prototypical insulating surfaces such as KBr(001), NaCl(001), or CaF 2 (111) is
rather weak [292, 293]. This fact constitutes a severe challenge when aiming at
complex structure formation because molecules tend to de-wet and pile up to form a
3D molecular bulk crystal.
Especially when aiming at thermal activation of an on-surface linking reaction,
this rather weak molecule–surface interaction poses problems because many
molecules desorb from the substrate upon annealing at temperatures well below
the reaction temperature. Moreover, it is known that metal atoms act as catalysts in
the dehalogenation that has been demonstrated before [285, 286]. Thus, as a first
step, exploring suitable anchor groups turned out to be mandatory. Based on our
previous results using the calcite as a substrate [294, 295], we identify the natural
cleavage plane of calcite, namely calcite(10.4) (Fig. 67a), as an ideal substrate
surface due to its comparatively high surface energy of 0.590 J/m
2 [295].
As simple model monomers, we explored the applicability of small benzoic acids
derivatives (Fig. 67b), which were chosen for three reasons. First, these molecules are
available in high purity and can readily be sublimated under ultrahigh vacuum
conditions. Second, carboxylic acid groups are known to interact strongly with the
calcite(10.4) surface [296]. This comparatively high molecule–substrate binding
strength is employed to prevent clustering at step edges and to avoid desorption
upon thermal activation of the coupling reaction. Depending on the acidity of the
benzoic acid derivative, the as-deposited molecules can be deprotonated at room
temperature or might require annealing for the deprotonation step to occur. We could
directly follow such a deprotonation step at room temperature upon deposition of
2,5-dihydroxy benzoic acid having a pK A value of 2.97 [297]. This study gives a first
estimation of the protonation state of other benzoic acid derivatives as a function of
the pK A value. For all molecules shown here, deprotonation takes place on a surface
held at room temperature when the pK A value is around 3 or smaller. For molecules
having a higher pK A value, annealing is required to induce deprotonation. Third, by
varying number and position of the halide substitution, we are able to rationally
design the resulting polymer architecture.
6.2 Proof of Concept: Dehalogenation and Covalent
Coupling on an Insulating Surface
For providing a proof-of-concept for covalent linking on a bulk insulator surface,
we present the results of a systematic noncontact AFM study, investigating the
reactions of four different halide-substituted benzoic acids. The molecules used
were 4-iodo benzoic acid (IBA), 2,5-diiodo benzoic acid (DIBA), 2,5-dichloro
benzoic acid (DCBA), and 3,5-diiodo salicylic acid (DISA), as shown in
Fig. 67b. Within this series of systematically varied benzoic acid derivatives, IBA
can be regarded as the conceptually simplest molecule. As only one halide atom is
available for the linking reaction, dimer structures are expected. Based on the pK A
196
K. Binder et al.
rather weak [292, 293]. This fact constitutes a severe challenge when aiming at
complex structure formation because molecules tend to de-wet and pile up to form a
3D molecular bulk crystal.
Especially when aiming at thermal activation of an on-surface linking reaction,
this rather weak molecule–surface interaction poses problems because many
molecules desorb from the substrate upon annealing at temperatures well below
the reaction temperature. Moreover, it is known that metal atoms act as catalysts in
the dehalogenation that has been demonstrated before [285, 286]. Thus, as a first
step, exploring suitable anchor groups turned out to be mandatory. Based on our
previous results using the calcite as a substrate [294, 295], we identify the natural
cleavage plane of calcite, namely calcite(10.4) (Fig. 67a), as an ideal substrate
surface due to its comparatively high surface energy of 0.590 J/m
2 [295].
As simple model monomers, we explored the applicability of small benzoic acids
derivatives (Fig. 67b), which were chosen for three reasons. First, these molecules are
available in high purity and can readily be sublimated under ultrahigh vacuum
conditions. Second, carboxylic acid groups are known to interact strongly with the
calcite(10.4) surface [296]. This comparatively high molecule–substrate binding
strength is employed to prevent clustering at step edges and to avoid desorption
upon thermal activation of the coupling reaction. Depending on the acidity of the
benzoic acid derivative, the as-deposited molecules can be deprotonated at room
temperature or might require annealing for the deprotonation step to occur. We could
directly follow such a deprotonation step at room temperature upon deposition of
2,5-dihydroxy benzoic acid having a pK A value of 2.97 [297]. This study gives a first
estimation of the protonation state of other benzoic acid derivatives as a function of
the pK A value. For all molecules shown here, deprotonation takes place on a surface
held at room temperature when the pK A value is around 3 or smaller. For molecules
having a higher pK A value, annealing is required to induce deprotonation. Third, by
varying number and position of the halide substitution, we are able to rationally
design the resulting polymer architecture.
6.2 Proof of Concept: Dehalogenation and Covalent
Coupling on an Insulating Surface
For providing a proof-of-concept for covalent linking on a bulk insulator surface,
we present the results of a systematic noncontact AFM study, investigating the
reactions of four different halide-substituted benzoic acids. The molecules used
were 4-iodo benzoic acid (IBA), 2,5-diiodo benzoic acid (DIBA), 2,5-dichloro
benzoic acid (DCBA), and 3,5-diiodo salicylic acid (DISA), as shown in
Fig. 67b. Within this series of systematically varied benzoic acid derivatives, IBA
can be regarded as the conceptually simplest molecule. As only one halide atom is
available for the linking reaction, dimer structures are expected. Based on the pK A
196
K. Binder et al.