Figure 7.r
Cross-sectional profile of deposition shown in Figure 7.q for an applied bias of -
3.80V for 3.0 seconds.
Figure 7.s (below) shows Ag deposition on HOPG in a meniscus of AgNO3 using a Si
tip. The applied voltage was -4.5V for 4.5 seconds in contact mode with a set point of
0.39V. The deposition volume is 6,865.2 nm3; with the associated pit volume being
30,868.9nm3. The deposition volume is 22.2% of the pit volume.
Page 37 of 61
Figure 7.s
Ag deposition on HOPG, bias applied -4.50V for 4.5 seconds in contact mode. Image
acquired in AM-AFM mode, set point 0.39V, free air amplitude 0.49337V, using a Si
tip.
Figure 7.t
Cross-sectional profile of deposition shown in Figure 7.s for an applied bias of -
4.50V for 4.5 seconds using a Si tip.
Figure 7.u
Surface of HOPG shown in Figure 7.s prior to any bias being applied. Image
acquired in AM-AFM mode, set point 0.35V, free air amplitude 0.45266V.
Page 38 of 61
Figure 7.v (below) below shows Ag deposition on HOPG in a meniscus of AgNO3
using a Si tip. The applied voltage was -5.0V for 3.0 seconds in contact mode with a
set point of 0.39V. The deposition volume is 671,390.0 nm3; with the associated pit
volume being 8,257.2 nm3. The deposition volume is 8,131% per the pit volume.
Figure 7.v
Ag deposition on HOPG, bias applied -5.0V for 5.0 seconds in contact mode. Image
acquired in AM-AFM mode, set point 0.39V.
Figure 7.w
Cross-sectional profile of deposition shown in Figure 7.v for an applied bias of -5.0V
for 5.0 seconds using a Si tip.
Figure 7.x (below) shows Ag deposition on HOPG in a meniscus of AgNO3 using a Si
tip. The applied voltage was -5.0V for 3.5 seconds in contact mode with a set point of
0.4V. The deposition volume is 3,658.6 nm3; with the associated pit volume being
803.9 nm3. The deposition volume is 455% per the pit volume.
Page 39 of 61
In addition to the bias site deposition decorating the atomically-smooth regions of
the graphite surface, it is clear from the AM-AFM image shown in Figure 7.x (below)
that nanoscopic silver particles/ribbons are also appearing along step edges on the
graphite surfaces. This deposition could be triggered by the photosensitive nature of
the AgNO3, see 5.2 (above) and or/coupled with the higher conductivity of the step
edges compared to basal planes (see above).
Figure 7.x
Ag deposition on HOPG, bias applied -5.0V for 3.5 seconds in contact mode. Image
acquired in AC mode, set point 0.4V.
Figure 7.y
Cross-sectional profile of deposition shown in Figure 7.x for an applied bias of -
5.0V for 3.5 seconds using a Si tip.
Page 40 of 61
Figure 7.z (below) shows Ag deposition on HOPG in a meniscus of AgNO3 using a Si
tip. The applied voltage was -5.0V for 4.0 seconds in contact mode with a set point of
0.4V. The deposition volume is 257,457.2 nm3; with the associated pit volume being
26,416.8nm3; the deposition volume being 974% per the pit volume.
Figure 7.z
Ag deposition on HOPG Bias applied -5.0V for 4.0 seconds in contact mode. Image
acquired in AM-AFM mode, set point 0.4V.
Figure 7.aa
Cross sectional profile of deposition shown in Figure 7.z for an applied bias of -5.0V
for 4.0 seconds using a Si tip
Page 41 of 61
Figure 7.bb (below) shows further Ag deposition on HOPG in a meniscus of AgNO3
using a Si tip. The applied voltage was -5.0V for 4.5 seconds in contact mode with a
set point of 0.4V. The deposition volume is 41,283.7 nm3; with the associated pit
volume being 37,044.1 nm3; the deposition volume being 111% per the pit volume.
Figure 7.bb
Ag deposition on HOPG, bias applied -5.0V for 4.5 seconds in contact mode. Image
acquired in AM-AFM mode, set point 0.4V.
Figure 7.cc
Cross-sectional profile of deposition shown in Figure 7.bb for an applied bias of -
5.0V for 4.5 seconds using a Si tip.
Figure 7.dd (below) shows the typical grid scenario that we employed for bias site
selection for testing the bias/duration/nucleation parameters.
Figure 7.dd
Typical grid scenario used for testing the bias/duration/nucleation
parameters. The Nucleation shown corresponds to Figure 7.o (above)
Page 42 of 61
Figure 7.ee (below) shows Ag deposition on a gold substrate in a meniscus of AgNO3
using a Si tip. The applied voltage was -2.5V for 3.0 seconds in contact mode with a
set point of 0.39V. The deposition volume is 9.25 x 10-2 m3; with no associated pit.
Figure 7.ee
Ag deposition on Gold, Bias applied -2.5V for 3.0 seconds in contact mode. Image
acquired in AM-AFM mode, set point 0.39V.
Figure 7.ff
Cross-sectional profile of deposition shown in Figure 7.ee for an applied bias of -
2.5V for 3.0 seconds using a Si tip.
Page 43 of 61
HOPG Substrate Results
Experiment
No.
Tip Voltage, V Duration, s Set Point, V Tip Force,
nN
Deposit
Volume, nm3
Pitt Volume, nm3
1 HQ:DPE -5.00 30.00 0.8 9.1 88,539.1 119,192.8
2 HQ:DPE -3.75 5.50 0.6 6.8 1,475,104.0 n/a
3 HQ:DPE -3.80 5.50 0.65 7.4 2,833.3 n/a
4 HQ:DPE -3.85 5.5 0.64 7.2 205,089.6 n/a
5 HQ:DPE -3.85 3.5 0.80 9.1 219,750.3 n/a
6 HQ:DPE -3.85 3.5 0.80 9.1 135,985.9 n/a
7 HQ:DPE -3.90 2.5 0.80 9.1 687,902.4 n/a
8 HQ:DPE -3.82 3.4 0.40 4.5 27,038.2 n/a
9 HQ:DPE -3.85 3.0 0.40 4.5 113,187.1 n/a
10 HQ:DPE -3.80 3.0 0.40 4.5 72,617.3 200,864.4
11 HQ:DPE -3.80 3.0 0.40 4.5 69,235.7 283,636.6
12 HQ:DPE -3.80 3.0 0.40 4.5 44,298.4 n/a
13 Si, SSS -4.5 4.5 0.39 * 6,865.2 30,868.9
14 Si, SSS -5.0 3.0 0.39 * 671,390.0 8,257.2
15 Si, SSS -5.0 3.5 0.40 * 3,658.6 803.9
16 Si, SSS -5.0 4.0 0.40 * 257,457.2 26,416.8
17 Si, SSS -5.0 4.5 0.40 * 41,283.7 37,044.1
Page 44 of 61
Gold Substrate Results
Experiment
No.
Tip Voltage, V Duration, s Set Point, V Tip Force,
nN
Deposit
Volume, nm3
Pitt Volume, nm3
18 Si, SSS -2.5 3.00 0.39 * 92,529,623.7 n/a
19 HQ:DPE -1.0 1.50 0.47 5.3 7,208,942.9 23,294.6
20 HQ:DPE -1.25 0.75 0.36 4.1 3,219,836.0 n/a
21 HQ:DPE -1.0 1.75 0.066 0.7 3,704,274.9 n/a
* Tip Force not calculated, this is because the Si tips Invols data was not collected at the time of experiment. The Si tips are far stiffer than the HQ:DPE tips, (see
Materials 5 (above)) so the tip-substrate force will be far higher for experiments using Si tips than HQ:DPE tips.
Page 45 of 61
8 Discussion
Jiang, Yan et al in their 2008 paper19 investigate convex and concave nanodots they
created on HOPG in ambient air by applying a voltage pulse between a metal-coated
AFM tip and the sample surface. Using a linear scan with a positive substrate bias,
nanoscale lines were also etched on the HOPG surface. Depending on the amplitude
and duration of the voltage pulse, the nanostructures were either convex or concave.
The depth of the concave structure sharply increased with the amplitude and
duration of the voltage pulse, while the height of the convexity stayed at a low level
and varied in a small range with the voltage lower than a threshold value. Under
negative substrate bias or in a vacuum, no change occurred on the HOPG surface in
the experimental range.
The formation of the nanostructures was ascribed by the authors to the primary
dissociative adsorption of water and oxygen in air induced by the intensive hole
concentration and the subsequent defect-assisted oxidation of graphite (with the
proviso that in our experiments we also have additional dissociated NO3- ions
available to play a role).24 The external electric field can induce a reaction of the
carbon surface with absorbed gases, which has been used in the fabrication of carbon-
based nanostructures20. Pits with minimum diameter down to 2 nm were produced by
applying positive voltage pulses of 3–8 V to the HOPG for 10–100 μs (note the
substrate is positively biased with respect to the tip, so electrons are flowing from the
tip to substrate which mirrors our experimental set up).21 When the HOPG and the tip
are immersed in liquid, convex structures on the HOPG surface can be created by
using electrical methods22
Jiang, Yan et al19 report that both concave and convex topography nanostructures
were produced in ambient air. The convex structures were created on the surface at
low electric voltage or short pulse duration. On increasing the amplitude or duration
of the voltage, ‘the convex profile will convert into concave morphology’.
Page 46 of 61
The authors propose that the convex profiles are attributed to concentrated holes
inducing water and oxygen dissociative adsorption onto the graphite layers while the
concave ones are formed by defect-assisted oxidation of the carbon layers under
enhanced electric field.24 We see evidence of pit formation in our experiments
(experiment no. 1, 10, 11, 13 through 17, & 19), which probably follow the mechanism
as outlined by Jiang, Yan et al. When pulses with short duration (<=1,000 ms) were
applied, nanodots with convex profiles were formed, However, where the voltage
pulses with longer durations were applied, concave dots were produced. The depth of
nanodots formed increased apparently with the duration. It is shown that the pulse
duration has a strong influence on the formation of the nanostructures.
When lower voltage pulses were applied for the same duration of 10s, the
nanostructures were not observed at the corresponding sites. This indicates that the
formation of nanodots occurs only at voltages larger than a threshold value.24 In
Jiang, Yan et al’s case, the threshold voltage is estimated to be in the range of 5–6 V,
and is dependent on the etching speed and air humidity.20Convex etching lines were
formed at the voltage lower than a specific threshold. On increasing the voltage over
the threshold, the etching lines became concave. It is dependent on the etching speed
and air humidity.20
The authors go on to suggest that the reaction between the water or oxygen and the
carbon layers most likely causes the formation of the nanostructures. The authors
further go on to outline the mechanism for nanodot and nanopit formation: under
the applied voltage one of the oxygen-free23 electron pairs of a water molecule close to
the HOPG surface can interact with the hole and a C–O bond is formed. The
formation of C–O bonds may cause the carbon lattice to produce a strain of the
topmost layer of HOPG and form a protrusion. This can be described as the first
stage. At the following stage, on increasing the amplitude or duration of the applied
voltage, more C–O bonds are formed and the carbon lattice strain increases. When
the strain reaches a limitation, some C–C bonds on the top area of the protrusions
begin to fracture and a small pit is produced. The whole process is shown in Figure
8.a (below).24
Page 47 of 61
Figure 8.a
Schematic diagrams of nanostructures on the HOPG surface produced in ambient
air by the AFM tip under an external electric field.24
Whilst our experiments were not conducted under the same conditions as Jiang, Yan
et al.: our experiments were conducted in a meniscus of AgNO3 and not solely in
ambient air (i.e. H2O meniscus/electrochemical cell) which would affect the reactions
occurring at the substrate surface, also our applied electric field was stationary- no
scan-speed. We did observe pit formation (experiments 1, 10, 11, 13, 14, 15, 16, 17 &
19). Pits were preferentially formed in our experiments that were conducted with Si
tips, but not exclusively (experiments 1, 10, 11 & 19, see results tables above). It is
possible that our pit formation follows a similar path as outlined above ‘defect
induced oxidation of graphite facilitates the formation of concave structures’24.
Furthermore the stiffer Si tips would make a better contact with the HOPG surface
thus lowering the impedance and facilitating the oxidation of the HOPG. This could
be further investigated by repeating our experiments whilst at the same time
measuring voltage and current.
Park, Jin Gyu, et al.24 report in their 2007 paper that sub-100 nm holes were made on
HOPG surfaces using a metal-coated AFM tip and carbon nanotubes. The hole-
formation mechanism is related to the chemical reaction of graphite with adsorbed
water and tunnelling electrons from the tip to substrate. The authors suggest that
chemical reactions between HOPG and tunnelling electrons are a more important
mechanism than field-emission electrons. The substrate (HOPG) was always
maintained at a higher voltage than the AFM tip (this is the same configuration as our
experiments). However the authors applied −10 V pulses to the metal-coated tip with
a 50 ms pulse width (50% duty ratio). With 1000 repetition times, a hole was
fabricated on the HOPG surface.24
Page 48 of 61
The authors go on to contend that ‘hole diameter depended on the applied voltage,
contact force, number of pulses, and humidity; there was a threshold voltage to make
a hole and any amount over this voltage created a larger hole proportional to the
voltage amplitude.’
Further the authors submit that ‘the surface experienced too much damage with high
current (several μA) over the threshold voltage. Therefore higher contact force in this
voltage range gives higher current; hence the size cannot be controlled.’ The authors
suggest several mechanisms for hole fabrication such as mechanical contact between
tip and sample, heating, electro-migration, electronic contact, field-induced
electrochemical etching, and field-induced evaporation. However, the mechanical
effect can be ruled out based upon previous tests and other reports25. Our contact
force is below the HOPG yield strength (see ensuing discussion below on this point).
The local heating effect was controversial due to the high thermal conductivity of
graphite and metals.26 If the tip apex has low thermal conductivity, then the local heat
can build up to 1000K which is enough to dissociate graphite (C) to gas (CO2).26
Again this experimental set-up was similar to our configuration in that the substrate
(HOPG) was always maintained at a higher potential than the AFM tip, however the
voltages were larger (10V) and applied for shorter durations 50 ms pulse width (50%
duty ratio) with 1000 repetitions. However there are enough similarities between the
studies for the authors suggestion that the possible machining mechanism can be
ascribed to the chemical reaction of graphite with tunnelling electrons to be
applicable in our work, but as the authors note ‘further research is needed in a more
controlled atmosphere’ before we can make firm conclusions in this regard.
Xu (2003)27 reported on the nano-indentation of HOPG surface and pure elastic
deformation up to a maximum load of ≈610 μN.28 Fraxedas et al.29 (2002) concluded
that the plastic yield threshold was not reached when forces as large as 16 μN are
applied,29 with the penetration at plastic yield larger than 25 nm. Compared with the
above references, the constant contact force ≈4 nN in our work (with the HQ-DPE
tips) shall be negligible in playing a part in patterning or modification of the HOPG
sample surface.
Page 49 of 61
Zoval, Jim V. et al report on voltage pulses employed to electrochemically deposit
silver nanocrystallites on atomically smooth graphite basal plane surfaces.17 Voltage
pulses with amplitudes of 100, 250, and 500 mV vs Ag0 and durations of 10 or 50 ms
were applied to graphite surfaces immersed in dilute (≈1.0 mM) aqueous silver
nitrate. Whilst the experimental set up is not identical to ours the studies are similar
enough for us to gain insight from their findings. Zoval’s experimental set up is
shown in Figure 5.c (above).
Further the authors state that any nucleated silver nanocrystals interact weakly with
the graphite surface and are removed by the sweeping action of the probe tip from the
imaging area. This effect has been previously documented for gold particles on
graphite by Schaefer et al18 (see Introduction 1 above).
Further the authors state that silver micro and nanocrystallites which nucleate at
defect sites are observed by STM and AFM, and, consequently, it had been
concluded30 that silver over potential deposition (OPD) on graphite is initiated by
nucleation exclusively at defects, such as step edges, on the graphite surface. The NC-
AFM data presented by the authors suggest that on low-defect density surfaces such
as the graphite basal plane, STM and repulsive-mode AFM data can provide a
misleading view of nucleation by “ignoring” the presence of weakly adsorbed metal
nanocrystallites which are not associated with defects.17 The researchers also note
that in successive silver deposition trials in which the graphite surface was cleaved
prior to each experiment, the apparent capacitance of the surface fluctuated by 10-
20%, presumably due to fluctuations in the areal density of defects on the graphite
surface which is exposed during fresh cleavage.
The researchers note that the diameters of the silver particles associated with defects
such as step edges were smaller by 20-50% compared with silver particles which were
present on nearby basal plane regions of the same graphite surface.17 Further they
elucidate that the size disparity probably derives from the fact that the diffusional
transport of Ag+ to growing nuclei arrayed along step edges had a cylindrical
symmetry, whereas nuclei on the basal plane, which were farther removed on average
from nearest neighbours, experienced a more efficient hemispherical diffusional flux
leading to faster growth and a larger terminal radius, see Figure 7.z above for graphic
nucleation/deposition similarities in our work.
Page 50 of 61
However we are not seeing these results replicated entirely in our experiments: it is
worth bearing in mind our experimental parameters were different in that we
generally applied higher voltages for longer durations in a meniscus (as opposed to
bath) which may lead to a masking of the Zoval trends observed at smaller time
frames and lower voltages.
Significantly the researchers note that at values of coulometric loading (quantity of
matter transformed during an electrolysis reaction by measuring the amount of
electricity in coulombs consumed or produced) 31, QAg greater than ≈15 μC cm-2, a
branching occurs in which an ever increasing fraction of the deposition charge is
consumed with the deposition of silver onto micron-scale crystallites instead of onto
nanocrystallites on the surface. The results of our experiments are generally micron
sized deposits with some smaller associated nano-deposits in the general vicinity
(and sometime nano-pits (see above)). Probably our results are predominantly
micron sized because the voltages and durations we employed are higher than in
Zoval’s experiments, and our coulometric loading was therefore higher and favoured
eventual micron sized deposits. We would need to measure and monitor the induced
current during deposition to confirm this.
Zoval et al conclude that the silver electrodeposition mechanism is the following:
within ≈5 ms of the application of a large potentiostatic pulse to the graphite surface,
critical silver nuclei are established both at defect sites on the surface and at high
areal density on the defect-free graphite basal plane. The number density of these
silver nuclei does not increase appreciably with time (up to ≈50 ms). Following their
formation and for the ensuing 15 ms, critical nuclei grow at a rate limited by the
hemispherical (for deposition not located on step edges) diffusive flux of Ag+ to each
nucleation site. At approximately 15ms, the rate of growth of most silver
nanocrystallites slows dramatically, and further silver deposition is concentrated at a
small fraction of crystallites which increase very rapidly in size-attaining micron-scale
dimensions within ≈20-40ms. Again Zoval conducted his experiments in a bath of
AgNO3, but the tendency toward micron scale nucleation at a higher bias and