Studies of self assembled monolayers on highly oriented pyrolytic graphite using scanning tunneling microscopy and computational simulation 7

FORMATION OF 2D CRYSTALS ON HOPG
CHAPTER 7
FORMATION OF 2D CRYSTALS ON HOPG

7.1 Motivation
In materials science, a crystal is built up from regularly repeating ‘structural
motifs’, which may be atoms, molecules, or groups of atoms, molecules, or ions [1].
Traditionally, it is widely accepted that the repeating pattern of the crystal reproduces
itself into three spatial dimensions (X, Y, Z), which leads to a three dimensional
crystal.
If the regularly repeating ‘structural motifs’ are restricted within a flat plane, they
can only grow in two dimensions (X, Y). The resulting structures can be considered as
crystals in two dimensions, and we call them two dimensional crystals (2D crystals).
Although not as popular as 3D crystals, 2D crystals are not very rare in nature with
surfaces as their favorite media [2]. The studies of physical and chemical properties of
two dimensional objects have become more and more important in daily life and
industries. Thus, the two dimensional systems are gaining great interest from many
sectors of industries, including microelectronics, catalysis, etc.
The investigations of the two-dimensional crystals are traditionally part of the
surface studies. Two-dimensional phases and phase transitions between them were
first observed by Langmuir in the 1920s in the experiments with layers of organic
molecules (salts of fatty acids) on the surface of a liquid. It results in the indirect
discovery of a crystalline phase in a two-dimensional system [2]. For existence of

109
FORMATION OF 2D CRYSTALS ON HOPG
most 2D crystals in three-dimensional world, an external field is needed. Most 2D
crystals are formed on the surface of a liquid or a crystal. One representative example
is the organic monolayers on the HOPG, which will be discussed further in this
chapter.
As a unique tool in the study of the physisorbed organic monolayers, the use of

scanning tunneling microscopy and the scanning results had been discussed in detail
in reviews by De Feyter [3, 4], Wan [5] and Tao [6]. But Matzger and his coworkers’
attempt is the first systematic work to link the organic monolayers on the HOPG with
the 2D crystals [7]. Since very little systematic work has been done towards the two
dimensional crystals till now, we hope our research can make some useful
contribution to this area.
As defined [2], there are two classes of 2D structures based on the dependence of
the adatoms’ period to the period of the substrate. One is the commensurate structure
with periods that are multiples of the substrate period, and anther is the
incommensurate structure with at least one period incommensurate with the
substrate’s period.
In our research the HOPG was chosen as the substrate to support 2D crystals. Its
advantages include natural flatness; minimum interaction between the adsorbates and
substrate; excellent conductivity, which allows us to study the monolayers using STM.
Unlike some metal and semiconductor surfaces, e.g. gold and silicon, the sp2 carbons
of the graphite crystal are chemically inert. They are unlikely to form covalent bonds
with adsorbates. Therefore the organic molecules on the graphite surface only

110
FORMATION OF 2D CRYSTALS ON HOPG
experience the weak molecule-substrate interactions, e.g. van der Waals forces.
Some examples of two dimensional crystals are: freestanding smectic liquid
crystal films; lattices of adatoms adsorbed on metal surfaces, for example: La-W(112)
system; overlayers on exfoliated graphite, e.g. noble gases on the basal plane of
graphite; and intercalated graphite, e.g. bromine atoms implanted into the graphite [2].
The number of 2D crystals is significantly increased as the new family member SAMs
on HOPG is counted in, since there are thousands of SAMs. The development in the
STM technology enhances the possibility of further investigation for 2D crystals
system.

7.2 Classification of crystals
The most common method is to classify 3D crystals according to their crystalline
structures. There are seven crystal lattice systems [8]: cubic (isometric), tetragonal,
orthorhombic, hexagonal, trigonal, triclinic and monoclinic. Another way is to
categorize the 3D crystals according to their chemical and physical properties. There
are four types of them: covalent crystals, metallic crystals, ionic crystals, and
molecular crystals.
For the case of crystals in 2D, among 17 space groups (p1, pg, p2gg, p4, p3, p6,
p2, pm, p2mg, p4gm, p31m, p3m1, c2mm, cm, p2mm, p4mm, and p6mm) [7], the
examined 359 SAMs belong to 9 groups ( p1, pg, p2gg, p3, p6, p2, p2mg, c2mm, cm).
However, SAMs in eight highly symmetric groups (p4, p4gm, p4mm, p3ml, p3lm,
p6mm, p2mm, and pm) have not been observed. The most possible reason is that the

111
FORMATION OF 2D CRYSTALS ON HOPG
number of the sample is insufficient to make the complete statistical studies.
Furthermore, HOPG surface lattice can affect the configuration of the adsorbates,
especially for the molecules with long alkyl chain groups, through van der Waal’s
forces (Chapter 4). Therefore the low symmetric space groups are always more
popular.
With regards to the classification of 2D crystals according to their chemical and
physical properties, only the molecular crystals have been observed so far. Suzuki et
al have inserted the bipyridinium cations between the carboxylic acid groups [9]. But
the crystal is not held together by the ionic bonds (electrostatic forces), it is still
considered as a molecular crystal, which is held together by non-covalent interactions,
like van der Waals forces or hydrogen bonding.

7.3 Proposed process of formation of the 2D crystals
The theory of growth of perfect crystals has been well developed. For the growth
of a three dimensional crystal, it mainly occurs at steps of monomolecular height on

its surfaces, and the probability for the formation of these two-dimensional nuclei is a
very sensitive function of the supersaturation [10].
The principle of the crystal growth was employed to study the growth of SAMs
on HOPG. Although the graphite surface has a high degree of perfection, they still
have steps and defects, as shown in Fig 7.1.

112
FORMATION OF 2D CRYSTALS ON HOPG

Fig 7.1 STM image of HOPG shows steps-like structures. (Height profile, 172nm×172nm)

Two possible mechanisms for the growth of two dimensional crystals on the
HOPG surface were proposed. In the first mechanism, like most crystal growing
environment, the steps with kinks are the best adsorption sites for the adsorbates. The
adsorbates start to occupy these adsorption sites when the supersaturated solution was
deposited onto the HOPG surface. It is followed by the growth of two dimensional
islands with a dense structure around the first few adsorbed molecules with the
increase in surface coverage, until the substrate is covered by a continuous monolayer.
In the second mechanism, the molecules randomly physisorb on the graphite surface
when the supersaturated solution was deposited onto the HOPG surface. With

113
FORMATION OF 2D CRYSTALS ON HOPG
increasing number of adsorbates on the surface, the randomly distributed molecules
start to pack closely to form a monolayer. The first proposed mechanism is more
likely to occur during the monolayer formation process, as discussed in Chapter 4
based on our experimental results. In addition, the fact that only partial surface being
covered with the SAMs, which are always beside the steps, suggests that mechanism
one is more likely to occur. On the other hand, if the second proposed mechanism was
the correct, we should be able to observe HOPG surfaces that are fully covered by the

densely packed monolayers, which were not observed in any experiments.
The first mechanism can be divided into two steps: at the first step, the organic
molecules are physisorbed on the HOPG surface; at the second step, the molecules
start nucleation to form SAMs - 2D crystals (Fig 7.2).

Fig 7.2 Illustration of supersaturated solution deposited onto the HOPG surface

Consider the equilibrium between the HOPG surface empty sites S, the occupied
sites SP, and the particles P in the solution, Langmuir equation is applied:
S + P  SP (7-1)
Equilibrium constant K is given by

114
FORMATION OF 2D CRYSTALS ON HOPG
]][[
][
PS
SP
K  (7-2)
Let’s consider the second step: formation of 2D crystals. The top view of the surface
is shown in the Fig 7.3.

Fig 7.3 Top view of HOPG surface with molecules forming SAMs (‘locked’ molecules) and
‘unlocked’ molecules

In this two dimensional model, the number of adsorption sites is N. The number
of the ‘unlocked’ molecules is SP as discussed in equation (7-1). The number of the
‘locked molecules’ forming SAMs is M.
Consider at the equilibrium
N + SP  M (7-3)

The equilibrium constant c is given by the equation:
]][[
][
SPN
M
c  (7-4)
Rearrange (7-2) and (7-4):

115
FORMATION OF 2D CRYSTALS ON HOPG
]][][[
][
NPSK
M
c  (7-5)
Density of the adsorption sites on step [N] is a property of the graphite crystal,
which is considered as not a variable but a constant. [M] is the proportional to
molecules of monolayers, therefore it is proportional to surface coverage . [S] is
proportional to (1-). Thus:
])[1(][
1
PNK
c



 (7-6)
Rearrange (7-6):
][
1

][
1
][
1
1
P
k
PNKc


(7-7)
where
][
1
NKc
k  (7-8)
Based on the proposed mechanism the relationship between the surface coverage
of SAMs and the solution concentration [P] was derived.  is an equilibrium constant
which is determined by many factors (as we will discuss in later part of this chapter).

7.4 Conditions for Formation of 2D crystals
Through numerous experiments carried out in our group, both failed and
successful, together with literature studies, the attempt to figure out the conditions for
formation of 2D crystals on HOPG was made. Based on the equation
][
1
][
1
1
PNKc



(7-7), the value of the coverage  can be used to qualitatively
describe the possibility that a certain molecule can form SAMs on HOPG. If
formation can take place easily, its coverage  tends to be 1. On the contrary, if the

116
FORMATION OF 2D CRYSTALS ON HOPG
molecule is unlikely to form SAMs on HOPG surface, the coverage  tends to be 0.

7.4.1 Concentration
It is very obvious that the possibility of forming SAMs on HOPG is highly
concentration dependent.
][
1
P
k


(7-7)
When the concentration of the molecule increases,
][P
k
will decrease, which leads to
the increase in the surface coverage . In our studies of stearic acid SAMs on HOPG
surface at room temperature, the possibility of observing SAMs at a lower
concentration solution (~10 mg/mL in phenyloctane) was lower than that of 30mg/mL
solution. Meanwhile the solutions at higher concentrations, e.g. 50mg/mL and
80mg/mL, could be observed easily.
Because STM scan maximum size is 240nm240nm, it is difficult to determine

the surface coverage  precisely with studying such small surface area. Combining the
experimental observation and equation (7-10), a rough diagram [Fig 7.4] of the
surface coverage  vs [P] was proposed.

117
FORMATION OF 2D CRYSTALS ON HOPG

Fig 7.4 Propsed diagram of surface coverage  vs solution concentration [P]: an derived  vs [P]
equation based on our proposed model.

Besides the concentration, the surface coverage  is also dependent on the
equilibrium constants. In our model the equilibrium constant  consists of many
factors:
][
1
][
1
1
PNKc


(7-7)
where [N] is the density of the adsorption sites on the graphite crystal surface; K is the
equilibrium constant of reaction (7-1); c is the equilibrium constant of reaction (7-3).

7.4.2 The surface structure of HOPG and [N]
[N] is the density of the kinks on the step of the graphite crystal surface which is
the nature of the graphite crystal. There are several grades of commercially available
HOPG crystals: ZYA, ZYB, ZYD and ZYH [11]. The different mosaic angle and
lateral grain size of these crystals cause the variation of the density of steps. Therefore

118
FORMATION OF 2D CRYSTALS ON HOPG
the density of the [N] is dependent on the grade of the HOPG. In general, surface
coverage  will increase when the density of step [N] on HOPG surfaces increases.
During our STM experiments the HOPG crystals were grade ZYB which was suitable
for critical experimental requirements. Other grade crystals have not been used in
SAMs studies.

7.4.3 Equilibrium constant K of adsorbate/surface interaction
7.4.3.1 Molecular size
Consider the equilibrium between the HOPG surface empty sites S, the occupied
sites SP, and the particles P in the solution (Refer to Fig 7.2), Langmuir equation is
applied:
S + P  SP (7-1)
Equilibrium constant K is given by
]][[
][
PS
SP
K  (7-2)
The equation (7-1) is the equilibrium between the particles P in solution and
physisorbed particles SP on the HOPG surfaces. It is obvious that more products SP
will be formed when the interaction between the substrate and the adsorbate is
stronger. Therefore the value of k can be considered as an index of the strength of the
adsorbate/substrate interaction. The main contribution of the adsorbate/substrate
interaction is from the van der Waals forces between the adsorbed molecules and
HOPG surfaces.
In equation (7-11), surface coverage  increases with increasing in k. In other

119
FORMATION OF 2D CRYSTALS ON HOPG
words, molecules which experience stronger adsorbate/substrate interaction have
higher chance to form SAMs on HOPG. For example, the monolayers of n-C
32
H
66

which could be easily localized on the HOPG surface, while monolayers of n-C
17
H
36

were not as clear as n-C
32
H
66
[12]. This was because n-C
17
H
36
exhibits higher surface
mobility than n-C
32
H
66
due to the weaker adsorbate/substrate interaction. In previous
studies [16, 17] the molecules were substituted with long alkyl chain to increase the
molecular size, which in turn increases the interaction between the molecule and
substrate and possibility for molecules to form SAMs on HOPG. Therefore larger

molecule has larger equilibrium constant k of adsorbate/surface interaction.

7.4.3.2 Flatness
Not only the molecules size matters, but also does its geometrical structure. As
shown in Chapter 5, the derivatives of DDPER comprise large side groups have their
flat center deformed. More than that, the bulky side groups also hindered the major
parts of adsorbates from contacting with the substrates. In the review by Matzger A.J.
[7] most molecules are almost planar. Therefore the flatness of the sample molecule
will increase the equilibrium constant k of adsorbate/surface interaction because of
the better surface contact.

7.4.4 Equilibrium constant c of adsorbate/adsorbate interaction
Consider the equilibrium between the adsorption site N, ‘unlocked’ adsorbates SP
and ‘locked’ adsorbates M based on the model in Fig 7.3:

120
FORMATION OF 2D CRYSTALS ON HOPG
N + SP  M (7-4)
The equilibrium constant c is given by the equation:
]][[
][
SPN
M
c  (7-5)
Similar to K, the value of c can be considered as an index of the strength of the
adsorbate/adsorbate interactions. In equation (7-11), surface coverage  increases with
increasing adsorbate/adsorbate interaction c. Unlike the adsorbate/substrate
interactions, which are mainly van der Waals forces, the interactions between the
adsorbates arise from several interactions, namely: van der Waals forces, hydrogen
bonding, dipole-dipole interaction, and etc.

7.4.4.1 Molecular Structures and Symmetry
Linear organic molecules including saturated and unsaturated n-alkanes, n-alkyl
acids, and n-alcohols exhibit good ability of forming SAMs on HOPG surfaces. The
linear molecules with longer alkyl skeleton are more stable on HOPG surfaces than
analogs with shorter alkyl skeleton [12]. While for molecules with same length, their
structural difference induces the difference in the possibility of SAMs formation. For
example, either stearic acid (n-C
17
H
35
COOH) or elaidic acid (n-C
17
H
33
COOH) can
form stable SAMs on the HOPG surface, but SAMs formed by oleic acid
(n-C
17
H
33
COOH) can not be observed at room temperature [18, 19]. Elaidic acid is a
trans fatty acid while oleic acid is its cis isomer. In 3D space, the elaidic acid crystal
[17] consists of straight aligned acid molecules. Its melting temperature at one
atmosphere is 44°C. For its cis isomer - oleic acid, the crystal consists of bended

121
FORMATION OF 2D CRYSTALS ON HOPG
molecules [18]. The melting temperature of oleic acid at one atmosphere is 13-14°C.
The lower melting temperature of the cis isomer arises from the fact that the ‘U’

shape oleic acid is not able to pack as well as elaidic acid in crystal to maximize the
intermolecular interaction. This explanation is tenable for both 3D crystals and 2D
crystals (SAMs). The oleic acid molecules are not able to pack as dense as elaidic acid
molecules on HOPG surfaces as the steric repulsion between the adsorbed oleic acid
molecules is stronger than elaidic acid molecules. Therefore the elaidic acid has larger
adsorbate/adsorbate interaction constant c. Cai et al. have reported that the iodinated
oleic acid was able to form stable monolayers on HOPG with ‘bent’ molecule within
each unit cell [19]. It was possibly due to the increasing in its molecular weight from
282.46g/mol to 442.26g/mol when two bromine atoms were added. The van der
Waals forces became stronger thereafter.
Among 359 examples listed in Matzger’s review [7] most molecules are linear or
highly symmetric for non-linear ones. Asymmetric compounds exhibit poorer ability
to form SAMs, because it is difficult for them to come closer to max intermolecular
interactions, as roughly illustrated in Fig 7.5: closer packing can be achieved with
molecules A and B but not C.



122
FORMATION OF 2D CRYSTALS ON HOPG

Fig 7.5 Molecular symmetry vs packings density: it is illustrated that a symmetric unit can pack
more densely than an unsymmetric unit in general (note: this is not a universal rule).

Therefore the asymmetry will decrease the equilibrium constant c of
adsorbate/adsorbate interactions. In certain cases the molecule is not rigid. Its
conformation and symmetry can change with different environment. De Feyter et al.
observed dramatic differences in the 2D ordering depending on the nature of the
solvent [20]. By changing the polarity of the solvent, the conformation of the solute
would change accordingly. The different conformations will lead to different patterns

on the HOPG surfaces. Therefore the nature of solution will also affect the symmetry
of the molecule as well as the SAMs pattern.

7.4.4.2 H-Bonding
The presence of hydrogen bonding between adsorbed molecules significantly
increases the strength of intermolecular interaction. Those alkane molecules without
hydrogen bonding can not form SAMs as easily as their acidic analogs. Thus,
hydrogen bonding will increase the equilibrium constant c of adsorbate/adsorbate
interactions.

123
FORMATION OF 2D CRYSTALS ON HOPG
7.4.5 Temperature
Considering the equilibriums (7-1) and (7-4)
S* + P  SP (7-1)
N + SP  M (7-3)
When temperature rises, the solubility of the solution also increases. More molecules
will remain in solution rather than being physisorbed onto the HOPG surfaces. The
equilibrium (7-1) will shift to the left hand side. As [SP] decreases equilibrium (7-3)
will also shift to the left hand side, and cause the melting of the SAMs.

7.5 Common properties of 3D crystals and 2D crystals
3D crystals and 2D crystals have quite a number of common properties. They are
listed in the Table 7.1:
Table 7.1 Common properties of 3D crystals and 2D crystals

3D crystals SAMs/HOPG (2D crystals)
Repeating structures
√ (3D) √ (2D)
Melting temperature

√ √
Space groups
√
230
√
17
Effect of molecular structure
√ √
Since SAMs are considered as 2D crystals, there could be a phase transition
temperature. The melting temperature of alkanes on graphite and other metals have

124
FORMATION OF 2D CRYSTALS ON HOPG
been systematically investigated. Some results are listed in the following table [21]:

Table 7.2 Melting Temperatures (in K) of alkanes in bulk and monolayers on other substrates
[21]

Similar to the trends in 3D crystals system, the melting temperature of 2D
monolayers on graphite increases with the rising molecular weight, indicating there
are stronger adsorbate/substrate interaction and intermolecular interactions for larger
molecules.

7.6 Conclusion and Future Works
We have discussed several factors that can affect the formation of SAMs on the
HOPG surfaces based on our model. The model can be used as a guidance to design
molecules that have a higher chance to form SAMs on a graphite surface. Further
studies on the chemical or physical functions of the monolayers may allow us to
introduce functionality to the monolayers, so that we can synthesize desired SAMs
which may have great potential applications in nano-electronics and biosensors.

125
FORMATION OF 2D CRYSTALS ON HOPG
Reference
[1] Atkins, P.W. Physical Chemistry, 1998, Oxford University Press, p 620.
[2] Lyuksyutov, Two-Dimensional Crystals, 1992, Academic Press, Inc. San Diego
[3] De Feyter, S.; Gesquiére, A.; Abdel-Mottaleb, M.M.; Grim, P.C.M. De Schryver,
F.C. Acc. Chem. Res. 2000, 33, 520.
[4] De Feyter, S.; De. Schryver, F.C. J Phys. Chem. B 2005, 109, 4290.
[5] Wan, L.J. Acc. Chem. Res. 2006, 39, 334
[6] Tao, F.; Bernasek, S.L.; Chem. Rev. 2007, 107, 1408.
[7] Plass, K.E.; Kim, K.; J Matzger, A.J. Acc. Chem. Res., 2007, 40, 287.
[8]
[9] Qian, P.; Nanjo, H.; Yokoyama T.; Suzuki, T.M.; Chem. Commun. 1999, 1197.
[10] Burton, W.K.; Cabrera, N.; Frank, F.C. Phil. Trans. Roy. Soc. London A 1951,
243, 299.
[11]
[12] McGonigal, G.C.; Bernhardt, R.H.; Thomson, D.J. Appl. Phys. Lett. 1990, 57, 28.
[13] Qiu, X.H.; Wang, C.; Zeng, Q.D.; Xu, Bo.; Yin, S.X.; Wang, H.N.; Xu, S.D.; Bai,
C.L. J. Am. Chem, Soc. 2000, 122, 5550.
[14] Schuurmans, N.; Uji-I, H.; Mamdouh, W.; De Schryver, F.C.; Feringa, B.L.; van
Esch, J.; De Feyter, S. J. Am. Chem. Soc. 2004, 126, 13884
[15] Ibino, M.; Sumi, A.; Hatta, I. Jpn. J. Appl. Phys. 1995, 34, 3354.
[16] Ibino, M.; Sumi, A.; Hatta, I. Jpn. J. Appl. Phys. 1995, 34, 610.
[17] Low, J.N.; Scrimgeour, C.; Horton, P. Acta Cryst. 2005, E61, 3730.

126
FORMATION OF 2D CRYSTALS ON HOPG

127
[18] Abrahamsson, S.; Ryderstedt-Nahringbauer, I. Acta Cryst. 1962, 15, 1261.
[19] Cai, Y.G.; Bernasek, S.L. J. Phys. Chem. B 2005, 109, 4514.
[20] Mamdouh, W; Uji-i, H; Ladislaw, JS; Dulcey, AE; Percec, V; De Schryver, FC;
De Feyter, S, J. Am. Chem. Soc. 2006, 128, 317.
[21] Bruch, L.W.; Diehl, R.D.; Venables, A. Rev. Mod. Phys., 2007, 79, 1381
[22] Kaneko, F.; Kobayashi, M. Acta Cryst. 1990, C46, 1490.