Study and characterization of diamond surface for biosensoring applications

Introduction
Diamond is the only wide band gap representative of the elemental semi-
conductors, with a crystal structure identical to its more common relatives
silicon and germanium.
On first glance one might also expect similar surface properties in terms
of reconstructions, surface states, and surface band diagrams. In part, this
expectationisindeedfulfilled,butdiamondalsoexhibitsanumberofunusual
and potentially very useful surface properties [1] [2].
Particularly when the surface dangling bonds are saturated by monova-
lent hydrogen atoms, (donor-like) surface states are removed from the gap,
the electron affinity changes sign and becomes negative, and the material
becomes susceptible to an unusual type of transfer doping where holes are
injected by acceptors located at the surface instead of inside the host lattice.
In such a way diamond surface becomes a conductor material and can be
used as an electrode to collect electrical signals [3] [4] [5].
Sincediamondelectricalandchemicalpropertiesarestrictlyrelatedtoits
surface termination, a surface nondestructive analysis technique is essential
to check its status.
Among all the possible surface techniques, XPS (X-ray Photoelectron Spec-
troscopy) and UPS (UV-ray Photoelectron Spectroscopy) are the most suit-
able, since their scanning depth is nearly 3 nm and 1.5 nm respectively [6]
[7] [8].
Moreover they give information not only about chemical species present onto
diamondsurface,buttheycanalsostatehowtheseatomsarelinkedtogether,
thanks to their chemical shift [9] [10].
This particular behavior has made diamond a good candidate for the
implementation of biosensors [11] [6] [7].
An ideal biosensor should combine nature sensitivity and specificity with
the advantages of modern microelectronics: diamond is especially attractive
because, in addition to having good electrical and chemical properties, it is
widely considered to be biocompatible and chemically inert and can be de-
positedasarobust, thinfilmonsiliconandothermicroelectronic-compatible
5
TABLE OF CONTENTS
substrates [12] [13] [14] [15].
Moreover, if we use a diamond as a substrate for a biosensor, we can
also exploit its bulk properties such as extreme mechanical hardness and,
the most important, broad optical transparency, from the deep ultraviolet
to the far infrared. This makes possible simultaneous recording of electrical
and optical signals from living cells, as sensor elements.
Hydrogen terminated diamond surfaces have been used to attach DNA
fragments[11],showingthatdiamondisagoodplatformontowhichbiomolecules,
with specific funtional groups, can be linked.
The first evidence of selective attachment of mammalian neurons and or-
deredoutgrowthofneuritesonpatterneddiamondsurfaceshasbeenrecently
reported by Specht et al. [16]. However, to our best knowledge, no report
about the functional viability of neurons on functionalised diamond surfaces
has been published to date.
To explore the possibility of exploiting the above mentioned properties
of diamond for the realisation of biosensors it is essential to investigate ad-
hesion and neuronal excitability (i.e., the ability of neurons to generate and
propagate trains of electrical impulses) on hydrogenated (HTD) or oxidised
(OTD) diamond surfaces.
Thusthestudyofdiamondsurfaceproperties,withdifferentterminations,
is fundamental if we are interested in biosensoring applications.
My PhD research activity thus will be focused onto three main topics:
• Implementation and characterization of an XPS apparatus, in order to
characterize diamond surface properties and to check out its modifica-
tion due to hydrogenation and/or oxidation processes [17] [18].
• Study different methods to perform hydrogenation (plasma, molecular
orHFCVD[19][20])andoxidation(chemical, plasmaorUVtechnique
[21] [22]), in order to find out the best way to modify diamond surface,
without introducing any sort of contamination and damage.
• Study of neurons adhesion and viability onto diamond, to check their
bio-compatibility [7] [15] [23].
6
Chapter 1
Diamond:“the Biggest and the
Best”
1.1 General properties
Diamond has been prized for centuries as a gemstone of exceptional bril-
liance and lustre. But to a scientist diamond is interesting for its range of
exceptional and extreme properties.
Indeed, a glimpse at any compendium of material data properties will
prove that diamond is almost always ‘the biggest and best’ [24]. Among
other properties, diamond is the hardest known material, has the highest
thermal conductivity at room temperature, is transparent over a very wide
wavelength range, is the stiffest material, the least compressible, and is inert
to most chemical reagents.
A selection of some of these properties is given in tab.1.1. With such a
wide range of exceptional properties, it is not surprising that diamond has
sometimes been referred to as ‘the ultimate engineering material’.
But what is diamond exactly?
Diamond is composed of the single element carbon, and it is the arrange-
mentoftheCatomsinthelatticethatgivesdiamonditsamazingproperties.
Let us compare the structure of diamond and graphite, the other major al-
lotrope of carbon, both composed of just carbon. In diamond we have the
hardest known material, in graphite we have one of the softest, simply by re-
arranging the way the atoms are bonded together. The relationship between
diamond and graphite is a thermodynamic and kinetic one, as can be seen
in the phase diagram for carbon (fig.1.1).
At normal temperatures and pressures, graphite is only a few eV more stable
thandiamond, andthefactthatdiamondexistsatallisduetotheverylarge
7
1 – Diamond:“the Biggest and the Best”
Property Diamond Best Alternative
Mechanical Hardness
(g/mm
2
) 5700 - 10400 4500 (cubic BN)
Thermal Conductivity
(Wcm
−1
K
−1
) 20 6 (BeO)
Electrical Resistivity
(Ω cm) 10
15
10
15
(Al
2
O
3
)
Band Gap
(eV) 5.45 1.12 (Si)
Lattice Constant
(
˚
A) 3.56 5.43 (Si)
Optical Transmission 220 nm≤ λ≤ 2500 nm Sapphire
λ≥ 6000 nm 150 nm≤ λ≥ 5000 nm
Table 1.1. Diamond bulk properties [2]
Figure 1.1. Carbon phase diagram [25]
activation barrier for conversion between the two.
There is no easy mechanism to convert between the two and so inter-
conversion requires almost as much energy as destroying the entire lattice
and rebuilding it. Once diamond is formed, therefore, it cannot reconvert
8
1.1– General properties
back to graphite because the barrier is too high. So diamond is said to be
metastable, since it is kinetically stable, not thermodynamically stable.
Diamond is created deep underground under conditions of extreme pressure
andtemperature. Undertheseconditionsdiamondisactuallythemorestable
of the two forms of carbon, and so over a period of millions of years carbona-
ceous deposits slowly crystallise into single crystal diamond gemstones.
Diamonds typically crystallise in the face-centered cubic crystal system
(FCC) and consist of tetrahedrally bonded carbon atoms. The unit cell of
diamond, with a side length a
0
approximately equal to 3.567
˚
A at room
temperature, has a two atom basis at (0,0,0) and (1/4, 1/4, 1/4), which
means half of the atoms are at lattice points and the other half are offset by
(1/4, 1/4, 1/4), where 1 is the length of a side of the unit cell (fig.1.2).
The C – C bond length d is equal to 1.54
˚
A. The atomic density is 1.76×
10
23
atoms/cm
3
, while the diamond’s density is 3.52 g cm
−3
. Its covalent
bonds between hybrid sp
3
orbitals make it the hardest material in nature
(from the Greek “adamas” = indestructible).
The carbon atoms in graphite, which are bonded together with hybrid
sp
2
orbitals, display a different (nontetrahedral) connectivity and as a result
shows dramatically different physical characteristics: graphite is a soft, dark
gray, opaque mineral (fig.1.3).
Anyway, even if diamond possesses so many extraordinary physical prop-
erties, yet its practical use in science or engineering has been limited due
its scarcity and expense and also for the fact that natural diamond is only
available in the form of stones or grit, and can sometimes contain impurities,
which can lower its performances.
Natural diamonds are classified by the type and level of impurities found
within them:
• Type I
Type I diamonds make up about 99% of all diamonds. They contain
an abundance of nitrogen atoms and are electrical non-conductors.
They are further broken down as follows:
1. Type Ia
By far the most common in the Type I category, these diamonds
contain clusters of nitrogen atoms as impurities in the crystal lat-
tice. When N atoms cluster into pairs, they don’t cause color
because they absorb ultraviolet light beyond the visible blue end
of the spectrum. These diamonds show a dark line at the 415 nm
line in the spectroscope. This category comprises colorless and
some yellowish diamonds.
9
1 – Diamond:“the Biggest and the Best”
2. Type Ib
These are much rarer than Type Ia diamonds (less than 1% of all
Type I). They contain nitrogen atoms that are dispersed through-
out the crystal lattice. When N atoms are dispersed, they can
absorb light in the blue end of the spectrum, allowing yellowish
to fancy yellow stones with stronger color than Type Ia. Type Ib
and mixes of Type Ia/Ib can be subjected to high pressure and
high temperatures (HPHT) to lose color.
• Type II
A small number of diamonds are Type II.
1. Type IIa
These are rare diamonds with exceptionally pure chemical com-
position and contain no nitrogen or boron. They are often very
large and usually colorless, though they may be pink, brown or
blue-green. They are inert to shortwave ultraviolet radiation and
don’t conduct electricity but are efficient conductors of heat.
2. Type IIb
These are even rarer than Type IIa diamonds. Boron substitutes
for some carbon atoms in Type IIb diamonds. They are elec-
trical semiconductors and are extremely sensitive to temperature
changes. They phosphoresce to shortwave ultraviolet. Most blue
diamonds are Type IIb.
Figure 1.2. Diamond structure [26] Figure 1.3. Graphite structure [27]
10
1.1– General properties
To overcome natural diamond flaws, researchers realised that in order to
form diamond in an easy and not too much expensive way, conditions are
needed where diamond is the more stable phase.
The knowledge of the conditions under which natural diamond is formed
deep underground suggested that diamond could be formed by heating car-
bon under extreme pressure. This process forms the basis of the so-called
High-PressureHigh-Temperature (HPHT)growthtechnique,firstmarketed
by General Electric, and which has been used to produce industrial diamond
for several decades.
In this process, graphite is compressed in a hydraulic press to tens of thou-
sands of atmospheres, heated to over 2000 K in the presence of a suitable
metal catalyst, and left until crystallises and forms diamond [28].
The diamond crystals so produced are used for a wide range of industrial
processes, whichusethehardnessandwearresistancepropertiesofdiamond,
suchascuttingandmachiningmechanicalcomponents,andforpolishingand
grinding of optics.
However,thedrawbackoftheHPHTmethodisthatitstillproducesdiamond
in the form of single crystals ranging in size from nanometers to millimeters,
and this limits the range of applications for which it can be used.
What is required is a method to produce diamond in a form that can allow
many more of its superlative properties to be exploited, in other words, as a
diamond thin film.
1.1.1 Chemical vapor deposition
Chemical vapor deposition, as its name implies, involves a gas-phase chem-
ical reaction occurring above a solid surface, which causes deposition onto
that surface.
All CVD techniques for producing diamond films require a means of activat-
inggas-phasecarbon-containingprecursormolecules. Thisgenerallyinvolves
thermal (e.g. hot filament) or plasma (D.C., R.F., or microwave) activation,
or use of a combustion flame (oxyacetylene or plasma torches).
Figure 1.4 illustrates two of the more popular experimental methods and
gives some indication of typical operating conditions. While each method
differs in detail, they all share features in common.
For example, growth of diamond normally requires that the substrate be
maintained at a temperature in the range 1000-1400 K, and that the pre-
cursor gas be diluted in an excess of hydrogen (typical CH
4
mixing ratio
1-2 vol%). The resulting films are usually polycrystalline, with a morphol-
ogy that is sensitive to the precise growth conditions. Growth rates for the
various deposition processes vary considerably, and it is usually found that
11
1 – Diamond:“the Biggest and the Best”
Figure 1.4. Common types of CVD reactor. (a) Hot filament reactor,
(b) Microwave Plasma Enhanced Reactor [29]
higher growth rates can be achieved only at the expense of a corresponding
loss of film quality. Quality here is a subjective concept. It is taken to imply
some measure of factors such as the ratio of sp
3
(diamond) to sp
2
-bonded
(graphite) carbon in the sample, the composition (e.g. C-C versus C-H bond
content) and the crystallinity. In general, combustion methods deposit dia-
mond at high rates (typically 100-1000 μm/h, respectively), but often only
with poor process control, leading to poor quality films. In contrast, the hot
filament and plasma methods have much slower growth rates (0.2-15μm/h),
butproducehighqualityfilms. Oneofthegreatchallengesfacingresearchers
in CVD diamond technology is to increase the growth rates to economically
possible rates, (hundreds of μm/h, or even mm/h) without compromising
film quality.
Progress is being made using microwave deposition reactors, since the de-
position rate has been found to scale approximately linearly with applied
microwave power.
Currently, the typical power rating range for a microwave reactor is
(1.5 - 100) kW, with typical growth rates ranging between 0.5 to 15 μm/h,
passing from 1” to 12” wafers [30] [31]. In one case it has been possible to
grow homoepitaxial diamond by microvawe plasma CVD on small areas (≤
1 cm
2
) with growth rates up to 150 μm/h [32].
Hereafter a comparative table of different CVD deposition techniques,
with typical growth parameters (gas activation temperature, pressure in
chamber, gas flow rate, diamond growth rate and surface area), is reported:
12
1.1– General properties
Activation T
act
P Flow rate Growth rate Area
tecnique [
◦
C] [mbar] [sccm] [μm/h] [cm
2
]
Hot
filament ≤ 2800 10 - 100 ≤ 2000 0.2 - 1.5 ≤ 50000
Flame
∼
= 3100 1000 ≤ 2000 50 - 200
∼
= 1
Microvawe
Plasma 3000 - 4000 20 - 250 ≤ 2000 ≤ 15 ≤ 1000
DC-
plasma 5000
∼
= 200 20 - 1000 ≤ 250 ≤ 400
Table 1.2. Comparison of different deposition techniques parameters [33]
Thermodynamically, graphite, not diamond, is the stable form of solid
carbon at the T and P occuring during the CVD process. The fact that
diamond films can be formed by CVD techniques is inextricably linked to
the presence of hydrogen atoms, which are generated as a result of the gas
being ”activated”, either thermally or via electron bombardment. These H
atoms are believed to play a number of crucial roles in the CVD process
(fig.1.5):
• they undergo H abstraction reactions with stable gas-phase hydro-
carbon molecules, producing highly reactive carbon-containing radi-
cal species. This is important, since stable hydrocarbon molecules do
not react to cause diamond growth. The reactive radicals, especially
methyl, CH
3
, can diffuse to the substrate surface and react, forming
the C-C bond necessary to propagate the diamond lattice;
• H-atomsterminatethedangling carbonbondsonthegrowingdiamond
surface and prevent them from cross-linking, thereby reconstructing to
a graphite-like surface;
• atomic hydrogen etches both diamond and graphite but, under typi-
cal CVD conditions, the rate of diamond growth exceeds its etch rate
while for other forms of carbon (graphite, for example) the converse is
true. This is believed to be the basis for the preferential deposition of
diamond rather than graphite.
The surface morphology obtained during CVD depends critically upon
the gas mixing ratio and the substrate temperature. Under ’slow’ growth
conditions - low CH
4
partial pressure, low substrate temperature - it is pos-
sibletoobtainamicrocrystallinefilm,withtriangular(111)facetsbeingmost
13
1 – Diamond:“the Biggest and the Best”
Figure 1.5. Chemical processes involved in diamond CVD deposition [29]
evident, along with many obvious twin boundaries.
(100) facets, appearing both as square and rectangular forms, begin to domi-
nateastherelativeconcentrationofCH
4
intheprecursorgasmixture,and/or
the substrate temperature, is increased. At still higher CH
4
partial pressures
the crystalline morphology disappears altogether; a film, such as that shown
inFigureC.11, isanaggregateofdiamondnanocrystals, withanaveragesize
of 500 nm (measurement made with a Scanning Electron Microscope - SEM,
provided by the Department of Electron Devices and Circuits - University of
Ulm, Germany).
1.2 Diamond Surface
Until now we have talked primarily about diamond bulk properties, but for
our purpose, we are much more interested in diamond surface properties. In
the next subsections I will show how it’s possible to modify diamond surface
tocreateconductiveorinsulatinglayerbyterminatingtheC-danglingbonds
on top of the surface with H - atoms or O - atoms.
An explanation of the crystallographic notation, that will be used here-
after to describe the adsorbed species on single crystal surfaces, known as
14
1.2– Diamond Surface
Figure 1.6. SEM image of a nanocrystalline CVD diamond (grown on Si)
‘Wood’s notation’ [34], will be related in the appendix A.
1.2.1 Surface Properties
The (001)-orientated surface of diamond is the most technologically impor-
tant, followed by the (111) surface. This is because, during manufacture
of synthetic diamond by chemical-vapor deposition, these are the slowest-
growing surfaces and are therefore those that remain behind after the other
surfaces have grown themselves out of existence [35].
Clean surface
On the clean
1
(001) surface, neighboring C atoms come together to form
double-bondeddimers, introducingoccupiedπ andunoccupiedπ
∗
statesinto
the band-gap of the electronic structure. Perpendicular to the bonds, the
dimers are separated from one another by 1/
√
2 of the conventional lattice
parameter (i.e., by about 2.5
˚
A), and this (001)-(2×1) surface is nonmetallic.
Instead the clean, single-dangling-bond (SDB) surface with (111) orienta-
tion is well known to exhibit a dramatic reconstruction after which it ap-
pears more like that of a clean (110) surface. The uppermost C atoms form
zigzag chains that run in parallel across the surface. As these atoms are only
threefold coordinated, they share a delocalised π network extended along
1
In this case clean means with no adsorbates on top of the surface
15
1 – Diamond:“the Biggest and the Best”
the chain, causing this (111)-(2×1) surface to exhibit semi-metallic behavior
(fig.1.7).
Hydrogenated surface
The mono-hydrogenated, (001)-(2×1):H surface is widely accepted to be the
most stable under normal conditions [36] [37] [38]. The addition of hydro-
gen removes the π and π
∗
states from the electronic structure, as the sur-
face dimers change from being double to single bonded. For the (111) ori-
entation, mono-hydrogenation of both the SDB (bulk-terminated) and the
reconstructed surfaces was investigated, these becoming, respectively, the
(111)-(1×1):H and (111)-(2×1):H surfaces (fig.1.7).
Figure 1.7. Atomic geometries for the clean and hydrogenated
(001)-(2×1) diamond surfaces [35]
Oxygenated surface
For the (001)-(1×1):O surface, the two most-plausible configurations are the
following: (a) the “ketone” arrangement, in which the O atom is double-
bonded to a single surface C atom (C=O), with the axis of this carbonyl
groupnormaltothesurfaceand(b)the“ether”,inwhichtheOatombridges
two surface C atoms and makes a single bond to each (C-O-C). The last one
16
1.2– Diamond Surface
Figure 1.8. Atomic geometries for the oxygenated (001)-(1×1) diamond
surfaces; (a) ketone and (b) ether configurations [35]
is a more stable configuration, due to the fact that the highest occupied level
in the ether system is significantly lower in energy than the same level in the
ketone system (fig.1.8).
1.2.2 Surface Conductivity
Undoped diamond, with a band gap of 5.5 eV, is mostly considered an elec-
trical insulator or sometimes a semiconductor with a wide band.
Yet in the 1989 Landstrass and Ravi [39] reported in their work a substantial
conduction on the surface of hydrogenated diamond (on both single crys-
tals and thin films samples). During the last years many experiments have
shown the same behavior on H-terminated diamonds and have led to a sort
of explanation of this phenomenon.
The surface conductivity (SC)of hydrogenated diamond is of the order of
10
−4
to 10
−5
Ω
−1
/per square at room temperature (RT). The areal density
of the p-type carriers responsible for the conductivity is aboutρ
s
=10
13
cm
−2
and it is hardly temperature dependent between RT and 150 K.
The Hall mobility of the carriers varies also little with temperature (≈T
1.2
)
and is of the order of 30cm
2
V
−1
s
−1
with a maximum value of 70cm
2
V
−1
s
−1
reported for a carrier density of 1.2×10
12
cm
−2
. These mobilities are not too
differentfromthosemeasuredforB-dopeddiamond,andthereisthusgeneral
agreementthatthecarriersareholes residinginanaccumulationlayer atthe
surface.
17
1 – Diamond:“the Biggest and the Best”
The depth distribution of the acceptors responsible for the hole accumu-
lation is discussed controversially, ranging from species at the surface over
layers extending up to 10 nm into the diamond. It has even been suggested
that the acceptors form a layer buried 30 nm below the surface [40]. In the
limit of a quasi-two-dimensional acceptor layer at or up to 30 nm below the
surface, the observed areal density of 10
13
cm
−2
holes requires a band profile
such that the surface Fermi level position lies within a few kT at the valence
band maximum (VBM) [41].
Because the surface conductivity is observed only on hydrogenated di-
amond surfaces and disappears after dehydrogenation or oxidation of the
surface, it has been assumed that hydrogen is directly responsible for the
hole accumulation layer by forming particular defects that act as acceptors.
DespitetheexceptionalnatureofSCondiamond, thereisnoclearunder-
standingofthenatureofthedopingmechanismthatleadstothesurface-near
holelayer. Suchanunderstandingis, however, highlydesirableinordertobe
abletoexploitfullytheconsiderablepotentialofSCforapplications[42],[43].
Electrochemical Model
In the article ’Origin of Surface Conductivity in Diamond’ Maier et al. [41]
giveexperimentalevidence thathydrogenisonlyanecessaryrequirementfor
SC; exposure to air is also essential.
In order to explain this, they performed an experiment, using different
types of diamonds with different crystallographic orientations.
The first step was the measurement of the samples surface conductance,
whichwasestablishedbythefingerprintofNegativeElectronAffinity(NEA)
2
of the surface as seen in the total photoelectron yield spectrum.
After annealing the sample in UHV (Ultra High Vacuum) at ≈400
◦
C for
15 min, the conductance dropped to 10
−10
A/V while the hydrogenation
remained intact as demonstrated by the NEA property of the surface.
In fact, thermal desorption of chemisorbed hydrogen does not begin below
700
◦
C. Then they masked half part of each samples and removed hydrogen
from the other half by electron beam (1 keV, 0.2 mA/cm
2
, 90 min) induced
desorption.
The masked area still shows the fingerprint of NEA, while this feature is
absent on the irradiated surface, proving that the hydrogen termination has
beenremovedandtheelectronaffinityturnedpositive. Asalsodemonstrated
2
The electron affinity χ is the energy required to remove an electron from the Conduc-
tion Band Minimum (CBM) and take it to the ’vacuum level’. When the vacuum level
is below the CBM, χ becomes negative and the surface layer acts as an ’electrons well’,
where electrons can escape from with an energy E=|χ|.
18
1.2– Diamond Surface
Figure 1.9. Surface conductance of the hydrogenated (masked) and the
de-hydrogenated (irradiated) part of a diamond (100) [41]
in fig.1.9, both halves of the sample are in the low conductance state and
remain so as long as they are kept in UHV. However, when they are brought
up to air, the conductance of the masked and thus hydrogenated area rises
by 4 orders of magnitude within the first twenty minutes of exposure and
increases more slowly thereafter until it reaches 10
−5
A/V after three days.
By contrast, the dehydrogenated part of the sample remains in its low con-
ductance state with no sign of change whatsoever. This experiment clearly
demonstrates that the hydrogenation of diamond is a necessary but not a
sufficient condition for high surface conductivity. An additional ingredient,
that is obviously coming from the air and that thermally desorbs in UHV
above 400
◦
C, is necessary.
In order to act as an acceptor, the adsorbate must have its lowest unoc-
cupied electronic level below the VBM of diamond. With an electron affinity
χ
C:H
= -1.3 eV for hydrogenated diamond (fig.1.10), this requirement sets a
lower limit for the electron affinity χ
ad
of the adsorbate: χ
ad
= E
g
- 1.3 eV
= 4.2 eV, where E
g
= 5.5 eV is the band gap energy of diamond. Electron
affinities of molecular atmospheric species lie below 2.5 eV and even for halo-
gen atoms χ
ad
does not exceed 3.7 eV. Thus, direct electron transfer from
the diamond into an atmospheric adsorbate appears to be impossible.
However, a thin water layer, as it forms naturally on all surfaces exposed
to atmosphere, provides an electron system which can act as a surface ac-
ceptor for diamond. Electrons exchange from diamond to the water layer is
governed by the redox reaction:
19