KTH Chemical Science
and Engineering
Nanotribology, Surface Interactions
and Characterization: An AFM Study
Rubén Álvarez-Asencio
Doctoral Thesis at the KTH Royal Institute of Technology
Stockholm 2014
ii
Akademisk avhandling som med tillstånd av Kungliga Tekniska
Högskolan framlägges till offentling granskning för avläggande av
teknologie doctorsexamen den 13 juni 2014 kl 10 i hörstal F3, KTH
Lindstedtsvägen 26, Stockholm. Avhandling presenteras på engelska.
Nanotribology, Surface Interactions and Characterization: An
AFM Study
Rubén Álvarez-Asencio ([email protected])
Doctoral Thesis
TRITA-CHE Report 2014:13
ISSN 1654-1081
ISBN 978-91-7595-102-7
KTH Royal Institute of Technology
School of Chemical Science and Engineering
Surface and Corrosion Science
Drottning Kristinas väg 51
SE-100 44 Stockholm
Sweden
iii
Denna avhandling är skyddad enligt upphovsrättslagen. Alla
rättigheter förbehålles.
Copyright© 2014 Rubén Álvarez-Asencio. All rights reserved. No part
of this thesis may be reproduced by any means without permission
from the author.
The following items are reprinted with permission:
Paper I: Copyright© American Chemical Society
Paper II: Copyright© AIP Publishing LLC
Paper III: Copyright© Springer
Printed at US-AB, Stockholm 2014
iv
“God grant me the serenity
to accept the things I cannot change;
courage to change the things I can;
and wisdom to know the difference”.
Reinhold Niebuhr
v
To Antonio, Ángeles, Antonio-José and Alejandro
vi
Abstract
When two surfaces achieve contact, then contact phenomena such as
adhesion, friction and wear can occur, which are of great interest in many
disciplines, including physics, physical chemistry, material chemistry, and
life and health sciences. These phenomena are largely determined by the
nature and magnitude of the surface forces such as van der Waals, capillary
and hydration forces. Moreover these forces are length-dependent, and
therefore when the system scales down, their contribution scales up,
dominating the interaction between the surfaces.
A goal of my PhD work was to investigate fundamental contact phenomena
in terms of the surface forces that regulate their properties. The primary tool
applied in this PhD thesis work has been the atomic force microscopy
(AFM), which (with all of its sub-techniques) offers the possibility to study
such forces with high resolution virtually between all types of materials and
intervening media. Therefore, in this work it was possible to study the long
ranged interactions presented in air between different industrially relevant
materials and how these interactions are shielded when the systems are
immersed in an ionic liquid.
Also investigated was the influence of microstructure on the tribological
properties of metal alloys, where their good tribological properties were
related with the vanadium and nitrogen contents for a FeCrVN tool alloy
and with the chromium content for a biomedical CoCrMo alloy. Moreover,
the effect of the intervening media can significantly affect the surface
properties, and when the biomedical CoCrMo alloy was immersed in
phosphate buffer saline solution (PBS), repulsive hydration forces
decreased the friction coefficient and contact adhesion. On the other hand,
with the immersion of the FeCrVN tool alloy in the NaCl solution, small
particles displaying low adhesion were generated in specific regions on the
vii
surface with low chromium content. These particles are assumed to be
related to a prepitting corrosion event in the tool alloy.
The mechanical properties of stratum corneum (SC), which is the outermost
layer of the skin, were also studied in this work. The SC presents a highly
elastic, but stiff surface where the mechanical properties depend on the
nanoscale. A novel probe has been designed with a single hair fibre in order
to understand how the skin deforms locally in response to the interaction
with such a fibre probe. This study revealed that is mostly the lateral scale
of the deformation which determines the mechanical properties of the SC.
Finally, important achievements in this work are the developments of two
new techniques - tribological property mapping and the Hybrid method for
torsional spring constant evaluation. Tribological property mapping is an
AFM technique that provides friction coefficient and contact adhesion maps
with information attributed to the surface microstructure. The Hybrid
method is an approach that was originally required to obtain the torsional
spring constants for rigid beam shaped cantilevers, which could not be
previously determined from their power torsional thermal spectra
(conventional method). However, the applicability is shown to be general
and this simple method can be used to obtain torsional spring constants for
any type of beam shape cantilever.
viii
Sammanfattning
När två ytor kommer i kontakt, sker det en interaktion dem emellan varvid
grundläggande kontaktfenomen som adhesion, friktion och förslitning
uppkommer. Detta är av stort intresse inom flera discipliner till exempel
fysik, fysiskalisk kemi, materialkemi och medicinsk forskning. Dessa
fenomen bestäms till stor del av naturen och omfattningen hos de krafter
som verkar på ytorna såsom van der Waals-, kapillär- och hydrationskrafter.
Dessutom är dessa krafter beroende av avståndsberoende så för system i
liten skala blir deras vikt mer betydande och kommer att dominera
interaktionen mellan ytorna.
Målet med denna avhandlig har varit att undersöka grundläggande
kontaktfenomen i termer av de ytkrafter som styr deras unika egenskaper.
Det huvudsakliga verktyget som har tillämpats är atomkraftsmikroskopet
(AFM) som (med all dess sub-tekniker) erbjuder möjligheten att kunna
studera dessa krafter med hög upplösning mellan närapå alla typer av
material och mellanliggande media. Därför har det i detta arbete varit
möjligt att studera de brett varierande interaktionerna som förekommer i
luft mellan olika industriellt relevanta material och hur dessa interaktioner
är skärmade när systemen befinner sig i en jonvätska.
En annan studie var inverkan av mikrostruktur på de tribologiska
egenskaperna hos metallegeringar. De goda tribologiska egenskaperna
fanns vara relaterade till vanadin- och krominnehållet för FeCrVN
verktygslegeringen och med kromhalten för den biomedicinska CoCrMo
legeringen. Vidare kan effekten av mellanliggande media märkbart påverka
ytans egenskaper och när den biomedicinska CoCrMo legeringen var dränkt
i fosfat buffertsaltlösning (FBS), reducerades friktionskoefficienterna och
kontaktvidhäftningen av repulsiva hydrationskrafter. När FeCrVN
ix
verktygslegeringen var dränkt i en NaC1 lösning, alstrades små partiklar
inom specifika områden på ytan. Dessa partiklar uppvisade en låg
vidhäftning och antas vara relaterade till en tidigare gropfrätning i
verktygslegeringen.
Vidare studerades även de mekaniska egenskaperna av stratum corneum
(SC), vilket är det yttersta lagret av huden. SC uppvisar en hög elasticitet
men med stela ytor där de mekaniska egenskaperna är beroende av
nanonivån. En prob tillverkades med ett enstaka hårfiber bundet i änden av
kantilevern. Fibern slipades efteråt med hjälp av en fokuserad jonstråle.
Denna nya typ av prob användes för att kunna förstå hur huden deformeras
lokalt som gensvar på interaktionen med en sådan fiberprob. Studien
avslöjade att det till största del är den laterala skalan av deformationen som
avgör de mekaniska egenskaperna hos SC.
Slutligen, viktiga delar av detta arbete är utvecklingen av ”tribological
property mapping” och hybridmetoden.”Tribological property mapping” är
en AFM teknik som ger bilder av friktionskoefficienter och
kontaktvidhäftning vilket ger information om mikrostrukturen på ytan.
Hybridmetoden däremot är en uppskattning som från början var designad
för att erhålla vridmomentsfjäderkonstanterna för styva, rektangulära
kantilevrar, vilka tidigare inte kunde bestämmas utifrån deras termiska
kraftspektra. Emellertid visar det sig att tillämpningen är allomfattande, och
denna metod kan användas för att erhålla vridmomentsfjäderkonstanter för
alla typer av rektangulära kantilevrar.
x
List of Papers
Included Papers
I. Ionic Liquid Nanotribology: Stiction Suppression and Surface Induced
Shear Thinning
Álvarez-Asencio, R., Cranston, E. D., Atkin, R., Rutland, M. W.
Langmuir, 2012, 28, 9967-9976
II. Determination of Torsional Spring Constant of Atomic Force Microscopy
Cantilevers: Combining Normal Spring Constant and Classical Beam
Theory
Álvarez-Asencio, R., Thormann, E., Rutland, M. W.
Review of Scientific Instruments, 2013, 84, 096102
III. Tribological Property Mapping: Local Variation in Friction Coefficient and
Adhesion
Álvarez-Asencio, R., Pan, J., Thormann, E., Rutland, M. W.
Tribology Letters, 2013, 50, 387-395
IV. Nanotribology and Microstructure of a CoCrMo Alloy: A Tribological
Properties Mapping Study
Álvarez-Asencio, R., Bettini, E., Pan, J., Leygraf, C., Rutland, M. W.
Manuscript
V. Role of Microstructure on Pitting Corrosion Initiation of an Experimental
Tool Alloy: A PeakForce QNM Atomic Force Microscopy Study
Álvarez-Asencio, R., Sababi, M., Pan, J., Ejnermark, S., Rutland, M. W.,
L. Ekman
Corrosion Science, Submitted, 2014
xi
VI. Nanomechanical Properties of Human Skin Studied by AFM and a Novel
Hair Indenter
Álvarez-Asencio, R., Wallqvist, V., Kjellin, M., Luengo, G., Rutland, M. W.,
Nordgren, N.
Manuscript
The author contribution to the included papers:
I. Major part of the experimental work and manuscript preparation
II. All the experimental work and major part of manuscript preparation
III. All the experimental work and part of manuscript preparation
IV. Part of the experimental work and major part of manuscript preparation
V. Part of the experimental work and major part of manuscript preparation
VI. Part of experimental work and part of manuscript preparation
Other Papers not Included in this Thesis
I. Monolayer Study by VSFS: In Situ Response to Compression and Shear in a
Contact
Ghalgaoui, A., Shimizu, R., Hosseinpour, S., Álvarez-Asencio, R., McKee, C.,
Johnson, M., Rutland, M., W.
Langmuir, 2014, 30, 3075-3085
xii
Summary of Papers
The aim of Paper I was to investigate the friction and adhesion between
relevant pairs of materials (silica, alumina, polytetrafluoroethylene) and
interpret them with regard to the longer ranged interactions between the
surfaces. In ambient air the interactions are controlled by attractive van der
Waals and strong adhesion, leading to significant frictional forces. In
ethylammonium nitrate (EAN) the van der Waals attraction is shielded, and
the adhesive/attractive interactions which lead to stiction are almost
eliminated, reducing frictional forces 10-fold at high applied load. The
friction coefficients in EAN were also significantly reduced and the
variation between systems was correlated with surface roughness. The
hydrodynamic forces between the materials surfaces have been also
investigated in EAN. A linear increase of these forces with velocity is
observed, reducing the probability of stiction. An unexpected result was
found when the viscosity extracted from the data was almost 3 times lower
than the EAN bulk viscosity, indicating a surface ordering effect.
In Paper II, a calibration technique was developed for the calculation of
torsional spring constants for AFM cantilevers by combining the normal
spring constant measurement with plate/beam theory. Originally, the aim of
this method was the determination of torsional spring constants for stiff
cantilevers where it is not possible to find the necessary torsional resonance
frequency peak because of its very low signal/noise ratio. However, its
applicability is more general and it can in fact be used to obtain the
torsional spring constant in a simple manner for any beam shaped
cantilever.
Tribological property mapping (TPM) is a new imaging technique
developed in Paper III that generates friction coefficient and adhesion
xiii
maps. This technique is based on the combination of a series of lateral
atomic force microscopy (LAFM) images obtained as a function of load,
which are tiled and pixelwise fitted to a modified Amonton’s law, obtaining
both friction coefficient and an adhesion value for each pixel. This imaging
technique eliminates the uncertainty of the friction contrast in conventional
LAFM imaging since the friction coefficient and adhesion are independent
of the load applied during the mapping. As an example of the application of
the technique, a heterogeneous commercial powder metallurgical tool alloy
has been scanned using a silicon tip, immersed both in air and in
tetradecane, in order to obtain friction coefficient and adhesion maps. These
data provide unique information related to the heterogeneous microstructure
of the alloy as well as an enhanced understanding of the tribological
properties of the material.
TPM was used in Paper IV to study the local tribological properties of a
biomedical CoCrMo alloy in a phosphate buffer solution (PBS) that mimics
the saline conditions of the body. The biomedical alloy turned out to be
stable during the measurement in PBS and displayed low friction
coefficient and contact adhesion values. These low values were attributed to
the chromium oxide surface layer and the hydration forces which originated
in PBS, even led to positive values (repulsive) in the contact adhesion map.
In Paper V, the adhesion properties of a FeCrVN tool alloy in water and
sodium chloride have been studied by PeakForce® QNM in order to
understand the influence of the microstructure on adhesion and corrosion
initiation of the tool alloy. It turned out that the adhesion of the alloy is
strongly influenced by the vanadium and nitrogen contents in water.
However when the alloy is immersed in sodium chloride, other factors
affect the system and small particles are formed on the surface with mostly
xiv
very low adhesion. These particles are assumed to be related to prepitting
events that may lead to passivity breakdown of the alloy.
Nanomechanical properties of the outer most layer of the skin (stratum
corneum, SC) have been studied by atomic force microscopy in Paper VI.
Nanomechanical mapping reveals that the Young’s modulus of the SC
varies somewhat over the surface with a mean value of 0.39 GPa, and the
force indentation measurements show that the SC is deformed permanently
at high applied loads (above 4 µN). In Paper VI a novel probe has also
been designed using a single hair fiber and sharpened with a focused ion
beam. The force indentation measurements performed with this probe on
SC reveals that the lateral scale of the deformation determines the Young’s
modulus of the elastic outermost layer of the skin.
xv
Table of Contents
Abstract ................................................................................................................ vi
Sammanfattning ................................................................................................. viii
List of Papers ......................................................................................................... x
Included Papers .................................................................................................. x
Other Papers not Included in this Thesis .......................................................... xi
Summary of Papers ............................................................................................. xii
Table of Contents ................................................................................................ xv
Symbols ............................................................................................................. xvii
1 Introduction ........................................................................................................ 1
1.1 Surface Forces .............................................................................................. 1
1.1.1 van der Waals Forces ............................................................................. 1
1.1.2 Capillary Forces ..................................................................................... 2
1.1.3 Solvation Forces .................................................................................... 3
1.1.4 Hydration Forces .................................................................................... 4
1.1.5 Hydrodynamic Forces ............................................................................ 5
1.2 Adhesion ...................................................................................................... 5
1.3 Tribology ...................................................................................................... 6
1.3.1 Nanotribology ........................................................................................ 9
2 Experimental .................................................................................................... 11
2.1 Atomic Force Microscope .......................................................................... 11
2.1.1 Imaging ................................................................................................ 12
2.1.2 Force .................................................................................................... 12
2.1.3 Friction ................................................................................................. 14
2.1.4 PeakForce® QNM ................................................................................ 16
2.1.5 Colloidal Probe .................................................................................... 17
2.1.6 Determination of Spring Constants ..................................................... 18
xvi
2.1.7 Focused Ion Beam (FIB)...................................................................... 20
3 Materials and Fluids ......................................................................................... 22
3.1 Cantilevers and Probes ............................................................................... 22
3.2 Substrates ................................................................................................... 23
3.3 Liquid Media .............................................................................................. 25
4. Summary of Key Results ................................................................................ 27
4.1 Suppression of Surface Interactions by an Ionic Liquid ............................ 27
4.2 A Hybrid Route for the Determination of Torsional Spring Constant for AFM Cantilevers ......................................................................... 32
4.3 A New Technique to Characterize Heterogeneous Surfaces: Tribological Property Mapping ........................................................................ 36
4.4 Biomedical CoCrMo alloy: A Tribological Properties Mapping Study ................................................................................................................ 43
4.5 Surface Study and Corrosion Initiation of an Experimental FeCrVN Tool Alloy ......................................................................................... 48
4.6 Nanomechanical Properties of Human Skin .............................................. 56
4.7 A Novel AFM Probe: The Single Hair Fibre Probe .................................. 60
5. Conclusions ..................................................................................................... 66
6. Acknowledgment ............................................................................................ 69
7. References ....................................................................................................... 70
xvii
Symbols
FLoad Applied Load
∆Vd Average Lateral Output Voltage
k Boltzmann Constant
Sc Critical Shear Contact Stress
rk Curvature of the Meniscus
λ Decay Length
ρ Density
D Distance, Surface Separation
heff Effective Height
Reff Effective Radius
W Energy per Unit Area
µ Friction Coefficient
FFriction Frictional Force
A Hamaker Constant
CH Hydration Constant
FHydration Hydration Force
FHydrodynamic Hydrodynamic Force
() Imaginary Component of the Hydrodynamic Function for
Normal Vibrations
() Imaginary Component of the Hydrodynamic Function for
Torsional Vibrations
δ Lateral Deflection Sensitivity
L Length
Vm Molar Volume
F Normal Force
Qz Normal Quality Factor
xviii
fz Normal Resonance Frequency
ν Poisson’s Ratio
R Radius of the Sphere
A Real Contact Area
E* Reduced Young’s Modulus
p/p0 Relative Vapour Pressure
vR Relative Velocity
G Shear Modulus
γ Surface Tension
T Temperature
t Thickness
Qϕ Torsional Quality Factor
fϕ Torsional Resonance Frequency
kϕ Torsional Spring Constant
FvdW van der Waals Force
Vn* Vertical Deflection in Newton
Vn Vertical Deflection in Volts
η Viscosity of the Fluid
w Width
E Young’s Modulus
z Z – Piezo Position
1
1 Introduction
1.1 Surface Forces
1.1.1 van der Waals Forces
van der Waals forces arise from the interaction between electromagnetic
fields generated from the surface of any material. They are formed by the
sum of three contributions: the orientation force, the induction force and
the dispersion or London force.[1, 2]
Dispersion forces usually dominate over orientation and induction forces.
Besides, they are always present, playing an important role in a multitude of
fundamental phenomena, such as, adhesion and surface tension. Therefore,
they are considered the most important contribution to the van der Waals
interactions.
The determination of the van der Waals force between molecular pairs is
straight forward by adding the orientation, induction and dispersion
contributions,[1] but for bulk materials the situation gets more complicated.
Hamaker[3] in 1937 was able to estimate the overall van der Waals
interaction between two macroscopic solids by summation of all the
molecular pair interactions between the two bodies. His approximation was
based on the interaction of a sphere against a flat surface, according to Eq.
1:
= (1)
where FvdW corresponds to van der Waals force, D is the distance between
the surfaces, R is the radius of the sphere and A the Hamaker constant,
where typical values are around 10-19 J. The addition method suggested by
2
Hamaker for the calculation of A does not consider the influence of
neighbouring atoms in the interaction between the molecular pair.
Moreover, the Hamaker approach cannot be easily applied to two bodies
with an intervening liquid. Therefore, Lifshitz in 1956[1] proposed for the
calculation of the Hamaker constant a new theory that ignores the atomic
structure and treats the force as a continuous medium. This approximation
is able to calculate the Hamaker constant based on only bulk properties,
such as, dielectric responses. Hamaker constants are mostly positive,
generating an attractive van der Waals. However, there are cases for certain
combination of media[4-7] (two different surfaces and an intermediate
fluid) where the Hamaker constant is negative, generating a repulsive van
der Waals force.
1.1.2 Capillary Forces
Capillary forces tend to arise when a capillary bridge spontaneously forms
between two hydrophilic neighbouring asperities in a humid
environment.[1, 8] These forces mostly form with the water absorbed on
lyophilic surfaces. However, they can also appear in other cases, such as,
with hydrophobic surfaces immersed in water connected by air cavities.
The Kelvin equation (Eq. 2) is a useful approximation that describes the
thermodynamic equilibrium of a drop. It determines the dimensions of the
capillary bridge:
= (/) (2)
where rk is the Kelvin radius (mean radius of curvature of the meniscus),
p/p0 the relative vapour pressure, γ is the surface tension, Vm is the
molecular volume, k is the Boltzmann constant and T the temperature.
3
Capillary forces depend mostly on the surface roughness and the relative
humidity. When the two asperities are in contact, a capillary bridge is
formed if the Kelvin radius is larger than the height of the smaller asperity
(which is characteristic of surface roughness of both surfaces).[9-11] Only
in the case of the interaction between macroscopic smooth spheres, do they
become independent of humidity.
Capillary forces have a long range thus they can mask other short range
interactions, such as van der Waals. Therefore, in order to measure other
forces, the capillary bridge must be eliminated by either working at low
relative humidity or immersing the system in liquid.[2]
1.1.3 Solvation Forces
When liquid molecules are confined between two surfaces (Figure 1), they
tend to achieve a higher ordering. The increase of this order with the
decrease in surface separation generates the solvation force.[1] This
interaction depends on the properties of the liquid media, as well as, on the
physical and chemical properties of the confining surfaces.
Solvation forces can have an attractive, repulsive or oscillatory nature,
becoming dominant at short range. Moreover, as they can be stronger than
other forces at small separations (e. g. van der Waals), they can become
more important than other interactions, contributing more to the overall
adhesion between two particles or surfaces.
4
Figure 1. Schematic showing the ordering of liquid molecules between two surfaces that generates
the solvation forces.
1.1.4 Hydration Forces
The hydration force can be thought of as a type of solvation force where the
liquid compressed between the surfaces is water, and is originated by the
overlapping of structured water layers between hydrophilic surfaces.[1, 12,
13]
When two hydrophilic surfaces are immersed in a dilute ion solution, the
interaction between the surfaces obeys the DLVO theory.[1] However,
when the ion concentration increases, the hydrated cations adsorb on the
negatively charged hydrophilic surfaces leads to a repulsive hydration force
that dominates the interaction. This force depends on the hydration number
of the cations adsorbed on the surface, increasing in range and strength with
it (Mg2+>Li+~ Na+>K+).
The hydration forces have an exponential decay thus can be successfully fit
by Eq. 3:
5
!"#$%&'(() = )!*+ ,− ./ (3)
where FHydration is the hydration force, λ is the decay length and CH is the
hydration constant.
This interaction is strong, repulsive and short ranged, and occurs typically
below 2 nm.
1.1.5 Hydrodynamic Forces
The hydrodynamic force (FHydrodynamic) is the additional force between two
bodies corresponding to displacing the liquid between them.[14] This force,
which causes dissipation of energy in hydrodynamic flow, was first
experimentally determined by Chan and Horn[15] for the interaction of a
sphere against a flat surface as a function of the viscosity of the fluid (η)
and relative velocity (vR) according to Eq. 4:
!"#$#"'%01 = 2 345 (4)
This force becomes so dominant at high relative velocities and fluid
viscosities that it shields the contribution of shorter ranged forces, such as
solvation or van der Waals forces.
1.2 Adhesion
When two solid surfaces are brought close enough, an attraction induced by
intermolecular interactions causes them to stick at contact spots or
asperities. The force that is required to overcome this interaction is defined
as adhesion or adhesion force. The forces that control the measured
adhesion depend mostly on the material pair and their interfacial properties,
6
such as roughness, surface energy, cleanliness, crystalline structure,
solubility of material in contact with each other, separation rate, time in
contact, etc. With the immersion of the system in liquid, these surface
forces may be modified or shielded, and new surface interactions may arise
(e.g. solvation forces, capillary forces, hydrodynamics, double layer forces),
contributing to the overall adhesion between the surfaces.
Therefore, the interpretation of adhesion can be challenging due to the
number of factors involved, however it can provide useful information
about the surface properties, as well as, assist to understand, predict and
control other contact phenomena like lubrication and friction.
1.3 Tribology
Tribology literally means “the science of rubbing” and was defined in 1966
as “the science and technology of interacting surfaces in relative motion and
of associated subjects and practices”.[16] This science that studies friction,
wear and lubrication is of great interest and huge practical significance.
Friction is commonly defined as the force resisting sliding when two bodies
are brought in contact, and it has been part of our history since ancient
times where, for example, fire was produced by generating high friction
between two sliding sticks and large stone building blocks were transported
using water lubricated sleds.[16] Friction prevents slipping or sliding and
thus allows the possibility to walk, write, drive, grab, push, pull, etc.
However, friction can be an inconvenience also because it resists motion.
For example friction increases the energy required to drive a car, it can even
lead to failure in mechanical devices with moving parts such as
microelectromechanical motors and gears.
The first systematic study of friction registered was performed by Leonardo
D. Vinci (1452-1519). Amontons later in 1699 received the credit and
7
postulated[17] that the friction between two sliding materials (FFriction) is
proportional to the normal applied load (FLoad) according to Eq. 5
(Amontons’ first law):
$1&'( ) = 6 7%# (5)
where µ is the friction coefficient defined as the proportionality constant
between the friction force and the applied load
This friction coefficient, which is a suitable parameter to compare the
lubrication properties between systems, is independent of both the apparent
contact area[17] (according to Amontons’ second law) and the sliding
velocity[18] (Coulombs’ law of friction). However both these laws are
phenomenological and applied at the macroscopic scale. When the adhesion
forces in the system are in the range of the applied load, as is the case at the
nanoscale, the adhesion behaves as an additional loading force, and Eq. 5 is
no longer valid. As a consequence of this additional force, friction forces
extend to negative applied loads according to the following equation
proposed by Derjaguin[19]:
$1&'( 7%#) = 6 7%# + $1&'(0) (6)
where FFriction(0) corresponds to the friction force at zero applied load. Eq. 6
is a useful simplification and allows the friction coefficient and adhesion to
be obtained independently of each other, where the adhesion or “contact
adhesion”[20] is obtained from the intercept of the friction-load relationship
with the load axis (Figure 2).
8
Figure 2. Effect of the applied load on the friction force according Derjaguins’, Amontons’ and
JKR approximations.
Amontons’ and modified Amontons’ laws are based on experimental
observations, but for a single asperity contact, the friction-force relationship
is not linear at low loads and thus another approximation is required. One of
the first attempts to deal with the adhesive contact of spherical asperities
was performed by Johnson, Kendall and Robert (JKR theory, 1971).[1, 21]
This theoretical treatment describes quite well the adhesive contact by
contact mechanics, even during sliding. This model is simple in principle
and postulates that the friction force is not proportional to the load but to
the real contact area (Figure 2)[22] according to:
$1&' = :1; (7)
where Sc is the critical shear contact stress at the contacting interface and A
is the real contact area.
However, in measurements when the deformation is small over the
measured range, the contact area does not change too much and the
modified Amontons’ law describes generally well the adhesive contact
9
interaction. Therefore, this friction-force relationship can be generally used,
has been proven in many experimental systems,[11, 23-25] and is one the
most common approaches to describe quantitative friction between
surfaces.[21, 26, 27]
1.3.1 Nanotribology
Nanotribology is a branch of tribology that studies, friction, adhesion, thin–
film lubrication and wear between sliding surfaces, at the molecular and
atomic scale.[28] This is an exciting field where scientists of different
backgrounds meet in order to understand fundamental contact phenomena
that occur when two surfaces are sliding relative to each other.
However, nanotribology faces major challenges; for example, in
miniaturized devices with moving parts, such as, micro- and nano-
electromechanical systems (MEMS/NEMS), the small length scale and
surface-area-to-volume ratio makes interfacial phenomena become
dominant. Therefore, a good understanding is required of how surface
interactions, such as van der Waals and capillary forces affect friction and
adhesion at the molecular and atomic scale.[28-30]
Some light was shed on this topic around 1942 by Bowden and Tabor[31]
that demonstrated the real contact between two solids in contact is only a
fraction of the apparent contact area due to the surface roughness. At the
nanometre/micrometre scale all surfaces are rough and they contact at some
microscopic points (asperities). Therefore, the study of the surface
interactions between these asperities at the molecular and atomic scale
would provide an enhanced understanding of nanotribology.
Since the development of atomic force microscopy (AFM) by Binnig et al.
in 1986,[32] much effort has been devoted to the study of the interactions
10
that govern nanotribology. AFM allows the study of the interactions that
occur between the tip and the surface, which is used to mimic the
interaction between the asperities.[33] Therefore, this technique has
provided new insights at length scales not previously accessible, which are
essential in order to understand nanotribology. This technique will be
discussed in Section 2.1.
11
2 Experimental
2.1 Atomic Force Microscope
Atomic force microscopy (AFM) [32] is a popular, common and extremely
versatile technique for analyzing surface properties. AFM has been mostly
designed for imaging at the nano- and micro- scales but a more specialized,
and no less important area is the study of force and friction with pico-
Newton Force resolution.[34]
Typically, the sample is glued on the top of a metal disc that is magnetically
attached to the base of the AFM (Figure 3). The scanner can be either
located in the head or in the base of the AFM and moves the sample and
cantilever relative to each other in x, y and z directions. The cantilever is
placed in a holder located over the sample with the tip pointing down. A
laser is focused on the cantilever and deflected onto a quartered photodiode.
The photodiode voltage reveals the deflection and twist of the cantilever in
response to an interaction of the tip where the vertical sections are related
with topography and forces and lateral sections with friction.
Figure 3. Schematic illustration of an AFM where: (a) is the quartered photodiode, (b) is the laser,
(c) is the cantilever, (d) is the tip, (e) is the sample and (f) is the scanner (Molworx®).
12
2.1.1 Imaging
AFM was originally developed to study topography by determining the
height changes on the sample surface. The two most common modes for
topographical imaging are contact mode and tapping mode. In contact
mode, the tip is in permanent contact with the sample, and the scanner is
moved up and down to keep the force constant. These height changes in the
scanner position provide the topography. This mode is theoretically
supposed to give better image resolution and is appropriate for relatively
hard samples where the tip cannot damage them. In tapping mode, the
cantilever is forced to oscillate with a certain amplitude, making
intermittent contact with the sample surface. The topographic image is also
generated here by the changes in the scanner height. This mode provides
generally the best performance for imaging in general conditions, being
recommended for fragile samples. Recently, new modes have been
developed such as PeakForce® Quantitative NanoMechanics (QNM) from
Bruker®, and the very promising Intermodulation AFM (ImAFM®) from
Intermodulation products®.
PeakForce® QNM provides not only topographic information, but other
relevant material properties, such as Young’s modulus and adhesion. This
mode will be described in Section 2.1.5.
2.1.2 Force
Normal forces are measured in an AFM by moving the surface toward and
away (often referred to approach and retraction) from the cantilever tip by
the piezoelectric scanner, and the raw data are obtained as cantilever
vertical deflection (Vn) versus Z-piezo position (z). The raw data are then
transformed into force versus separation as described below.
13
Figure 4. Vertical deflection (Vn) versus Z-piezo position (z-top), where A corresponds to the
constant compliance region, Z is the point of zero separation and B is the zero deflection region.
The inset displays a schematic illustration of a force measurement before contact (A) and in
contact (B). Force (FL) versus separation data (D - bottom) obtained after processing.
There are three important regions in the force curve that are identified in
Figure 4 (top); the zero deflection, the constant compliance and the point of
zero separation. The zero deflection region corresponds to the value of
vertical deflection when the tip is not interacting with the surface and is
unperturbed. The constant compliance defines the region after contact when
the cantilever deflects by the same amount as the piezoelectric scanner
moves and thus the slope of the representation Vn versus z becomes
constant. This calibrates the cantilever deflection in terms of distance units.
The projections of these two areas intersect at the point of zero separation
(Z), which is where the tip comes in physical contact with the studied
surface.
The processing of the normal force curve starts with the transformation of
Vn from Volts via the inverse of the slope extracted from the constant
compliance region. Afterwards, Hooke´s law (Eq. 8) is applied in order to
convert deflection in metres (<'∗) to force in Newtons, where kz is the
normal spring constant of the cantilever. Finally, z is converted to
separation by taking into account the movement of the cantilever in Eq. 9:
14
= >'(<'∗) (8)
( = ? − <'∗ (9)
Although the measurement and processing of force curves are usually
straightforward particularly for rigid substrates, interpretation of these force
curves is not trivial at all. A good review of this topic is presented by
Ralston et al.[33]
2.1.3 Friction
Friction is measured in lateral atomic force microscopy (LFM) by sliding
the sample perpendicularly to the cantilever axis while in contact mode at a
specific applied load, and recording the resulting cantilever twist. While
changes in the vertical deflection of the cantilever give topography and
normal forces, variation of the lateral deflection generates the friction
information between the tip and the sample. During the scanning of the
sample with the tip in both directions (often referred to trace and retrace),
the cantilever twists in opposite directions, producing two photodetector
output signals with opposite signs for each direction scanned (Figure 5). In
the absence of anisotropy, the retrace lateral signal will be a mirror image
of the trace.
15
Figure 5. Schematic representation of a typical friction loop. The inversion of the voltage sign is
related to the change in the scanning direction (Paper II).
The conversion of the vertical deflection to applied load is calculated in the
same way as for the vertical deflections show in Section 2.1.2. On the other
hand, the average lateral output voltage (∆Vd), which is obtained from the
difference between the trace and retrace output lateral photodetector
signals, is transformed to friction forces by Eq. 10:
$1&' = ∆@ABCC DE (10)
where δ is the lateral deflection sensitivity, kϕ is the torsional spring
constant, and heff is the effective height, which corresponds to the height of
the tip plus half the cantilever thickness.
Afterwards, friction forces can be displayed as a function of applied load
(Figure 6) and the modified Amontons’ law (Eq. 6) can be applied to fit the
data and extract µ from the slope and a measure of the adhesion from the
negative applied load where the Ff becomes zero (Figure 2).
16
Figure 6. A Typical relationship between the applied load and the friction force.
2.1.4 PeakForce® QNM
PeakForce® QNM or PeakForce® Quantitative NanoMechanics is an atomic
force microscopy mode that allows high resolution topography imaging, as
well as providing nanomechanical properties such as adhesion, Young’s
modulus and deformation.[35, 36] Therefore, this technique is suitable for
the study of heterogeneous surfaces like metal alloys.
In PeakForce® QNM, the piezo scanner oscillates in the normal direction
with a standard frequency of 2 kHz and a force curve is thus generated
every 0.5 ms. During the scanning, the AFM feedback loop keeps the
chosen maximum applied force constant (Peak Force) by adjusting the
overall extension of the piezo. The possibility of controlling the applied
force provides the opportunity for non-destructive imaging.
A force curve is generated with every cantilever tap and its analysis
provides the nanomechanical properties (Figure 7), which afterwards are
presented in the images.
0
17
Figure 7. Schematic representation of a force curve as a function of separation on approach (solid
line) and retraction (dotted line), indicating which part of the force curve provides the
nanomechanical properties.
Figure 7 shows that adhesion is extracted from the difference between the
baseline and the minimum force during retraction, the deformation from the
distance between zero separation and a position with a given percentage of
the full deformation of the approaching curve and the Young’s modulus by
fitting the linear part of the retraction force curve using the Derjaguin-
Muller-Toporov (DMT)[35] contact mechanical model, which describes the
relationship between applied force and deformation of a material.
Afterwards, a set of images can be obtained for every sample scan, where
each image displays a different nanomechanical property of the surface.
2.1.5 Colloidal Probe
The colloidal probe technique developed by Ducker et al.[13, 37] and
Butt[38] is based on the exchange of the AFM cantilever tip by a colloidal
particle (1-20 µm in diameter), as shown in Figure 8. It has the advantage of
allowing forces to be measured with a probe made of virtually any material,
as long as the probe has a well defined shape and is almost incompressible,
18
where the magnitude of the measured force scales with the size of the
probe.
However, the comparison of forces obtained with different probes is not
straight forward, and is thus necessary to normalize them, by applying the
Derjaguin’s approximation.[39] This approximation relates the normal
force to the energy per unit area (W) between two flat surfaces, according
to:
()@2 BCC = F(() (11)
where Reff is the effective radius that depends on the interacting surfaces. In
the case of a colloidal probe interacting against a flat surface (Paper I), the
spherical shape of the probe simplifies the calculation of effective radius,
which can be approximated to the radius of the colloidal probe.
Figure 8. Scanning electron microscope image of a 5µm silica colloidal probe glued on a tipless
cantilever (Paper I).
2.1.6 Determination of Spring Constants
In force and friction measurements, the determination of the normal and
torsional spring constants is crucial because these constants transform the
cantilever bending and twisting to forces (see Section 2.1.2). During the last
19
two decades, a large number of solutions have been proposed, which are
theoretical, experimental or a combination of both, but there is one
approach that has been widely used due to its accuracy and simplicity. [40]
This technique, developed by Sader,[41-44] is commonly known as the
Sader method.
2.1.6.1 Determination of Spring Constants by the Sader Method
The Sader method is based on how the cantilever vibration frequency
response is affected by the surrounding fluid. The cantilever is allowed to
vibrate due to thermal motion in a fluid, which is generally air. The normal
resonance frequency (fz) (which can be thought of as the vertical vibration
frequency of for example a diving board) and the normal quality factor (Qz)
are obtained by fitting a simple harmonic oscillator function to the normal
resonance peak obtained from the thermal power spectra of the cantilever,
and afterwards they are combined with the measured length (L) and width
(w) of the cantilever, and the density (ρ) of the fluid, to determine the
normal spring constant, according to Eq. 12[43] where () is the
imaginary component of the hydrodynamic function for normal vibrations.
[45]
> = 0.1906LM@NO(2QR)@Γ() (12)
= ST@2UVW5 (13)
The determination of the torsional spring constant is very similar to the
calculation of kz presented above, but in this case the torsional resonance
frequency (fϕ) and the torsional quality factor (Qϕ) are obtained from the
torsional resonance peak. Therefore, kϕ is calculated using:[42]
20
> = 0.1592LMWNOY2QRZ@Γ() (14)
= ST@2UDW5 (15)
where Γ()is the imaginary component of the hydrodynamic function
for torsional vibrations.[46]
There is a limitation in the determination of the torsional resonance
frequency from the power torsional thermal spectra because of its lower
resolution, and for stiffer cantilevers this resonance is hard to measure.
Section 4.2 and Paper II contains an approach to avoid this issue.
2.1.7 Focused Ion Beam (FIB)
Focused ion beam (FIB)[47] is a technique that employs a focussed beam of
ions (usually gallium) to irradiate the sample (Figure 9). One of the
applications of FIB is the ability to generate images from the charged
particles (ions and electrons) that are released from the sample when the
beam of ions hits the surface. Another application is the milling that occurs
during beam exposure due to the atomic collision process that removes
atoms from the sample surface. An FIB can also be used to deposit material
via a gas delivery system that provides a chemical compound close to the
beam-sample interaction point. The FIB decomposes the gas locally, and
the products are deposited on the surface. In this thesis the FIB has been
used solely as a cutting tool for microscopic samples. (Paper VI)
21
Figure 9. Schematic representation of the focus ion beam operation principle.
22
3 Materials and Fluids
3.1 Cantilevers and Probes
The AFM cantilevers used for calibration, imaging, force and friction
throughout this thesis were crystalline silicon cantilevers with rectangular
shape from MikroMasch (Tallinn, Estonia).
The silica and alumina colloidal probes used in Paper I, were provided by
Bang Laboratories (Fishers, IN) and Sveriges Tekniska Forskningsinstitut
(SP, Stockholm), respectively. These particles were glued on tipless
cantilevers using a micromanipulator (Micromanipulator 5171, Eppendorf,
Hamburg, Germany) under an optical microscope (Nikon Optiphot 100,
Tokyo, Japan).
A hair fibre is composed of a layered structured with three regions: cuticle,
cortex and medulla. The cuticle is the outer part of the fibre and is
responsible for the surface properties of the hair (Figure 10). The cortex is
the main structural component and provides the special mechanical
properties of the hair. Finally, the medulla is the central part and is not
always present, only for thick hairs, which tend to protrude mostly from the
scalp, and its function remain unclear.[48] The facial human hair involved
in the Paper VI was provided by L’Oréal (Paris, France).
23
Figure 10. Scanning electron microscopy image of a human hair cuticle surface.
3.2 Substrates
In Paper I, the silica wafers (model Ultrapack Wafershield H9100-0302)
were purchased from Entegris (Dresden, Germany) and the smooth
polytetrafluoroethylene (PTFE) surface was prepared by compressing
pieces of PTFE between fleshly cleaved mica sheets inside of an oven
overnight at 500 °C.[4]
The metal alloy used in Paper III is a nitride tool steel (Vancron® 40)
provided by Uddeholm AB (Sweden), chemical composition of which is
given in Table 1. This material consists of a homogeneous distribution of
fine nitride particles embedded in a metallic matrix. Vancron® 40 does not
need surface coating in applications because the surface microstructure of
the metal alloy leads to low adhesion against soft working materials.
Moreover, it is a hard material with enough ductility and toughness to avoid
premature failure.[49, 50]
24
Table 1. Chemical composition (wt. %) of Vancron® 40, Fe is balanced.[49]
Wt. % C N Si Mn Cr Mo W V
Vancron® 40 1.1 1.8 0.5 0.4 4.5 3.2 3.7 8.5
In Paper IV, the biomedical CoCrMo alloy was supplied by Sandvik
Material Technology (Sweden) with a chemical composition (wt %)
described in Table 2. This CoCrMo alloy consists of two different types of
nitride particles embedded in a Co rich matrix that forms a native protective
thin film mostly composed of Cr2O3 and, (in minor amount) Co and Mo
oxides.[51, 52] These alloys are widely used for joint replacement due to
their corrosion and wear resistance, good mechanical properties, and
biocompatibility with the human body.[53, 54]
Table 2. Chemical composition (wt. %) of the CoCrMo alloy, Co is balanced.[52]
Wt. % Cr C Mo Mn Ni Fe
CoCrMo alloy 28 6.3 0.5 0.2 0.2 0.2
FeCrVN is the experimental nitrogen based tool alloy studied in Paper V
provided by Uddeholm AB (Sweden), the chemical composition of which is
given in Table 3. This metal alloy is formed by two different types of
nitrides (with different vanadium and iron contents), which are embedded in
the alloy matrix.[55]
Table 3. Chemical composition (wt. %) of the FeCrVN alloy, Fe is balanced.
Wt. % C Mn Si N Ni Cr Mo V
FeCrVN 0.2 0.3 0.3 4.2 0.05 21.2 1.3 9.0
25
The stratum corneum (SC) is the outer part of the skin and is formed by
keratin-rich dead cells called coenocytes, which are inserted in a lipid
cement providing high mechanical strength and good elastic properties.
This layer has important functions, such as skin barrier, preventing water
loss, for the appearance (i.e. optical properties) and photoprotection. It also
acts as contact surface for tactile perception.[56-59] The human abdominal
SC (~15µm thick) studied in Paper VI was kindly provided by L’Oréal
(Paris, France).
3.3 Liquid Media
The ultrapure water used throughout this thesis has a pH of ca. 5.7,
resistivity of 18.2 MΩ-cm and carbon content below 2 ppb, and was
obtained with a Milli-Q unit (Millipore, Molsheim, France). The ethanol
(99.5% purity) used in this work was acquired from Kemetyl (Haninge,
Sweden).
Ionic Liquids (ILs) are molten organic salts with melting temperatures
below 100°C, which have a wide range of applications due to their unusual
properties, such as, high temperature stability, thermal conductivity, low
volatility and electrical conductivity.[60] Ethylammonium nitrate (EAN),
discovered in 1912, is a protic ionic liquid formed by an ethyl ammonium
cation and a nitrate anion that forms a 3D H-bonded network structure. This
IL has a viscosity that is 30 times higher than water.[61, 62] In Paper I,
EAN was synthesized according to a process described previously,[63]
combining ethylamine and nitric acid, which were acquired from Sigma
Aldrich (Munich, Germany) and Merck (Darmstadt, Germany),
respectively.
The tetradecane with a purity of 99% used in Paper III was purchased from
Sigma (Haninge, Sweden).
26
The phosphate buffered saline (PBS) is a solution generally used in
biological research since it mimics the saline conditions of the body. This
salt used in Paper IV contains 8.77 g/L NaCl, 0.2 g/L KCl, 1.28 g/L
Na2HPO4 and 1.36 g/L KH2PO4, and was obtained from Sigma Aldrich
(Munich, Germany).
27
4. Summary of Key Results
4.1 Suppression of Surface Interactions by an Ionic Liquid
In Paper I, the friction and adhesion between pairs of industrially relevant
materials (silica, alumina and polytetrafluoroethylene-PTFE) have been
analyzed in terms of the long ranged surface interactions, such as, van der
Waals and capillary forces. The four systems (with the convention “probe-
surface”: silica-silica, silica-PTFE, alumina-silica and alumina-PTFE) were
studied in both ambient air and in EAN in order to understand the different
types of interactions involved between the surfaces and how EAN
lubricates the contact. In Figure 11a is presented a normalized force curve
for the silica-PTFE system in air, and in Figure 11b the frictional force as a
function of applied load for the same system. When the tip is getting close
to the surface (around 5 nm), a jump-in is observed that is characteristic of a
van der Waals interaction; when the tip is moved away from the surface,
adhesive forces keep the surfaces together until this interaction is overcome.
The fitting of the approach curve by using Eq. 1 (Figure 11a, inset)
confirms the van der Waals nature of this interaction.
28
Figure 11. Normalized silica–PTFE force curve in air (a) on approach (closed symbols) and
retraction (open symbols). The inset in (a) shows five normalized force curves for the same system
on approach fitted with van der Waals theory. (b) Friction as a function of applied load in air for
silica-PTFE (Paper I).
The friction response of the silica-PTFE system is also affected by the
adhesion forces observed in Figure 11a. Figure 11b presents large friction
values, where the large friction at zero applied load is produced by the
significant adhesion that acts as an extra load (see Section 1.3). The friction
coefficient extracted from the gradient of the friction-force data in the linear
part of the Figure 11b is constant and thus does not depend on the applied
load or frictional force according to Eq. 7. However, a careful examination
of the figure reveals that this linear regime is broken at low applied loads.
This is expected for PTFE because it is much softer than silica and is
deformed after contact, changing the real contact area. As a consequence,
the friction force becomes dependent on the area of contact which is
increasing due to contact mechanics (see Section 1.3).
The hydrophobicity of PTFE makes the silica-PTFE and alumina-PTFE
systems representative of minimal adhesion because of the absence of
capillary forces that greatly contribute to adhesion forces. However, the
other systems studied in air without PTFE (silica-silica and alumina-silica
in Paper I) presented strong capillary forces, leading to high adhesive and
friction values which depends on the ambient conditions, such as relative
29
humidity. The differences of the systems studied in air were highly
influenced by the materials chemistry, the environment (humidity and
temperature), and the number/sequence of measurements.
Figure 12. Normalized force curves on approach (closed symbols) and retraction (open symbols)
at an approach rate of 100 nm/s (Paper I).
Figure 12 shows that with the immersion in EAN, van der Waals and
capillary forces in every system were shielded because of the screening
effect of the IL. The systems studied in EAN showed no adhesion with
friction forces and friction coefficients lower than in air (Figure 13). Since
these systems show almost no adhesion at zero applied load, they follow a
more classical Amontons’ law (Eq. 5). The differences observed between
the systems immersed in EAN were mainly influenced by the sample/probe
roughness.
30
Figure 13. Friction measurement between a silica colloidal probe and a silica surface in EAN
(Paper I).
Previous works have been applying adsorbed thin films of ILs,[64-67] but
in this work the systems were immersed in the IL. This modification
eliminates potential problems, such as, film depletion and atmospheric
water absorption.[68]
EAN is a highly viscous liquid, and hydrodynamics (fluid dynamics in this
case) thus need to be addressed in the surface interactions. Therefore, in
Paper I, an investigation has been performed in order to quantify this
effect. Figure 14a shows normal force measurements, on approach, between
a silica probe and a silica surface with different scan speeds. They present a
relatively long-ranged repulsive force that increased with increasing
approach speed. This effect was not observed in the silica-silica force curve
presented above in Figure 12a because of the low scanning speed (100
nm/s). At high velocities this hydrodynamic resistance controls the
interaction and provides an extra barrier against contact, which from a
tribological point of view, adds an extra barrier to avoid adhesion and
decrease friction. On the other hand, the opposite effect occurs on retraction
where an attractive force is generated because of the viscous resistance to
flow into the contact area. These forces during approach and retraction
increase monotonically with the ramping speed.
31
This hydrodynamic resistance can be minimized by decreasing the ramping
velocity of the AFM scanner which controls the speed of approach. For
very slow approaches, where the hydrodynamic force is small (Eq. 4) it is
possible to detect steps in the normalized force curve in the inset of Figure
14a that correspond to structural layering of the EAN ion pairs, as shown
previously.[62, 69-72] These steps are separated by 0.5 nm which
corresponds to the IL ion pair diameter.[69] Thus this can be seen as a type
of solvation force where the ion pairs act as solvent molecules (see Section
1.2.3). However, these steps produced by EAN are not visible at high rates
because they become masked by the much larger and more long ranged
hydrodynamic force.
To verify the nature of the hydrodynamic forces in the Figure 14b, a
retraction force curve was fitted by Eq. 4, allowing the viscosity to be
extracted (all the other parameters in Eq. 4 are fixed. Details of the fitting
procedure are given in Paper I). The inset of Figure 14b shows that the
model describes the experimental data remarkably well but the viscosity
obtained by the fit is 12 mPa·s, which is much lower than the bulk viscosity
of EAN at room temperature (32 mPa·s).[61]
The interpretation of this astonishing result is provided by the consideration
of a lamellar ordering of the EAN at the surface which may offer a lower
resistance to sliding due to well-defined shear planes.
32
Figure 14. Normalized forces curves on (a) approach and (b) retraction between a silica probe and
a silica surface in EAN for different ramping velocities. The inset in (a) presents steps with the
EAN ion pair dimensions obtained at 12 nm/s. Inset in (b) shows the fit using Eq. 6 in the
retraction experimental data at 4360 nm/s (Paper I).
4.2 A Hybrid Route for the Determination of Torsional Spring
Constant for AFM Cantilevers
For the determination of the normal and frictional forces in Paper I, III and
IV, it was necessary to transform the photodiode voltages (which are
related with the deflection and twist of the cantilever) into friction force
through the normal and torsional spring constants of the cantilever, (kz and
kθ, respectively, see Sections 2.1.2 and 2.1.3).[33, 40, 73] However, due to
the high stiffness of the cantilevers used, it was not possible to apply the
Sader method for the determination of the torsional constant. As the rigidity
of the cantilever increases, the signal/noise ratio of the torsional resonance
33
peak decreases, and at certain rigidity, the peak cannot be found. Therefore,
for very rigid cantilevers the Qθ and fθ values cannot be extracted and the
torsional resonance frequencies cannot be calculated. A new approach is
thus needed for the determination of kθ for rectangular rigid cantilevers that
does not rely on the determination of Qθ and fθ.
The determination of these spring constants for rectangular cantilevers can
be performed by using theoretical approximations, such as standard beam
theory,[74, 75] where the spring constant is related to the beam dimensions
and the Young’s (E) or Shear (G) moduli of the cantilever material,[76] and
t is the thickness of the cantilever:
>[\%0 = &]^_W7] (16)
>[\%0 = &]^`a7 (17)
Another approach is to consider experimental approximations, such as the
Sader method (see Section 2.1.6.1), which are preferred because the
cantilevers are very thin and it is difficult to obtain the thickness accurately
or to know how the thickness varies. Furthermore, there is no guarantee that
the moduli of a micron thickness cantilever should have the same as the
bulk. Besides, the standard beam model for the calculation of kθ does not
consider the inherent restraint on axial warping. This term, which becomes
important for short cantilevers,[44] has been considered in a derivation
from plate theory:[44]
>b%&\ = &]^`a7 c1 − (^&%'Adefga,Whi/j7ga,Whi/ )kl
(18)
34
Eq. 18 takes into consideration the above issue in the theoretical treatment
and improves the accuracy of Eq. 17 but the other problems still remain.
Therefore, an appropriate experimental calibration technique is needed to
determine the spring constant of stiff cantilevers.
In Paper II, a hybrid route has been developed, where the normal Sader
method (nSm) to obtain kz is combined with the theoretical approximation
based on the plate theory for the determination of the torsional spring
constant for stiff cantilevers. This method, which is referred to as the
Hybrid model, does not need to extract any parameters from the power
torsional thermal spectra and does not depend on a measured cantilever
thickness. The hybrid method was obtained by combining Eq. 18 with Eq.
16 leading to Eq. 19:
>!"T$# = VW7`a_ c1 − (^&%'Adefga,Whi/j7ga,Whi/ )kl
(19)
To validate the above derived equation for the Hybrid method, it was tested
it against the torsional Sader method (tSm) for several cantilevers with
different dimensions, and with and without coatings.
35
Figure 15. Comparison between the torsional spring constants (>!"T$#) calculated by the Hybrid
method with respect to the torsional spring constants calculated by the established Sader method (>m%#\$), with dashed linear fit. The solid line in the figure corresponds to a perfect agreement
with the torsional Sader method (Paper II).
Figure 15 displays a comparison between the torsional spring constant
calculated by the tSm and Eq. 19 where the data are obtained from a set of
beam shaped monocrystalline silicon cantilevers, both coated and uncoated
(see Paper II). The error bars were obtained from the standard deviations
of three consecutive measurements. This error was omitted for clarity when
the error bar was smaller than the symbol. The solid diagonal line in the
figure corresponds to the case where both methods return the same values,
and the almost indistinguishable dashed fit is the linear regression of
>!"T$# with respect to >m%#\$. Standard deviations of kθ for stiffer
cantilevers were larger than for softer ones because the amplitude/noise
ratio of the torsional resonance peaks becomes smaller for more rigid
cantilevers. Besides, there is a larger error in the x direction than in the y
direction that is generated by the greater uncertainty in the tSm method,
especially for the more rigid cantilevers. The only required factor in Eq. 19
is the ratio of G/E which is used as a fitting parameter such that the dashed
fit is as close as possible to the solid line in Figure 15. The 3.0 value
36
provides the best fit, but the fact that the results are linear is an indication
that the approach is correct.
It could also be considered that the quality of the fit in Figure 15 should be
affected by the surface coating of the cantilever, however a linear fit was
obtained, irrespective of whether coated cantilevers were used or not.
Therefore, this coating effect appears to be negligible or to be adequately
addressed in the measurement of kz.
In this work, it has been established a working value of G/E for a family of
cantilevers by comparing >!"T$# with >m%#\$. It can be seen that this value
is general and thereafter, can be safely used in conjunction with the kz to
obtain the torsional spring constant, irrespective of whether the cantilever is
stiff or not. This agreement between the hydrodynamic approach and the
mechanics calculations provides further confidence to the applicability of
this model, not only for stiff, rectangular, silicon cantilever, but also for any
other type of rectangular beam cantilevers.
4.3 A New Technique to Characterize Heterogeneous Surfaces:
Tribological Property Mapping
The friction constant of a material depends on the material properties and
through them the strength of the adhesive force at contact, for example the
magnitude of the van der Waals force.[77] In Paper I the effect of
screening the van der Waals force completely through the use of an IL was
clearly observed in the magnitude of the friction coefficient. In that paper
homogeneous materials were used, but in an application, a heterogeneous
material is usually used where the properties of the surface can vary
significantly from place to place, reflecting the local composition. Often, as
in a tool steel, the various properties can be harnessed in the performance of
37
the material. Thus, it is useful to be able to map this local variation in the
property of a system.
Scanning probe techniques are used for the characterization of material
properties at the nanoscale, and lateral atomic force microscopy (see
Section 2.1.3) has long been a part of this armoury. As the scanning tip
transverses the sample surface in LFM, the twisting of the cantilever
generates a friction force that depends not only on the applied load but, also
depends on the local adhesion between the tip and the surface, which varies
with the surface composition and microstructure. Therefore, the friction
force depends on the tip, surface and ambient conditions. On the other hand,
the friction coefficient is a more useful parameter to compare lubrication
properties between different systems (see Section 1.3) but is not obtained
directly from LFM measurements.
LFM can be used to map different surfaces[78-83] providing qualitative
friction images (by sliding the tip across the surface). However, quantitative
measurements are more difficult but possible to obtain by transforming the
lateral photodetector output to friction force (see Section 2.1.3). Both of
these measurements provide relevant information of the sample-tip
interactions, but to map fully the tribological response of the sample, the
adhesion and friction coefficient are required for each pixel of the image.
Therefore, tribological property mapping (TPM) has been developed in
Paper III for a complete tribological analysis of the surface, where a set of
images containing information about the twisting of the cantilever at
different applied loads are transformed into two new images in which each
pixel contains friction coefficient and adhesion values, respectively. In this
study, the commercial powder metallurgical tool alloy, Vancron®40, was
employed because its heterogeneous microstructure allows the
identification of local tribological properties.
38
Figure 16. Friction images of Vancron®40 in air obtained from trace (a, d, g), retrace (b, e, h) and
average (c, f, i) friction data applying a load of 15.19, 80.29 and 145.39 nN. The insets show the
friction values vary through a cross section of 3.7 µm and the symbols x and + correspond to two
positions with lower and higher friction contrast respectively (Paper III).
Figure 16 presents LFM trace, retrace and average friction force obtained at
three different applied loads, where the average friction was created by the
combination of the trace and retrace data (Paper III). Afterwards, these
average friction images obtained at different applied loads are employed to
generate the friction coefficient and adhesion maps. The line sections
inserted in every image in Figure 16 demonstrate how the friction contrast
varies locally with the applied load, indicating that there are different
regions with significantly different frictional properties. This structure of
particles embedded in a matrix is consistent with previous studies, which
39
show that Vancron®40 has a heterogeneous microstructure of fine carbide
and carbide-nitride particles inserted in the alloy matrix.[49, 50]
The images in Figure 16 only provide frictional information at the applied
load that they were obtained. However, it is unknown what would occur at
other applied loads and neither could we predict how the friction would
vary with load. Therefore, it is more useful to obtain images containing
tribological information which are not related to the applied load. With
these images, it is possible to describe fully the frictional properties of the
studied system and predict their variation with load.
To show how these images are obtained, the calculation was simplified to
two positions on the image, one located over a particle (+) and another over
the matrix (x), which display higher and lower friction contrast,
respectively. The average friction forces were extracted at each of the two
pixels for all of the seven images obtained at different applied load and
plotted with respect to the applied load (Figure 17).
Figure 17. Effect of applied load on friction force, at individual pixels. The filled symbols
correspond to the particle (+) and the empty symbols to the matrix (x) (Paper III).
Afterwards, linear regressions were applied according to Eq. 6 (see Section
1.3) and the friction coefficients and the adhesion values were obtained.
These two parameters provide load-independent tribological information
related to the local microstructure of the two positions analyzed on the
40
images. In Paper III, such treatment was automatically applied to every
pixel-position on the seven images, and two maps containing friction
coefficients and adhesion values were generated.
Figure 18. Vancron®40 surface scanned with a silicon cantilever in air, (a) height, (b) friction
coefficient and (c) contact adhesion (Paper III).
Figure 18a is a topographic image obtained in contact mode performed in
air (19% relative humidity), which presents the microstructure of
Vancron®40. It is composed of a matrix containing particulate features with
a diameter smaller than 2 µm. These tops of the particles are located both
above and below the matrix surface. Figure 18b is the friction coefficient
image obtained by TPM for the same area as in Figure 18a, where the
features show a larger friction coefficient than the matrix. It appears that
whether or not the particles protrude has no effect on the friction coefficient
values. Therefore these particles, with friction coefficient values ranged
between 0.65 and 0.85, may have the same or similar kind of surface
composition (note that composition differences for such materials are
nominal, and the composition of the both the matrix and the particles can
vary within the same sample).[49, 50] The matrix displays a lower friction
coefficient ranged between 0.40 and 0.70.
The other contribution of TPM is the adhesion image shown in Figure 18c.
This adhesion is obtained without actually separating the surfaces from
contact and for convenience this value is referred as “contact adhesion”
(Paper III). In this case, the correlation with Figures 18a-b is possible but
41
not easy because the contact adhesion values for both the particle and
matrix against the silicon tip in air are very similar, around -40 nN. This
result is not unexpected because the surface is exposed to ambient relative
humidity (19% in this case), and when the tip comes in contact with the
surface, a capillary bridge forms due to vapour condensation, increasing the
adhesion contribution. This contribution is expected to be relatively similar
for all the surface components and thus independent of the scanning
position. The magnitude is also expected to be comparable to, or larger
than, the influence of other types of surface interactions, such as van der
Waals forces.[25] To confirm this contention, normal force measurements
were performed between a silicon tip and the Vancron®40 surface in air. A
result is displayed in Figure 19 where the jump in and jump out in the force
curve are consistent with this type of interaction. Note that the values of the
conventional adhesion extracted from force curves such as that in Figure 19
are slightly lower than the contact adhesion obtained from Figure 18c
because of the underestimation of the adhesion expected in the traditional
pull off measurements. The lifetime of the capillary condensate is also very
different in the two cases, and vapour harvesting[25] may have contributed
to a larger condensate with greater adhesion in the TPM images.
Figure 19. Normal force curves on approach and retraction between a silicon tip and a
Vancron®40 surface in air.
42
Thus it is very important to measure the humidity when the measurements
are performed in to order to compare tribological measurements, - this is
true for any system where adhesion is important, irrespective of the
technique.
On the other hand when the system is immersed in liquid, capillary
condensation may not occur and the sole contributor to adhesion is most
likely to be the van der Waals interactions between the tip and the surface.
Therefore, TPM was performed between Vancron®40 and a silicon
cantilever in tetradecane. Figure 20a shows again a topographic image, and
the surface microstructure is very similar to that in Figure 18a. Figure 20b
displays the friction coefficient image obtained once again by TPM. As for
Figure 18b the particulate features show a larger friction coefficient than the
matrix, but in this case, the values are lower than those in air due to the
lubricating contribution of tetradecane. Thus, the particles display friction
coefficients ranging between 0.25 and 0.55, and the matrix displays values
between 0.25 and 0.45. The friction coefficient map also shows less noise
than the previous measurement in air probably because of the absence of a
capillary condensate during the scanning.
Figure 20. Vancron®40 surface scanned with silicon cantilever in tetradecane, (a) height, (b)
friction coefficient and (c) contact adhesion (Paper III).
In contrast to Figure 18c, the contact adhesion image in Figure 20c shows
heterogeneity. The absence of the capillary condensate leads to contrast
between the matrix and the particulates, which can be clearly distinguished
43
in the image. The contact adhesion values observed on the asperities ranged
between -10 and -40 nN and on the matrix between -15 nN and -25 nN. The
considerably lower adhesion values in Figure 20c and the jump in and jump
out of the force curve in Figure 21 reflect the absence of the condensate.
Thus the major contribution to the contact adhesion is from other forces,
such as van der Waals which depend upon the local composition of the
sample.
Figure 21. Normal force curves on approach and retraction between a silicon tip and a
Vancron®40 surface in tetradecane (Paper III).
4.4 Biomedical CoCrMo alloy: A Tribological Properties Mapping
Study
TPM[20] was applied once again to another heterogeneous material in
Paper IV. In this case a biomedical CoCrMo alloy was selected because of
its special properties such as corrosion and wear resistance, good
mechanical properties, and biocompatibility with the human body, and the
intermediate medium was PBS because it mimics the saline conditions of
the body (see Section 3.3). Under these conditions, the local tribological
properties could be studied and the stability of the alloy in similar
biological environment to which, for example, hip-joint replacements are
exposed during their working life.
44
The biomedical CoCrMo alloy is formed by two different types of carbides
embedded in a Co-based matrix according to the atomic composition shown
in Table 4. These two types of carbides are observed in the backscatter
SEM image in Figure 22, where the bright areas correspond to the M6C
carbides with heavier elements (9.70% molybdenum) and the darker areas
are the M23C6 carbides and matrix with lighter elements, (2.26% and 3.80%
Mo respectively).[52] Both carbides have face centred cubic structure
(FCC), but the M6C carbides with a diamond like crystalline structure Fd-
3m[84] are the hardest part of the biomedical alloy.
Table 4. Element composition (at. %) of the respective phases obtained from energy dispersive
spectroscopy (EDS) and transmission electron microscopy (TEM).[52]
Element (at. %) Cr Co Mo Mn Ni Fe C Si
M6C 18.66 17.91 9.70 - - - 44.40 9.33
M23C6 37.42 6.45 2.26 - - - 53.87 -
Alloy matrix 31.20 63.17 3.80 0.52 0.21 0.21 0.98 -
Figure 22. BSE-SEM image of a CoCrMo alloy, displaying the matrix and the two different types
of carbides.
Figure 23a shows a topographic image of the studied medical alloy in
contact mode performed in PBS after one hour of immersion. It indicates
45
particulate features contained inside the matrix that protrude above the
surface. Two sets of particles with different heights are observed in the
image. During the polishing of the sample the harder, more resistant part of
the metal abrades less than the softer parts, thus the features of harder
particles are higher than the softer matrix. Figure 23b corresponds to the
same area as Figure 23a and it displays a TPM image of friction coefficient
variation. The particles can be clearly distinguished from the matrix. TPM
shows clearly that the particles, which might be speculated as being of
different hardness due to their different heights, are in fact of very different
nature. One type of particle displays the highest friction coefficient values
(between 0.22 and 0.33), and the other type of particle presents the lowest
friction coefficient values in the image, ranging between 0 and 0.15. Thus
the matrix has frictional properties that are intermediate between the two
types of particles described above. The contact adhesion information is
presented in Figure 23c, where particles with the highest friction
coefficients in Figure 23b also show higher contact adhesion values ranging
between 10 and -20 nN. On the other hand, particles with the lowest friction
coefficient also present the lowest adhesion, with values that range between
25 and 5 nN. The matrix displays intermediate contact adhesion values
between 15 and -5 nN. The contact adhesion according to the definition in
Section 1.3 is the intercept of the friction force data with the load axis. If
the surfaces experience an adhesion then the value is a negative load, but
there is no reason why a “negative adhesion” i.e. an intercept at a positive
load, cannot occur. This corresponds to the case of a repulsive force which
must be overcome before the surfaces can achieve contact and dissipate
frictional energy.[85] I clarify that a true adhesion is denoted for TMP by
negative numbers, and a more negative value corresponds to a more
attractive interaction. Positive values thus correspond to repulsion.
46
Figure 23. Si - CoCrMo alloy images in PBS, (a) height, (b) friction coefficient and (c) contact
adhesion.
TPM provides relevant tribological information of the CoCrMo alloy but in
order to understand why differences phases have different tribological
responses, it is necessary to study the influence of the surface
microstructure of this alloy.
The chromium content is a very important parameter that affects the
tribological properties of the studied alloy. When Cr is exposed to oxygen,
it readily oxidizes forming a chromium oxide layer, which decreases the
friction coefficient and the contact adhesion. The good tribological
properties of the chromium oxide were attributed in previous work[86] to
the poor wettability and high hardness of the thin oxide formed on a
chromium coated steel tool. Moreover, the Cr content at the interface is
slightly enhanced with respect to the bulk due to its preference to form a
thin oxide.[87] Therefore, the phases that displayed higher friction
coefficients and contact adhesions in Figures 23b-c can be identified as the
M6C carbides because of their low chromium content (Table 4). On the
other hand, the phases with lower friction coefficient and adhesion values
correspond to the M23C6 carbides with more than 37% chromium content.
This identification agrees with the previous speculation about the relative
hardness, where the M6C carbides with a diamond like hardness will
protrude more than the M23C6 carbides which abrade more during the
polishing. As a conclusion, it is more likely that the thin chromium oxide
47
formed on the biomedical alloy surface is responsible for the good
tribological properties of the material. Moreover, the M23C6 carbides with
the highest chromium content generates the more effective oxide layer
displaying the lowest friction coefficient and contact adhesion values in
Figures 23 b-c.
Another parameter that affects the tribological response is the intervening
medium. After one hour of immersion in PBS, TPM was performed and no
indication of corrosion or alteration of the biomedical alloy surface was
observed. This supports the inertness of the CoCrMo alloy within this type
of biological environment.[54]
Figure 24. Normal force curves on approach (closed symbols) and on retraction (open symbols)
between a silicon tip and a CoCrMo alloy surface in PBS. The continuous line in the approach
curve represents the fit according to Eq. 20. The inset shows an enlargement of the data.
Figure 24 shows the surface force at an arbitrary point in PBS. As the tip
approaches the surface, it experiences first, from around 7 nm, an attraction
attributted to a van der Waals force (see Section 1.2.1). Afterwards, an
attraction is overcome by a hydration force (see Section 1.2.4), generating a
repulsion around 2 nm of separation distance. A threshold-force of about 15
nN is needed to reach the hard contact region (constant compliance), which
corresponds to the range of positive values (repulsive) observed in Figure
23c and explains why the contact adhesion values are repulsive in this case.
The fitting in Figure 24 is not perfect, but verifies that the forces between
48
the tip and the biomedical alloy are likely explicable by the superposition of
van der Waals (Eq 1) and hydration forces (Eq 3) according to Eq. 20:
(() = #n + !"#$%&' (20)
where the fitting parameters are A = 3x10-20 J, λ = 0.2 nm and CH = 3.7 N.
On retraction, the superposition between hydration and van der Waals
forces generated a light repulsion below 2 nm and an attractive interaction
between 2 and 8 nm. Therefore, the system formed by the silicon tip and
the CoCrMo alloy in PBS experiences repulsive interactions mostly because
of hydration forces, which generate a decrease in the friction coefficient and
contact adhesion values in Figures 23b-c. The effects of these forces are so
important than even positive contact adhesion values (repulsion) are
observed in Figure 23c, especially in the case of the M23C6 carbides.
4.5 Surface Study and Corrosion Initiation of an Experimental
FeCrVN Tool Alloy
Mechanical properties describe how materials deform when they are
exposed to external forces. These mechanical properties can be combined
with tribological information in order to understand the microstructure of a
material as well as shed light onto surface phenomena such as corrosion.
The aim in Paper V was to study a FeCrVN stainless steel tool alloy by
PeakForce QNM® in water and in NaCl (0.1 M) in order to understand the
influence of microstructure on corrosion initiation of the tool alloy.
49
Figure 24. (a) Back-scatter SEM image of FeCrVN experimental tool alloy and (b) Volta potential
image of the alloy.
Figure 24a is a back-scatter SEM image of the FeCrVN alloy showing its
heterogeneous structure. The alloy is formed by particles of 0.5-3.0 µm
enriched in chromium and molybdenum, which are harder than the iron-
based martensitic matrix that contain them, and are referred to in this work
as nitride particles. There are two types of particles embedded in the matrix
and their compositions are displayed in Table 5. The darker ones in Figure
24a have more vanadium and nitrogen, and the lighter ones have more
chromium and molybdenum. For simplicity, I will further refer to the two
types of hard phase particles as Cr-V rich and Cr-Fe rich nitrides.
Table 5. Average of at least five analysis points showing the EDS chemical analysis of the matrix
and the two different particles that form the alloy (lighter and darker nitrides).
Element (wt. %) Cr V N Mo Fe
Darker nitride 34.1 ± 0.3 33.9 ± 0.7 18.9 ± 0.5 0.8 ± 0.1 13.3 ± 0.5
Lighter nitride 38.2 ± 0.7 5.4 ± 0.3 8.4 ± 0.2 2.2 ± 0.3 43.3 ± 0.8
Alloy matrix 18.7 ± 0.5 1.2 ± 0.3 2.7 ± 0.4 1.7 ± 0.6 73.1 ± 0.5
Figure 24b is a Volta potential image obtained from a Kelvin Force
Microscopy (KFM)[88] where the nitride particles have higher Volta
30 μm
a) b)
5 μm
50
potential than the alloy matrix. Differences are also observed between the
particles in Figure 24b, where Cr-V rich nitrides present higher Volta
potential than the Cr-Fe rich nitrides. Therefore under corrosion conditions
the latter ones would likely show more tendency to corrode[55] because
corrosion typically starts where the Volta potential is lower, i.e., lower
relative nobility.[55, 89-91]
Figure 25a presents a topographic image taken in PeakForce® mode in MQ-
water after one hour of immersion, showing the microstructure of the metal
alloy; a matrix containing particles with diameters below 3 µm located
above and below the matrix surface.
The adhesion image in Figure 25b displays the adhesion force generated by
interactions between the metal alloy and the silicon tip, where the matrix
displays high adhesion with values between 25 and 28 nN. The particles
contained in the matrix present two different adhesion values, one very
similar to the matrix, and other with values between 19 and 25 nN.
Furthermore, there are regions around the particles displaying lower
adhesion. All the differences in adhesion observed in Figure 25b suggest
that there are different compositions in the alloy which agree with the
microstructure described below in Figure 24 and Table 5.
51
Figure 25. Peak Force® QNM images of FeCrVN tool alloy using a silicon tip, in water (a-b) and
NaCl solution, 0.1 M (c-h). This series of images in NaCl display how the small particles are
gradually generated in the salt solution, until it is impossible to continue imaging due the
streakiness presented in (g and h), which is typical of particles adhering to the tip. The closed and
open white circles correspond to Cr-V rich and Cr-Fe rich nitride particles, respectively.
52
Vanadium and nitrogen content in the tool alloy are important parameters to
consider because they form vanadium nitride (VN), which is characteristic
of low adhesion.[92] The origin of this low adhesion is still not fully
understood, but it has been reported that VN is prone to lower adhesion
than TiN[93] and the surface VN oxide formed at high temperature leads to
crystallographic planes that are easy to shear.[94] Therefore, it is most
likely that the particles with lower adhesion in Figure 25b correspond to Cr-
V rich nitrides and the particles with higher adhesion to Cr-Fe rich nitrides.
In relation to the lower adhesion observed around the rich nitrides in Figure
25b, a possible explanation relates the low adhesion in these corona regions
with a different microstructure or a higher surface charge than in the rest of
the alloy. This effect is not fully understood, but is probably related to the
transition in composition between the particles and the matrix.
The tool alloy surface was immersed in a NaCl solution (0.1 M), and after a
waiting time of 60 minutes the sample was scanned three times with
PeakForce® QNM mode, where every image took 8 minutes. After this first
scan, Figure 25c shows a surface topography almost identical to the earlier
case in water (Figure 25a), but small particles appear on the surface mostly
on the matrix and around the nitride particles (especially, around the Cr-V
rich nitride particles). In contrast, the adhesion data (Figure 25d) present
important changes where the adhesion decreased dramatically in the three
regions. This reduction was larger for the matrix where the adhesion
decreased dramatically from 19.98 to 0.95 nN, becoming almost
indistinguishable from the Cr-V rich nitride particles.
Figures 25e-f show that the topography and adhesion maps generated after
the second scan are rather similar to the previous case (Figures 25c-d).
However, in Figure 25f the contrast difference between the Cr-V rich
nitride particles and matrix decreased even more, and specific areas located
53
in the matrix and around nitride particles display very low adhesion -
around 1 nN.
After a further waiting time of 22 minutes (total immersion 82 minutes), the
surface was scanned a third time, and Figures 25g-h present important
differences with respect to Figures 25e-f. First, the horizontal artifacts
displayed in Figures 25e-f leads to loss of quality on the images. Second,
the overall adhesion in Figure 25f decreases, likely as a consequence of the
artifacts observed on the images. In all the images obtained in NaCl
solution, small spot features are observed on the surface (Figures 25c, e and
g), which are mostly on the matrix and around the nitrides (especially
around the Cr-V rich nitride particles). After 90 minutes of immersion in
NaCl, the probability of attachment of the spots to the cantilever tip
increases which prevents imaging. The generation of horizontal artifacts in
imaging due to loose particles attaching to the tip is well described in the
literature.[95]
With the NaCl solution in the system, new forces may arise contributing to
the adhesion, which generally decrease in Figures 25d and f. This is likely a
consequence of hydration forces originated by cations absorbed on the
surfaces. To understand these changes, the surface chemistry of the silicon
tip and the alloy has to be considered. Both surfaces oxidize in contact with
oxygen, and the oxide evolves after water exposure.
Silicon forms silicon oxide with oxygen exposure. Afterwards it gets
hydrolyzed by water and forms silanol groups on the surface. These groups
ionize, generating a negative charge on the surface according to:[96, 97]
:opq + q@p ↔ :op + qaps
54
Therefore, the oxide layer on the silicon tip is negatively charged, and when
it is immersed in the NaCl solution it attracts and accumulates hydrated Na+
ions.
The alloy surface has also a native oxide-like passive layer which is formed
in contact with water. However, the structure of this native oxide layer is
different than the silicon one, and is formed by a gel-like structure with a
large amount of bound water with different bridging structures.[87, 98] In
NaCl, the Cl- ions start to attack weak sites in the native oxide, creating
metal-salt or corrosion products,[99] which can be assumed to be negatively
charged and thus also attract and accumulate Na+ ions.
During a PeakForce® measurement, the tip and the alloy achieve contact
and the interaction between the hydrated Na+ ions accumulated on the
surfaces generates a repulsion or a hydration force[2, 12, 100, 101] that
could be responsible for the decrease in adhesion observed in Figures 25d
and f. This adhesion force appears to be region specific, and thus related to
the microstructure of the experimental tool alloy.
The corrosion of this tool alloy generally leads to metal dissolution and
pitting formation.[55] However, in this work, small protrusions produced
by metal salts or corrosion products appear on the surface, which are mostly
correlated with small spots of very low adhesion displayed in the adhesion
images. It is hypothesized that these areas are related to pre-pitting events
due to the action of the NaCl solution that generates the formation of
probably some kind of oxides, hydroxides, and/or chloride complex
compounds.
Chromium and nitrogen are two elements that are commonly used to
enhance the passivity of stainless steel[87, 102, 103] leading to a high
corrosion resistance. The matrix is the region of the tool alloy where the
chromium content is lower in the passive layer, thus this area is weaker
55
against Cl- attack,[55, 87, 102, 103] and has more tendency to form metal
salts or corrosion products. Moreover, these protrusions appear to
concentrate around the nitrides which correspond to regions in the matrix
with deficiency of Cr and N (Figure 25f).[55, 87, 104-106] Likely, pre-
pitting events preferentially take place in these areas, resulting in higher
concentration of negatively charged corrosion products that lead to stronger
hydration forces and thus lower adhesion, as described previously.
However, a new study is necessary to verify this hypothesis where the
challenge is the in-situ chemical analysis of such small features on the tool
alloy surface.
Adhesion, topography and chemical composition provide us with relevant
information about microstructure and phenomena such as corrosion, but
when deformation and Young’s modulus were expected to provide
additional information, a difficulty arose. PeakForce® QNM provides the
Young’s modulus of a material by deforming the studied substrate with the
cantilever tip applying a threshold force. This threshold will increase with
the hardness of the studied material. Therefore, by using an AFM tip softer
than the studied surface, the tip tends to deform and even to break (Figure
26) without reaching the necessary threshold force to deform the surface.
Consequently, no relevant information based on the Young’s modulus and
deformation of the tool alloy was obtained by using an AFM silicon tip in
PeakForce® QNM mode.
56
Figure 26. SEM image of a silicon tip after (a) and before (b) a PeakForce® QNM measurement
applying high force.
4.6 Nanomechanical Properties of Human Skin
The limitation for obtaining the Young’s modulus and deformation faced in
Paper V is of course lifted for softer materials. For unambiguous studies
the material of the AFM tip should be harder than the surface scanned. The
rigidity of the chosen cantilever has to be selected carefully because first,
the applied force by the AFM tip has to be high enough to induce a relevant
deformation on the surface. Furthermore the cantilever laser-reflection must
always be located in the linear regime of the AFM-photodetector, otherwise
the load applied would incorporate a high error. Therefore, it is necessary to
use a cantilever made of a harder material than the sample, where the
cantilever should be of the right rigidity.
The stratum corneum (SC - see Section 3.3) has mechanical properties that
contribute significantly to its special functions as the outer protective layer
such as skin barrier and photoprotection. Of particular interest are
biointeractions, for example how much deflection is caused as a human hair
interacts with the skin, but there are limited data available in the area.[56-
58] In order to gain a better understanding of the mechanical properties of
57
the SC, surface indentation and PeakForce® QNM measurements have been
performed.
Indentation[107-109] is a commonly used technique to study the
mechanical properties of materials (such as hardness and elastic modulus)
where the load and the indentation depth are obtained simultaneously
during the load and unloading process. Indentation can be designed for use
with very small indenters, applying very small loads in the range of
nanometres, causing nanometre-indentation depths. In this way was born
nanoindentation[109, 110] where the atomic force microscopy became a
very important asset for this type of measurement.
The aim of this study was to investigate the nanomechanical properties and
the magnitude of the force needed to induce elastic/plastic deformation of
the outer layer of skin (SC) by combining AFM imaging and
force/nanoindentaion measurements.
Figure 27. AFM PeakForce QNM images with Topography (a-c) and DMT modulus histograms
(c-e) at scan sizes of 20µm, 10µm and 5µm respectively. The vertical axes in (a-c) indicate the
maximum difference between the darkest and the brightest parts.
58
Figures 27 (a-f) shows the heterogeneous topography of the outer part of
the SC layer obtained after a PeakForce® QNM measurement. The reduced
Young’s moduli (E* in Eqs. 21-22)[111] were obtained by fitting the
contact mechanical theory of Derjaguin, Muller and Toropov (DMT) to the
force curve obtained at each pixel. The extracted values which were
obtained at the nanoscale can be plotted as histograms (see Figures 27 d-f),
where the mean reduced Young’s modulus obtained was 0.51 GPa.
Afterwards, the values of the reduced Young’s moduli were transformed
into Young’s moduli (ESC) for comparison purposes according to Eq. 22
(this equation was derived from Eq. 21[112] by ignoring the contribution of
the silicon because of its much higher relative stiffness):
l_∗ = lνtu_vw + lνt_v (21)
lx∗ = lνtyxty (22)
where the subscript S corresponds to the silicon, SC to the stratum corneum
and ν to Poisson’s ratio (νSC = 0.48). [113]
Therefore, the reduced Young´s modulus of 0.51 GPa was transformed into
a value of 0.39 GPa, which is consistent with the relative high stiffness of
the SC reported in literature.[56, 113, 114]
Afterwards, the deformation of the SC surface was studied by normal force
measurements (nanoindentation) using the same silicon cantilever. First, the
topography of a 1 µm2 region of the surface was obtained. Afterwards,
nanoindentation was performed, and finally the surface was imaged again in
order to detect any type of permanent deformation. This nanoindentation
experiment was performed at different locations by applying 2.5 µN once
and then twice applying 4.8 µN.
59
Figure 28. PeakForce topography images (a-d from left to right) of SC using a sharp tip. (a) 1 µm
scan of neat SC sample (b) 1 µm scans after force indentation using a sharp tip at 2.5 µN (c) 1 µm
scan after force indentation using a sharp tip at 4.8 µN (d) 1 µm scan after repeat force indentation
in a new location using a sharp tip at 4.8 µN. The vertical axes indicate the maximum difference
between the darkest and the brightest parts in the topographic images.
Figure 28a presents the topographic image before indentation and Figure
28b after applying 2.5 µN, and it can be seen that no permanent
deformation occurred. Afterwards, a larger force of 4.8 µN was applied in
two adjacent spots and Figure 28d demonstrates that this nanoindentation
force was enough to produce permanent deformation in both areas with
similar shape and depth.
Figure 29 shows the characteristic normal force curves obtained from the
nanoindentation measurements from which mechanical properties of the SC
can also be extracted.
Figure 29. (a) Force indentation using a sharp tip on approach at a ramp size of 2 µm (1 Hz).
Maximum applied force 2.5 µN (black curve) and 4.8 µN (grey curve). Inset: Indention curve on
retraction for 4.8 µN maximum applied force (grey curve). (b) Indentation profile (4.8 µN)
obtained from the cross section of the topography image that corresponds to the permanent
deformation of the SC (Figure 28c), where the tip image (which corresponds to a SEM image of a
cantilever tip) in inset indicates maximum indentation depth.
60
The black curve in Figure 29a is obtained by applying a load of 2.5 µN,
which causes an indentation depth of approximately 125 nm. This
indentation has mostly an elastic deformation since no permanent
deformation was observed in Figure 28b. On the other hand, the grey curve
generated with an applied load of 4.8 µN, which corresponds to an applied
pressure of 0.47 GPa (Paper VI), presents an even larger indentation depth,
of around 200 nm. Figures 28 c-d show that this value is large enough to
produce a permanent deformation on the SC. On retraction, the elastic
recovery of the SC reduced the indentation depth to the permanent
deformation of around 37 nm observed in the cross section of the
topography in Figure 29b. When the tip contacts a new spot on the SC
surface, the surface may undergo both plastic and elastic deformation. On
retraction the material tries to recover its original shape, but is partially
prevented because of the permanent or plastic deformation. The only
recovery undergone by the material is due to its elastic relaxation, which
corresponds to the apparent indentation depth on retraction.[108] Therefore,
a way to extract the permanent deformation is by taking the difference of
the apparent indentation depths of the force curves measured during
approach and retraction (Figure 29a). This treatment generates a value of 45
nm, which is quite consistent with the 37 nm observed in the cross section
of Figure 29b and thus confirms the elastic-plastic behavior of the SC.
4.7 A Novel AFM Probe: The Single Hair Fibre Probe
Another advantage of AFM is the possibility to use different probes in order
to understand the tip-sample interactions. A material which often interacts
with SC is wool, the fibres of which can cause discomfort. Our own hair, or
that of other individuals, can also cause discomfort and to be able to treat
this issue it is necessary to understand how the SC deforms locally in
response to interaction with such a “probe”.
61
Thus as part of Paper VI, a nanoindentation AFM study was performed
using a novel probe, which was designed from only a single hair fibre
(Figure 30).
Figure 30. SEM images of (a) a single vertical hair fibre attached to the end of a cantilever pre-cut
using sharp scissors and (b) the attached fibre cut after using FIB.
The indentation measurements were initiated by imaging the topography of
the SC sample using a silicon tip. Afterwards, a single force measurement
was performed in the same area using the novel hair probe at an applied
load of 4.8 µN. Subsequently, the surface area was imaged again with the
silicon tip.
Figure 31. PeakForce® topography images. 100 µm2 scans of SC (a) before and (b) after force
measurement using the single hair fibre indenter.
Figure 31 presents the topographic images obtained before and after the
nanoindentation measurement, and a comparison of both images suggests
62
that there is no apparent alteration of the surface. This result might be a bit
unexpected considering the permanent deformation observed in Figure 28,
however there are more parameters to be considered, such as the tip radius
and effective area of contact. Thus, in order to further analyze this
nanoindentation experiment, the force distance profile should also be
considered.
Figure 32. Force indentation (grey curve) on retraction using single hair fibre probe indenter on
SC (ramp size of 2 µm and rate 1 Hz). The black curve represents reference measurement on
retraction of the single hair fibre probe against a bare mica substrate.
Figure 32 displays an example of normal force curves on retraction
obtained during the nanoindentation AFM measurement using the hair fibre
probe. The grey curve corresponds to the measurement against the SC
surface (negative separation corresponds to the indentation depth) and the
black one against a stiff mica surface. The sharp 90º angle in the force upon
contact is characteristic behavior of a stiff non-deformable substrate. The
analysis of the force curve is performed on the assumption that constant
compliance is achieved, and the linearity of the constant compliance region
strongly suggests that the probe does not deform. If it does then the
deformation is pseudo linear and thus the deformation is taken into account
in the deflection sensitivity which is then employed for the interaction with
SC. In contrast, the grey curve shows an indentation depth of around 550
nm and a nonlinearity which is presumably related to the viscoelastic nature
63
of the SC.[115] Comparison of the maximum indentation depths of the hair-
SC force profiles during approach and retraction shows that they differ by
about 164 nm, which corresponds to the plastic deformation that is
observed only during approach (Figure 32). However, in contrast to the
apparent plastic deformation observed in Figures 28c-d, this 164 nm of
plastic deformation are not visible in Figure 31 because the deformation is
spread over a larger area, which reflects the much larger radius of the hair
probe (at least in one dimension). At first, the difference in penetration
depths observed for the different probe diameters is difficult to justify, but it
can be resolved by considering that the Young’s moduli are actually a
function of the interaction size.
Three different combinations of Hertzian mechanical models were assumed
in Paper VI for the approximation of appropriate Young’s moduli for the
hair probe against the SC. The first model (Figure 33a) is based on a sphere
against an elastic half space, where the hair probe is assumed to behave as a
sphere of radius corresponding to the large radius of curvature of the probe.
This model simplifies the situation because it does not account for the large
topographic variation of the SC, and generates a Young’s modulus of 0.98
MPa and a maximum applied pressure of 0.09 MPa. The second (Figure
33b) and the third (Figure 33c) models take into consideration the
topography of the SC, assuming that the hair probe contacts the SC surface,
deforming it at two levels; both individual features and a more general flat
surface of the SC, simultaneously. These interactions can be coarsely
approximated by a sphere-sphere contact (i.e. the hair probe interacting
with a SC asperity) in series with a sphere-flat contact (i.e. the hair probe
interacting with the underlying SC surface). The differences between
second and third models are that in the second it is assumed the same
Young’s modulus for the SC asperity and the SC surface and in the third
they are different (Paper VI).
64
Figure 33. Schematic showing the different Hertzian models applied between the hair probe
(sphere) and SC (surface) before (left) and after indentation (right). (a) A sphere interacting with
an elastic surface, where EH is the Young’s modulus of the hair and ESC2 is a fitting parameter that
corresponds to the Young’s modulus of the SC. (b) A combination of a sphere interacting against
an elastic sphere (SC-asperity), and a sphere interacting against an elastic surface, where ESC3 is a
fitting parameter that corresponds to both the Young’s modulus of the SC-asperity and the
underlying SC surface. (c) A combination between a sphere interacting against an elastic sphere
(SC-asperity), and a sphere interacting against an elastic surface, where ESC1 is the Young’s
modulus of SC-asperity (obtained from Figure 27) and ESC4 is a fitting parameter that corresponds
to the Young’s modulus of the SC.
65
Finally, the second and third models generate Young’s moduli of 2.79 and
1.01 MPa, and maximum applied pressure of 0.70 and 17.7 MPa,
respectively.
It is stressed that there is no good contact model for the geometry
employed, and the above values are not more than an approximation, but
should be of the right order of magnitude. The most important development
here is that it has been identified the scale of applied pressures that are
appropriate to describe hair induced deformation of the SC.
66
5. Conclusions
The combination of a nanotribology approach with corrosion showed that
this synergy can lead to new insights and research avenues. Significant
technique development in atomic force microscopy (AFM) was achieved
which enabled investigation of metal surfaces in a new light. Such expertise
has the ability to be applied over a wide variety of fields-for example in
understanding interactions in ionic liquid, as well as how the stratum
corneum deforms in response to probes of different sizes.
The long ranged interactions between pairs of materials have been studied
by AFM in air and after immersion into ethylammonium nitrate (EAN). In
air the tribological properties depends on the long ranged interactions
present between them (van der Waals and capillary forces), which are
significantly affected by the environmental conditions, material properties
and number/sequence of measurements. However, in EAN the ionic liquid
suppress the long ranged interactions and reduces the friction and adhesion,
which become only affected by the roughness of the surfaces in contact.
When the surfaces were approaching each other (in EAN) with a relative
fast speed, there was a repulsion that precluded contact because of
hydrodynamic forces. The analysis of the hydrodynamic force provided an
interfacial viscosity which was almost a 3-fold reduction compared to the
bulk viscosity. It is assumed that this low viscosity value is due to a
lamellar ordering of the EAN at the interface that reduces the resistance to
sliding.
Tribological property mapping (TPM) is a new technique that generates
friction coefficient and especially contact adhesion maps, by using lateral
atomic force microscopy images. TPM provides tribological information
which can be related to the sample microstructure. This technique has been
67
applied to two different metal alloys showing their heterogeneous
microstructure which is formed by hard particles embedded in a softer
matrix. In case of the CoCrMo alloy, it was shown that the chromium oxide
and the phosphate buffer saline solution are responsible for the low friction
coefficient and contact adhesion.
Another AFM technique, PeakForce® QNM was used in water to study the
surface microstructure of an experimental FeCrVN tool alloy, and in NaCl
solution to understand the influence of microstructure in the corrosion
initiation of this tool alloy. The results show that the low adhesion is
attributed to the vanadium and nitrogen contents, and the corrosion
initiation or prepitting depends on the chromium content. Moreover, this
prepitting originates in specific areas leading to small particles with low
adhesion.
Force and PeakForce® QNM measurements with a Si tip were also used to
understand the mechanical properties of the human stratum corneum (SC).
The SC is heterogeneous at the nanoscale where substancial permanent
deformation occurs at an applied pressure of about 0.47 GPa.
A novel probe was developed by attaching a single hair probe to the end of
a cantilever to perform nanoindentation measurements on the SC sample.
These measurements show that the SC is extremely elastic at the nanoscale,
and that the lateral deformation regulates the effective Young’s modulus.
Three models were proposed in order to calculate the pressure applied for
the novel hair probe in SC.
A new approximation referred as Hybrid method was originally developed
to obtain the torsional spring constants of rigid cantilevers where the power
torsional thermal spectra are difficult to obtain because of the high
resonance frequency and low signal/noise ratio. This method that combines
the normal spring constant with the length and the width of the cantilever to
68
obtain the torsional spring constant can be used for the determination of any
type of rectangular cantilever with or without coating.
69
6. Acknowledgments
First of all, I would like to express my sincere gratitude to my supervisors
Prof. Mark Rutland and Jinshan Pan, for giving me the opportunity to do
my PhD in KTH. Especially to Prof. Mark Rutland for his support and
guidance, as well as, the attitude to find always the positive side of any
situation has really inspired me.
I also want to thanks Emily Cranston and Esben Thormann (co-supervisors)
for patiently teaching me to use the AFM and help me to become the
professional that I am now.
Thanks to my friends Deborah Wakeham and Beatrice Johannson for
helping with my Thesis.
There are so many people who have contributed in their own way to the
completion of my PhD, coauthors, friends, and colleagues. They work not
only at the Division of Surface and Corrosion Science, but also at SP,
Nanologica AB, L´Oreal and Stockholm University. You know who you
are and even I don’t name you will be always with me wherever I go.
I want to especially thank my dear friend Rodrigo Robinson, for being
willing to listen and share complaints and for giving me good advices.
SSF (Swedish foundation for Strategic Research) is gratefully
acknowledged for sponsoring this project.
Alfredo Metere at Molworx (www.molworx.com) is also acknowledged for
the design of Figure 3.
My sweet Linnéa Bengstsson, thank you for your love, support, faith and
patience. I could never have made it without you.
The last and not the least, my family because they have always believed in
and supported me. Os quiero y os echo de menos, siempre estaréis conmigo
esté donde esté.
70
7. References
1. J.N. Israelachvili, Intermolecular and Surface Forces. 3rd ed. 2008: Academic Press.
2. B. Cappella and G. Dietler, Force-Distance Curves by Atomic Force Microscopy. Surface Science Reports, 1999. 34(1-3): p. 1-104.
3. H.C. Hamaker, The London-van der Waals Attraction between Spherical Particles. Physica, 1937. 4(10): p. 1058-1072.
4. A.A. Feiler, L. Bergstrom and M.W. Rutland, Superlubricity Using Repulsive van der Waals Forces. Langmuir, 2008. 24(6): p. 2274-2276.
5. A. Milling, P. Mulvaney and I. Larson, Direct Measurement of Repulsive van der Waals Interactions Using an Atomic Force Microscope. Journal of Colloid and Interface Science, 1996. 180(2): p. 460-465.
6. A. Meurk, P.F. Luckham and L. Bergstrom, Direct Measurement of Repulsive and Attractive van der Waals Forces between Inorganic Materials. Langmuir, 1997. 13(14): p. 3896-3899.
7. S.-W. Lee and W.M. Sigmund, Repulsive van der Waals Forces for Silica and Alumina. Journal of Colloid and Interface Science, 2001. 243(2): p. 365-369.
8. H.-J. Butt and M. Kappl, Capillary Forces, in Surface and Interfacial Forces. 2010, Wiley-VCH Verlag GmbH & Co. KGaA. p. 127-161.
9. Y.I. Rabinovich, J.J. Adler, A. Ata, R.K. Singh and B.M. Moudgil, Adhesion between Nanoscale Rough Surfaces: II. Measurement and Comparison with Theory. Journal of Colloid and Interface Science, 2000. - 232(- 1): p. - 24.
10. Y.I. Rabinovich, J.J. Adler, A. Ata, R.K. Singh and B.M. Moudgil, Adhesion between Nanoscale Rough Surfaces: I. Role of Asperity Geometry. Journal of Colloid and Interface Science, 2000. 232(1): p. 10-16.
11. A.A. Feiler, P. Jenkins and M.W. Rutland, Effect of Relative Humidity on Adhesion and Frictional Properties of Micro- and Nano-scopic Contacts. Journal of Adhesion Science and Technology, 2005. 19(3-5): p. 165-179.
12. J.J. Valle-Delgado, J.A. Molina-Bolívar, F. Galisteo-González, M.J. Gálvez-Ruiz, A. Feiler and M.W. Rutland, Hydration Forces between Silica Surfaces: Experimental Data and Predictions from Different Theories. The Journal of Chemical Physics, 2005. 123(3): p. -.
13. W.A. Ducker, T.J. Senden and R.M. Pashley, Measurement of Forces in Liquids Using a Force Microscope. Langmuir, 1992. 8(7): p. 1831-1836.
14. H.-J. Butt and M. Kappl, Hydrodynamic Forces, in Surface and Interfacial Forces. 2010, Wiley-VCH Verlag GmbH & Co. KGaA. p. 163-187.
71
15. D.Y.C. Chan and R.G. Horn, The Drainage of Thin Liquid Films between Solid Surfaces. J. Chem. Phys., 1985. 83: p. 5311-5324.
16. B. Bhushan, Introduction to Tribology. 2002: John Wiley and Sons. 17. G. Amontons, Proceedings of the French Royal Academy of Science,
1699: p. 257−282. 18. D. Downson, History of Tribology. 1979, London: Longman. 19. B. Derjaguin, Molekulartheorie der Äußeren Reibung. Zeitschrift für
Physik A Hadrons and Nuclei, 1934. 88(9): p. 661-675. 20. R. Álvarez-Asencio, J. Pan, E. Thormann and M. Rutland, Tribological
Properties Mapping: Local Variation in Friction Coefficient and Adhesion. Tribology Letters, 2013. 50(3): p. 387-395.
21. J. Gao, W.D. Luedtke, D. Gourdon, M. Ruths, J.N. Israelachvili and U. Landman, Frictional Forces and Amontons’ Law: From the Molecular to the Macroscopic Scale. The Journal of Physical Chemistry B, 2004. 108(11): p. 3410-3425.
22. A.M. Homola, J.N. Israelachvili, M.L. Gee and P.M. McGuiggan, Measurements of and Relation Between the Adhesion and Friction of Two Surfaces Separated by Molecularly Thin Liquid Films. Journal of Tribology, 1989. 111(4): p. 675-682.
23. M.A. Plunkett, A. Feiler and M.W. Rutland, Atomic Force Microscopy Measurements of Adsorbed Polyelectrolyte Layers. 2. Effect of Composition and Substrate on Structure, Forces, and Friction. Langmuir, 2003. 19(10): p. 4180-4187.
24. G. Bogdanovic, F. Tiberg and M.W. Rutland, Sliding Friction between Cellulose and Silica Surfaces. Langmuir, 2001. 17(19): p. 5911-5916.
25. A.A. Feiler, J. Stiernstedt, K. Theander, P. Jenkins and M.W. Rutland, Effect of Capillary Condensation on Friction Force and Adhesion. Langmuir, 2006. 23(2): p. 517-522.
26. A. Berman, C. Drummond and J. Israelachvili, Amontons’ Law at the Molecular Level. Tribology Letters, 1998. 4(2): p. 95-101.
27. S. Yamada and J. Israelachvili, Friction and Adhesion Hysteresis of Fluorocarbon Surfactant Monolayer-Coated Surfaces Measured with the Surface Forces Apparatus. The Journal of Physical Chemistry B, 1998. 102(1): p. 234-244.
28. C.M. Mate, Tribology on the Small Scale : A Bottom Up Approach to Friction, Lubrication, and Wear. 2008, Oxford: Oxford University Press. xiii, 333 p.
29. S. Izabela, C. Michael and W.C. Robert, Recent Advances in Single-Asperity Nanotribology. Journal of Physics D: Applied Physics, 2008. 41(12): p. 123001.
30. Handbook of Micro/Nano Tribology, ed. B. Bhushan. 1999: CRC Press.
72
31. F.P. Bowden and D. Tabor, Mechanism of Metallic Friction. Nature 1942. 150: p. 197-199.
32. G. Binnig, C.F. Quate and C. Gerber, Atomic Force Microscope. Physical Review Letters, 1986. 56(9): p. 930-933.
33. J. Ralston, I. Larson, M.W. Rutland, A.A. Feiler and M. Kleijn, Atomic Force Microscopy and Direct Surface Force Measurements - (IUPAC Technical Report). Pure and Applied Chemistry, 2005. 77(12): p. 2149-2170.
34. J.R. Smith, C. Larson and S.A. Campbell, Recent Applications of SEM and AFM for Assessing Topography of Metal and Related Coatings - a Review. Transactions of the Institute of Metal Finishing, 2011. 89(1): p. 18-27.
35. M. Sababi, J. Kettle, H. Rautkoski, P.M. Claesson and E. Thormann, Structural and Nanomechanical Properties of Paperboard Coatings Studied by Peak Force Tapping Atomic Force Microscopy. ACS Applied Materials & Interfaces, 2012. 4(10): p. 5534-5541.
36. B. Foster, New Atomic Force Microscopy (AFM) Approaches Life Science Gently, Quantitatively, and Correlatively. American Laboratory, 2012. 44: p. 4.
37. W.A. Ducker, T.J. Senden and R.M. Pashley, Direct Measurement of Colloidal Forces Using an Atomic Force Microscope. Nature, 1991. 353(6341): p. 239-241.
38. H.-J. Butt, Measuring Electrostatic, van der Waals, and Hydration Forces in Electrolyte Solutions with an Atomic Force Microscope. Biophysical Journal, 1991. 60(6): p. 1438-1444.
39. B.V. Derjaguin, Kolloid. Zh., 1934. 69: p. 155-164. 40. T. Pettersson, N. Nordgren, M.W. Rutland and A.A. Feiler, Comparison of
Different Methods to Calibrate Torsional Spring Constant and Photodetector for Atomic Force Microscopy Friction Measurements in Air and Liquid. Review of Scientific Instruments, 2007. 78(9): p. 093702.
41. J.E. Sader, I. Larson, P. Mulvaney and L.R. White, Method for the Calibration of Atomic Force Microscope Cantilevers. Review of Scientific Instruments, 1995. 66(7): p. 3789-3798.
42. C.P. Green, H. Lioe, J.P. Cleveland, R. Proksch, P. Mulvaney and J.E. Sader, Normal and Torsional Spring Constants of Atomic Force Microscope Cantilevers. Review of Scientific Instruments, 2004. 75(6): p. 1988-1996.
43. J.E. Sader, J.W.M. Chon and P. Mulvaney, Calibration of Rectangular Atomic Force Microscope Cantilevers. Review of Scientific Instruments, 1999. 70(10): p. 3967-3969.
73
44. J.E. Sader, Susceptibility of Atomic Force Microscope Cantilevers to Lateral Forces. Review of Scientific Instruments, 2003. 74(4): p. 2438-2443.
45. J.E. Sader, Frequency Response of Cantilever Beams Immersed in Viscous Fluids with Applications to the Atomic Force Microscope. Journal of Applied Physics, 1998. 84(1): p. 64-76.
46. C.P. Green and J.E. Sader, Torsional Frequency Response of Cantilever Beams Immersed in Viscous Fluids with Applications to the Atomic Force Microscope. Journal of Applied Physics, 2002. 92(10): p. 6262-6274.
47. L.A. Giannuzzi and F.A. Stevie, Introduction to Focused Ion Beams. 2005, New York: Springer.
48. S.G. Luengo, A. Galliano and C. Debief, Aqueous Lubrication in Cosmetic, in Aqueous Lubrication D.N. Spencer, 2014, L´Oréal: Zurich.
49. S. Hatami, A. Nafari, L. Nyborg and U. Jelvestam, Galling Related Surface Properties of Powder Metallurgical Tool Steels Alloyed with and without Nitrogen. Wear, 2010. 269(3-4): p. 229-240.
50. I. Heikkilä, E. Van der Heíde, E.D. Stam, H. Giraud, G. Lovato, N. Akdut, F. Clarysse and P. Caenen, Tool Material Aspects in Forming of Stainless Steel with Easy-to-Clean Lubricants, in Innovations in Metal Forming. 2004.
51. C. Valero Vidal and A. Igual Muñoz, Electrochemical Characterisation of Biomedical Alloys for Surgical Implants in Simulated Body Fluids. Corrosion Science, 2008. 50(7): p. 1954-1961.
52. E. Bettini, T. Eriksson, M. Boström, C. Leygraf and J. Pan, Influence of Metal Carbides on Dissolution Behavior of Biomedical CoCrMo alloy: SEM, TEM and AFM Studies. Electrochimica Acta, 2011. 56(25): p. 9413-9419.
53. J.A. Disegi, R.L. Kennedy and R. Pilliar, Cobalt-Base Alloys for Biomedical Applications. 1999, Virginia: ASTM STP1365.
54. J.J. Jacobs and L.T. Craig, Alternative Bearing Surfaces in Total Joint Replacement. 1998, Fredericksburg, VA: ASTM.
55. M. Sababi, S. Ejnermark, J. Andersson, P.M. Claesson and J. Pan, Microstructure Influence on Corrosion Behavior of a Fe–Cr–V–N Tool Alloy Studied by SEM/EDS, Scanning Kelvin Force Microscopy and Electrochemical Measurement. Corrosion Science, 2013. 66(0): p. 153-159.
56. Y. Yuan and R. Verma, Measuring Microelastic Properties of Stratum Corneum. Colloids and Surfaces B: Biointerfaces, 2006. 48(1): p. 6-12.
57. A. Potter, S.G. Luengo, R. Santoprete and B. Querleux, Stratum Corneum Biomechanics, in Skin Moisturization. 2009. p. 259-278.
74
58. W. Tang and B. Bhushan, Adhesion, Friction and Wear Characterization of Skin and Skin Cream Using Atomic Force Microscope. Colloids and Surfaces B: Biointerfaces, 2010. 76(1): p. 1-15.
59. L. Norlén, Stratum Corneum Keratin Structure, Function and Formation – a Comprehensive Review. International Journal of Cosmetic Science, 2006. 28(6): p. 397-425.
60. M. Palacio and B. Bhushan, A Review of Ionic Liquids for Green Molecular Lubrication in Nanotechnology. Tribology Letters, 2010. 40(2): p. 247-268.
61. T.L. Greaves and C.J. Drummond, Protic Ionic Liquids: Properties and Applications. Chemical Reviews, 2008. 108(1): p. 206-237.
62. O. Werzer, G.G. Warr and R. Atkin, Compact Poly(ethylene oxide) Structures Adsorbed at the Ethylammonium Nitrate-Silica Interface. Langmuir, 2011. 27(7): p. 3541-3549.
63. P. Niga, D. Wakeham, A. Nelson, G.G. Warr, M. Rutland and R. Atkin, Structure of the Ethylammonium Nitrate Surface: An X-ray Reflectivity and Vibrational Sum Frequency Spectroscopy Study. Langmuir, 2010. 26(11): p. 8282-8288.
64. K. Ueno, M. Kasuya, M. Watanabe, M. Mizukami and K. Kurihara, Resonance Shear Measurement of Nanoconfined Ionic Liquids. Physical Chemistry Chemical Physics, 2010. 12(16): p. 4066-4071.
65. Y. Mo, W. Zhao, M. Zhu and M. Bai, Nano/Microtribological Properties of Ultrathin Functionalized Imidazolium Wear-Resistant Ionic Liquid Films on Single Crystal Silicon. Tribology Letters, 2008. 32(3): p. 143-151.
66. W. Zhao, L. Wang, M. Bai and Q. Xue, Micro/Nanotribological Behaviors of Ionic Liquid Nanofilms with Different Functional Cations. Surface and Interface Analysis, 2010. 43(6): p. 945-953.
67. J. Pu, D. Huang, L. Wang and Q. Xue, Tribology Study of Dual-Layer Ultrathin Ionic Liquid Films with Bonded Phase: Influences of the Self-Assembled Underlayer. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 2010. 372(1-3): p. 155-164.
68. M. Palacio and B. Bhushan, Molecularly Thick Dicationic Ionic Liquid Films for Nanolubrication. Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films, 2009. 27(4): p. 986-995.
69. D. Wakeham, R. Hayes, G.G. Warr and R. Atkin, Influence of Temperature and Molecular Structure on Ionic Liquid Solvation Layers. The Journal of Physical Chemistry B, 2009. 113(17): p. 5961-5966.
70. R. Atkin and G.G. Warr, Structure in Confined Room-Temperature Ionic Liquids. The Journal of Physical Chemistry C, 2007. 111(13): p. 5162-5168.
75
71. J.A. Smith, O. Werzer, G.B. Webber, G.G. Warr and R. Atkin, Surprising Particle Stability and Rapid Sedimentation Rates in an Ionic Liquid. Journal of Physical Chemistry Letters, 2010. 1(1): p. 64-68.
72. R. Hayes, G.G. Warr and R. Atkin, At the Interface: Solvation and Designing Ionic Liquids. Physical Chemistry Chemical Physics, 2010. 12(8): p. 1709-1723.
73. S.S. Perry, Scanning Probe Microscopy Measurements of Friction. Mrs Bulletin, 2004. 29(7): p. 478-483.
74. R.J. Roark and W.C. Young, Formulas for Stress and Strain. McGraw-Hill. 1975, Nee York.
75. R.G. Cain, S. Biggs and N.W. Page, Force Calibration in Lateral Force Microscopy. Journal of Colloid and Interface Science, 2000. 227(1): p. 55-65.
76. D. Faoite, D. Browne, R.F. Chang-Díaz and K. Stanton, A Review of the Processing, Composition, and Temperature-Dependent Mechanical and Thermal Properties of Dielectric Technical Ceramics. Journal of Materials Science, 2012. 47(10): p. 4211-4235.
77. E.W. van der Vegte and G. Hadziioannou, Scanning Force Microscopy with Chemical Specificity: An Extensive Study of Chemically Specific Tip−Surface Interactions and the Chemical Imaging of Surface Functional Groups. Langmuir, 1997. 13(16): p. 4357-4368.
78. J.R. Cannara, M. Eglin and W.R. Carpick, Lateral Force Calibration in Atomic Force Microscopy: A New Lateral Force Calibration Method and General Guidelines for Optimization. Review of Scientific Instruments, 2006. 77(5): p. 053701.
79. K. Sasaki, Y. Koike, H. Azehara, H. Hokari and M. Fujihira, Lateral Force Microscope and Phase Imaging of Patterned Thiol Self-Assembled Monolayer Using Chemically Modified Tips. Applied Physics A: Materials Science & Processing, 1998. 66(0): p. 1275-1277.
80. R.L. McMullen and S.P. Kelty, Investigation of Human Hair Fibers Using Lateral Force Microscopy. Scanning, 2001. 23(5): p. 337-345.
81. J.R. Smith and J.A. Swift, Lamellar Subcomponents of the Cuticular Cell Membrane Complex of Mammalian Keratin Fibres Show Friction and Hardness Contrast by AFM. Journal of Microscopy, 2002. 206(3): p. 182-193.
82. F.Z. Sidouni, N. Nurdin, P. Chabrecek, D. Lohmann, J. Vogt, N. Xanthopoulos, H.J. Mathieu, P. Francois, P. Vaudaux and P. Descouts, Surface Properties of a Specifically Modified High-Grade Medical Polyurethane. Surface Science, 2001. 491(3): p. 355-369.
83. M.D. Levi, Y. Cohen, Y. Cohen, D. Aurbach, M. Lapkowski, E. Vieil and J. Serose, Atomic Force Microscopy Study of the Morphology of
76
Polythiophene Films Grafted onto the Surface of a Pt Microelectrode Array. Synthetic Metals, 2000. 109(1-3): p. 55-65.
84. L. Ping, Phase Analysis in Steel Using Analytical Transmission Electron Microscopy. 2004, Sandviken: Sandvik Materials Technology.
85. N. Nordgren, P. Eronen, M. Österberg, J. Laine and M.W. Rutland, Mediation of the Nanotribological Properties of Cellulose by Chitosan Adsorption. Biomacromolecules, 2009. 10(3): p. 645-650.
86. P. Beer, In Situ Examinations of the Friction Properties of Chromium Coated Tools in Contact with Wet Wood. Tribology Letters, 2005. 18(3): p. 373-376.
87. P. Marcus, Corrosion Mechanisms in Theory and Practice. 3rd ed. Corrosion technology. 2012, Boca Raton, Fla.: CRC. xii, 929 p.
88. M. Nonnenmacher, M.P. O’Boyle and H.K. Wickramasinghe, Kelvin Probe Force Microscopy. Applied Physics Letters, 1991. 58(25): p. 2921-2923.
89. N. Sathirachinda, R. Pettersson, S. Wessman and J. Pan, Study of Nobility of Chromium Nitrides in Isothermally Aged Duplex Stainless Steels by Using SKPFM and SEM/EDS. Corrosion Science, 2010. 52(1): p. 179-186.
90. M. Li, L.Q. Guo, L.J. Qiao and Y. Bai, The Mechanism of Hydrogen-Induced Pitting Corrosion in Duplex Stainless Steel Studied by SKPFM. Corrosion Science, 2012. 60(0): p. 76-81.
91. V. Guillaumin, P. Schmutz and G.S. Frankel, Characterization of Corrosion Interfaces by the Scanning Kelvin Probe Force Microscopy Journal of the Electrochemical Society, 2001. 148(5): p. B163-B173.
92. I. Heikkilä, The Possitive Effect of Nitrogen Alloying of Tool Steels Used In Sheat Forming, in Department of Engineering Sciences. 2013, Uppsala Universitet: Uppsala.
93. L. Vitos, K. Larsson, B. Johansson, M. Hanson and S. Hogmark, An Atomistic Approach to the Initiation Mechanism of Galling. Computational Materials Science, 2006. 37(3): p. 193-197.
94. A. Glaser, S. Surnev, F.P. Netzer, N. Fateh, G.A. Fontalvo and C. Mitterer, Oxidation of Vanadium Nitride and Titanium Nitride Coatings. Surface Science, 2007. 601(4): p. 1153-1159.
95. S. Vijendran, H. Sykulska and W.T. Pike, AFM Investigation of Martian Soil Simulants on Micromachined Si Substrates. Journal of Microscopy, 2007. 227(3): p. 236-245.
96. L.T. Zhuravlev, The Surface Chemistry of Amorphous Silica. Zhuravlev Model. Colloids and Surfaces A: Physicochemical and Engineering Aspects, 2000. 173(1–3): p. 1-38.
97. R.K. Iler, The Chemistry of Silica : Solubility, Polymerization, Colloid and Surface Properties, and Biochemistry. 1979, New York: Wiley.
77
98. G. Okamoto, Passive Film of 18-8 Stainless Steel Structure and Its Function. Corrosion Science, 1973. 13(6): p. 471-489.
99. J. Sedriks, A., Corrosion of Stainless Steel. 1996, Virginia: John Wiley & Sons.
100. K. Holmberg, D.O. Shah and M.J. Schwuger, Handbook of Applied Surface and Colloid Chemistry. 2002, Chichester: Wiley.
101. J.N. Israelachvili and R.M. Pashley, Molecular Layering of Water at Surfaces and Origin of Repulsive Hydration Forces. Nature, 1983. 306(5940): p. 249-250.
102. C.R. Clayton, G.P. Halada and J.R. Kearns, Passivity of High-Nitrogen Stainless Alloys: the Role of Metal Oxyanions and Salt Films. Materials Science and Engineering: A, 1995. 198(1–2): p. 135-144.
103. S. Yuan, B. Liang, Y. Zhao and S.O. Pehkonen, Surface Chemistry and Corrosion Behaviour of 304 Stainless Steel in Simulated Seawater Containing Inorganic Sulphide and Sulphate-Reducing Bacteria. Corrosion Science, 2013. 74(0): p. 353-366.
104. C. García, F. Martín, Y. Blanco and M.L. Aparicio, Effect of Ageing Heat Treatments on the Microstructure and Intergranular Corrosion of Powder Metallurgy Duplex Stainless Steels. Corrosion Science, 2010. 52(11): p. 3725-3737.
105. S.S.M. Tavares, F.J. da Silva, C. Scandian, G.F. da Silva and H.F.G. de Abreu, Microstructure and Intergranular Corrosion Resistance of UNS S17400 (17-4PH) Stainless Steel. Corrosion Science, 2010. 52(11): p. 3835-3839.
106. S.X. Li, Y.N. He, S.R. Yu and P.Y. Zhang, Evaluation of the Effect of Grain Size on Chromium Carbide Precipitation and Intergranular Corrosion of 316L Stainless Steel. Corrosion Science, 2013. 66: p. 211-216.
107. Y.-T. Cheng and C.-M. Cheng, Scaling, Dimensional Analysis, and Indentation Measurements. Materials Science and Engineering: R: Reports, 2004. 44(4–5): p. 91-149.
108. A.C. Fischer-Cripps, Nanoindentation. 2nd ed. Mechanical engineering series. 2004, New York ; London: Springer. xxii, 263 p.
109. W.C. Oliver and G.M. Pharr, Measurement of Hardness and Elastic Modulus by Instrumented Indentation: Advances in Understanding and Refinements to Methodology. Journal of Materials Research, 2004. 19(01): p. 3-20.
110. J.R. Withers and D.E. Aston, Nanomechanical Measurements with AFM in the Elastic LImit. Advances in Colloid and Interface Science, 2006. 120(1–3): p. 57-67.
78
111. B.V. Derjaguin, V.M. Muller and Y.P. Toporov, Effect of Contact Deformations on the Adhesion of Particles. Journal of Colloid and Interface Science, 1975. 53(2): p. 314-326.
112. K.L. Johnson, Contact Mechanics. 1985, Cambridge: Cambridge University Press.
113. F. Xu and T. Lu, Introduction of Skin Biothermomechanics and Thermal Pain. 2011: Science Press Beijing and Springer-Verlag Berlin Heidelberg.
114. P.G. Agache, C. Monneur, J.L. Leveque and J. Rigal, Mechanical Properties and Young's Modulus of Human Skin in Vivo. Archives of Dermatological Research, 1980. 269(3): p. 221-232.
115. T. Jee and K. Komvopoulos, Skin Viscoelasticity Studied in Vitro by Microprobe-Based Techniques. Journal of Biomechanics, 2014. 47(2): p. 553-559.
