Over the past two decades, organic molecules adsorbed on atomically defined metal
surfaces have been intensively studied to obtain an in-depth understanding of their
self-assembly behavior, on-surface reactivity, as well as their structural and electronic
properties [1–6]. An important aspect to unravel their potential use in electronic
and optoelectronic devices is how their functionality can be preserved when adsorbed
on surfaces. Unfortunately, the (strong) interaction of the molecules with the metallic
surface, for example, due to hybridization of molecular states with electronic bands
from the metallic substrate, often alters the electronic properties of the molecules
and, moreover, can even turn off their sought-after functionality. As a result of
the (strong) interaction, the molecular scaffolds can also become distorted, electronic
states may be significantly broadened and shifted, and vibronic states may even be
quenched. Decoupling strategies offer unique opportunities to reduce these (strong)
interactions. In the following, recent progress to decouple both single molecules
and molecular assemblies physically and electronically from a (strongly) interacting
support is briefly reviewed.
The molecule–substrate decoupling can either be achieved by suitable molecular design
or manipulation of the molecular building blocks, or passivation of the substrate.
Molecular engineering approaches include rigid spacers added to the molecular core
[7–10] or specially designed double-decker molecules [11] to maintain the active part
of the molecule sufficiently high above the surface to prevent it from interacting
with the latter. Alternatively, molecular/oligomeric structures can be attached to
the tip of a scanning probe microscope and mechanically lifted from the metallic surface
such that they hang freely between metal contacts. This manipulation technique allows
for measuring, amongst others, the electronic conductance, magnetic properties, reversible
switching, and electroluminescence of free-standing molecules [12–16]. Physical decoupling
strategies involving the design or manipulation of the molecular building block significantly
limit the selection of building blocks and often hinder probing molecular properties
with intramolecular resolution. Therefore, modifications of the substrate, for instance
by adding or intercalating a decoupling layer, are often the better choice. In the
best case, these interfacial layers have a large bandgap to prevent a hybridization
with molecular states as well as with the metallic/semiconducting substrate. All the
strategies for physical and electronic decoupling have been developed in view of fundamental
studies as well as application in devices.
Ultrathin semiconducting or insulating decoupling layers can be epitaxially grown
as mono- and multilayers on many metallic substrates by either physical or chemical
vapor deposition. Among others, ultrathin dielectric layers of either alkali halides
(e.g., NaCl [17]) or metal oxides (e.g., MgO [18], Al2O3 [19], and CuO [20]), or nitrides
(CuN [21]) have been shown to be beneficial for successfully reducing or even completely
switching off the unwanted interaction between the metal substrate and the organic
building blocks. Recently, two-dimensional (2D) materials, including hexagonal boron
nitride (hBN) [22–23], graphene [24–27], and MoS2 [28], have emerged as monatomically
thin decoupling layers. Van der Waals 2D materials are generally well suited due to
their chemical inertness and the low density of states near the Fermi level. However,
the electronic decoupling efficiency also depends on the electronic structure of the
2D material. Sometimes, only molecular states in the bandgap of the 2D material can
be decoupled. Moreover, ultrathin organic spacer layers can efficiently electronically
decouple further organic layers from a metal surface. The decoupling strategy with
organic layers relies on both an increased separation between the organic layer of
interest and the metal as well as the different electronic gaps of the organic interface
layer and following ones [29–32]. Additional concepts to weaken adsorbate–surface
interactions involve the post-deposition intercalation of atomic species such as iodine
[33].
For semiconductors, for example, bare silicon or germanium, electronic decoupling
of molecules can be achieved by either the growth of ultrathin dielectric layers on
top of the surface [34–35] or a chemical modification of the surface to saturate the
dangling bonds. In surface-science-based studies, for the latter approach hydrogenation
of semiconductor surfaces is frequently applied as effective passivation against chemisorption
of adsorbates [36–39], while also B deposition was shown to result in effective passivation
of the Si surface [40–41]. In particular for electronic devices, oxidized semiconductor
surfaces (e.g., silicon dioxide layers formed on bare silicon) are mostly used as
substrates for fabricating devices [42].
Most of these ultrathin interfacial layers significantly reduce both the molecular
adsorption energy and the hybridization of molecular states with the electronic bands
in metals and semiconductors. However, charge transport is not completely inhibited,
and electrons can still tunnel from the metal through these decoupling layers to the
organic molecules and vice versa. This has the great advantage that these systems
can still be examined by scanning tunneling microscopy (STM) and spectroscopy (STS),
which gives insight into structural and electronic properties of individual molecules.
For applications such vertical tunneling through the interfacial layers can be undesirable
since it prevents the current from flowing across electrode–molecule–electrode junctions
[43] or leads to the charging of physisorbed molecules on top of such layers [18].
Hence, ultrathin insulating layers are often not sufficient to truly electronically
insulate molecular structures. For this, the usage of either thicker insulating films
(mostly alkali halides) or bulk insulators is required.
For studying molecules on bulk insulators, ionic crystals (e.g., KBr, NaCl, CaF2,
and calcite) have mostly served as model systems. To a lesser extent, metal oxides
have also been used, for which defects and charging often pose additional challenges
[44–46]. On electronically insulating surfaces, non-contact atomic force microscopy
(AFM) is the method of choice to study molecular assemblies and individual molecules
in real space. Molecular adsorption and self-assembly are significantly altered compared
to metals due to an often weak interaction between organics and bulk insulators. In
contrast to face-on adsorption on metals, a tilted and edge-on adsorption becomes
possible for planar aromatic molecules on bulk insulators. Due to lowered diffusion
barrier and adsorption energy, the two-dimensional molecular layers can be affected
by dewetting and may change into three-dimensional clusters [47]. In return, the reduced
molecule–surface interaction on insulating films or bulk insulators can stabilize
highly reactive molecules, offering unique opportunities for the bottom-up assembly
of novel carbon-based materials using on-surface chemistry [48–49]. However, the significantly
reduced catalytic activity on non-metallic substrates requires exploring alternative
reaction mechanisms beyond thermal activation, for example, photon-induced reactions
[46,49–51] or electron-induced reactions by electrons from a probe tip [52–53].
In the following, we would like to highlight a few examples for physical properties
and built-in functionalities of molecular systems, which were successfully accessed
by employing one of the decoupling strategies mentioned above. Electronic decoupling
significantly increases the lifetime of excited molecular states and improves the
effective energy resolution (down to a few millielectronvolts) of molecular resonances
observed in tunneling spectroscopy since the hybridization of molecular states with
the ones of the metallic/semiconducting support is prevented. Thereby, it became feasible
to investigate molecular orbitals [17], observe and switch the charge state of individual
molecules [54–59], and resolve individual vibronic states in single molecules and
molecular assemblies [23,28,60–61]. Similarly, the lack of electronic states around
the Fermi level in a superconductor was used to preserve electronic properties in
adsorbed molecules. For example, the spin relaxation in magnetic molecules was suppressed
on a superconducting surface, which then resulted in a significant enhancement of
the excited-state spin lifetimes [62]. Concerning optical properties, successful decoupling
made the examination of fluorescence from both single molecules and molecular assemblies
feasible by tunneling electron excitation [19,63–66]. Also, sub-molecularly resolved
Raman images were successfully demonstrated [67]. Additionally, electroluminescence
was reported for suspended molecular wires between a metallic surface and the tip
of a scanning tunneling microscope [14]. Moreover, the decoupling brings considerable
benefits for surface-supported molecular switches, in particular for preserving reversible
switching capabilities [68–69], but also conformational [9–10
70], tautomeric [71], and charge state switching could be shown [56].
This Thematic Issue highlights recent experimental and theoretical developments in
realizing and understanding physically and electronically decoupled single molecules
and molecular assemblies on surfaces. Several decoupling strategies at the solid–vacuum
and solid–liquid interface were explored to elucidate structural, electronic, vibronic,
and chemical properties of decoupled molecular structures.
Physical decoupling by molecular design often relies on non-planar adsorbates with
bulky spacer groups, which can adopt various conformations. From a theoretical point
of view, finding the energetically most stable conformational structure can be challenging
and costly because conventional atomistic simulations are often limited to the partial
exploration of the potential energy landscape due to the complexity of the system.
Recently developed structure search methods that combine machine learning with density
functional theory provide the possibility of reliable structure identification of
non-planar molecules, as demonstrated for the example of (1S)-camphor on Cu(111) [72].
Such computational tools become relevant for molecules with bulky spacer groups since
they are very valuable for predicting and interpreting the structural and conformational
properties as well as the decoupling of such molecules on surfaces.
With an appropriate molecular design, the built-in functionality of the active part
of the molecule can be preserved upon adsorption on a surface. An example of the preservation
of catalytic properties is demonstrated for the redox behavior of manganese porphyrins
at the solid–liquid interface. Redox reactions at the axial ligands attached to the
metal center of the porphyrin were observed regardless of the type of surface (highly
oriented pyrolytic graphite (HOPG) and Au surfaces were used), solvent (1-phenyloctane
and n-tetradecane) and tip material (Pt/Ir, Au, and W), which indicates that the ligands
have to be decoupled from the substrates [73].
Suitable functionalization of molecules is another concept to vary their adsorption
strength on metal substrates. For instance, partial fluorination of pentacene molecules
decreased the adsorption strength on strongly interacting substrates such as Cu [74]
but did not result in notable effects on Ag(111) [75]. Although this decoupling concept
is only practical on Cu, the fluorination significantly changed the molecular multilayer
growth on Ag(111) and led to a physical decoupling with a nearly bulk crystalline
structure for the fluorinated pentacene.
Two articles within this Thematic Issue discuss structural templating effects at the
solid–liquid interface by systematically looking at the influence of organic decoupling
layers. Reynaerts et al. [76] investigated the suitability of long-chain alkanes as
physical decoupling layers from a graphite surface. The occurrence of the same polymorphs
for 4-tetradecyloxybenzoic acid assemblies in the presence and absence of the long-chain
alkane buffer layer indicated that the influence of the substrate could not solely
explain the self-assembled structures. However, the alkane buffer layer provided the
possibility to monitor the STM-induced nucleation, growth, and ripening of self-assembled
monolayers in a more controlled fashion. Söngen et al. [77] provide insight into the
interaction of organic molecules with bulk insulators by discussing the adsorption
of ethanol on both calcite and magnesite using three-dimensional AFM experiments.
Although molecules adsorbed on bulk insulators are electronically decoupled, molecular
self-assemblies can experience a substrate templating effect due to the presence of
heterogeneous adsorption sites. Therefore, Söngen et al. [77] found on bulk calcite
and magnesite that the first ethanol layer arranges in a laterally ordered way due
to ionic interactions, where ethanol adopts well-defined adsorption positions on the
carbonate surface. In contrast, the following layers lack this order as they reside
on ethanol layers. Hence, they experience a physical decoupling due to the changing
chemical environment.
Next, we outline articles that use 2D materials and ultrathin dielectric layers as
decoupling layers. While on the one hand, molecular functionalization is a powerful
approach to tune the electronic and optical properties of 2D materials, in particular
for many practical applications [78], 2D materials, on the other hand, offer an alternative
way for decoupling molecular structures from metal substrates [24]. 2D van der Waals
materials are generally inert and therefore, are potentially well suited for physical
decoupling of molecular structures. However, moiré patterns present due to the lattice
mismatch between 2D material and its substrate might serve as structural templates
for molecular adsorption and self-assembly [79–82]. The electronic decoupling depends
on the electronic properties of the 2D materials as they can be insulators, semiconductors,
semimetals, or metals [83].
Rothe et al. [84] demonstrated that semimetallic graphene is an appropriate buffer
layer for the physical and chemical decoupling of rubrene from Pt(111). The strong
molecule–surface interaction on Pt(111) is expressed by hit-and-stick adsorption due
to a substantial diffusion barrier. In contrast, on graphene/Pt(111) the growth of
molecular domains is facilitated. Electronically, the width of the highest occupied
molecular orbital (HOMO) resonance is reduced by a factor of ten on graphene/Pt(111)
compared to bare Pt(111) due to a reduction of the molecule–surface hybridization.
The significantly reduced resonance width allowed for resolving vibronic states in
both frontier orbitals on graphene/Pt(111) by STS.
The semiconducting 2D material MoS2 may act as a decoupling layer for molecules from
the underlying metal substrate if the molecular resonances lie within the MoS2 bandgap.
Hence, Yousofnejad et al. [85] found using MoS2 on Ag(111) as substrate that the HOMO
of tetracyanoquinodimethane (TNCQ) is not decoupled because it is located in the MoS2
valence band, while the lowest unoccupied molecular orbital narrows but still suffers
from lifetime broadening because it is situated at the conduction band onset of MoS2.
Despite this, the vibronic states of the transiently negatively charged TCNQ could
be resolved by STS.
hBN is an insulator and has therefore been widely used to decouple organic molecules
from metal substrates. Three articles within this Thematic Issue successfully employed
hBN to investigate the pristine properties of particular molecules. Schaal et al.
[86] showed that hBN on Ni(111) electronically decoupled tetraphenyldibenzoperiflanthene
such that the molecular vibronic progression was observable by in situ differential
reflectance spectroscopy, which is otherwise only achieved for multilayers on the
bare Ni. On hBN/Cu(111), Zimmermann et al. [87] could visualize the molecular orbitals
of pyrene derivatives by STM at the submolecular level, while Brülke et al. [88] measured
the fluorescence of monolayer perylenetetracarboxylic dianhydride (PTCDA), which is
quenched on bare Cu(111) and would require three molecular decoupling layers to be
probed on Cu(111). In all three studies using hBN, ordered molecular films were observed.
The decoupling even allowed for the formation of complex self-assemblies such as kagome
lattices by tuning the number and position of the substituents of the pyrenes derivatives
[87].
Metal oxide thin films on top of metal substrates are another interesting class of
ultrathin interfacial layers to decouple organic molecules and to enable the study
of their electronic properties without the contribution of the underlying metal substrate.
Hurdax et al. reported that both charged and neutral species of sexiphenyl can co-exist
on thin dielectric MgO films on Ag(100) [89]. Due to the changed work function of
the substrate, charging of the adsorbates is enabled by electron tunneling. The charge
transfer strongly influences the molecular conformation by planarizing the carbon
backbone as well as the self-assembly. Hurdax et al. [89] suggested that work function
measurements before and after the adsorption of molecules should give insight into
the electronic and physical decoupling. It should be noted that metal oxide thin films
feature heterogeneous adsorption sites, which can lead to the anchoring of organic
molecules having specific functional groups. Hence, the interplay of the potential
energy landscape of the substrate and the intermolecular interactions steers the self-assembly
in such systems. Xiang et al. [90] studied these aspects for the self-assembly of
porphyrin derivatives on cobalt oxide films on top of Ir(100). While the unfunctionalized
diphenylporphyrin self-assembled on the bilayer film but not on the two-bilayer film,
the opposite observation was made for cyanotetraphenylporphyrin.
Physical decoupling of molecules from a semiconducting substrate is discussed for
the example of both insulating CaF2 thin films on Si(111) [91] and hydrogen passivation
of Ge(001) surfaces [92]. In the first case, three scenarios were compared: PTCDA
on Si, on a thin CaF2, and on a thicker CaF2 layer. While isolated PTCDA molecules
were pinned to defects on Si and also on the thin CaF2 layer, PTCDA was physically
decoupled via the thicker CaF2 films and self-assembled into small islands. For FePc
on H-passivated Ge(001), efficient physical decoupling facilitated the growth of large
islands with upright oriented molecules, similar to their arrangement in molecular
crystals.
In summary, the articles collected in this Thematic Issue highlight recent experimental
and theoretical developments in the atomic- and molecular-scale understanding of physically
and electronically decoupled single molecules and molecular assemblies on surfaces.
Over the last decade, significant progress in this field led to a manifold of decoupling
strategies for single molecules and surface-supported molecular architectures. Decoupling
strategies are highly relevant to preserve the intrinsic structural, electronic, and
optical properties of the molecules for performing insightful fundamental surface-science-based
studies on them and, thus, play a crucial role in designing new molecule-based electronic
and optoelectronic devices.
Sabine Maier and Meike Stöhr
Erlangen and Groningen, July 2021