Molecules on surfaces

1. Introduction.-

The study of molecules on surfaces has had enormous thrust from the coating industry. The interest in preparing special coatings with design properties has heavily relied in the use of molecules because on one hand, molecules are very flexible with enormous possibilities regarding their properties, on the other hand, chemical procedures can become very efficient, reducing the manufacturing process of any coating layer. Fundamentally, molecules on surfaces have undeniable appeal.

Molecules are complex quantum objects that can be tailored in many different ways. From molecules displaying exotic magnetic states, to molecules strongly binding to specific sytems, passing by extraordinary dynamical properties. Molecules have been heralded as the future of minute electronic circuits. This has steered the field of molecular electronics. The magnetic properties of molecules have further expanded the field into the spintronics body of research.

Quantum computation has seen its first candidates in molecules protecting qubits encoded in f-electron atoms inside the molecules. Molecules are powerful and ubiquitous. In most of the applications we have described above, the molecule is in contact with a solid interface. Indeed if a molecular device is to be created, the device needs to be interfaced to the macroscopic world. The solid-surface/molecular interface is unavoidable.

And the surface brings in a lot of interesting physics. The reduced dimensionality of the surface contains its own field of research. The reactivity of the surface itself is an important field in some areas of chemistry. The conduction features of devices do depend on these properties. It is then a field in itself: molecules on surfaces.

We have developed a lot of expertise using different tools. Our workhorses are both the STM and density functional theory.

2. Vibrational properties of molecules.-

The first example of IETS using the STM, addressed the excitation and detection of vibrational modes in acetylene molecules adsorbed on a Cu (100) surface. The webpage of the Wilson Ho group features many interesting examples of molecules studied with the STM, particularly exploiting the vibrational signals under an STM tip.
The vibrational fingerprint of a molecule permits us to fully characterize the molecule as well as the environment where the molecule is. Vibrational modes can harden or soften (increase or reduce their frequencies respectively) depending on the bonding and charge transfer when the molecule sits on a surface. In order to understand the conformation, bonding and general features of a molecule that has bound to a substrate, the vibrational properties become particularly informative.

But vibrations have other appealing aspects. By exciting repeatedly a given vibration, energy can be efficiently transfer from tunneling electrons into the molecule. When the STM bias hits the energy of a mode, electrons can transfer a full quantum of vibration. This can be a lot of energy that is transferred in a single step. The largest vibrational mode is the hydrogen stretch mode. In the gas phase, this mode takes place at 500 meV. All other modes of any molecule will be at lower energy because nothing is as light as Hydrogen. However, many modes involving H atoms can in the hundred meV range. C--H bonds are a particularly ubiquituos mode involving basically all organic molecules and many other. This mode easily lies between 350 and 400 meV. A single electron exciting this mode will transfer almost 400 meV to a molecule in a single process. This is a very large amount of energy, comparable to the adsorption energy of molecules on surfaces. If the mode can be repeatedly excited, the transferred energy reaches levels where harder bonds can start breaking.

This example shows the translation of ammonia molecules on a Cu (100) surface by using tunneling electrons. The stretch mode associated with the N--H mode of the molecules lies at about 410 meV. This energy is above the energy needed to overcome the barrier between adsorption sites on the surface, as a consequence the molecule could hop.

3. Magnetic properties of molecules.-

In the same way that new electronic channels open when a molecular vibration  is excited, spin IETS consists in opening new channels by producing magnetic excitaitons. Molecules can have clear spin states at different energies due to the intrinsic magnetic anisotropy of the molecule. The anisotropy is caused by the spin-orbit interaction that couples the geometry of the molecular electronic states (the orbital degrees of freedom) with the electronic spin.

Axial anisotropy is the simplest case. The molecular geometry is characterized by a prevailing axis such that spins will try to align with it (easy axis) or avoid it (easy plane). Typically the anistropy is given by a longitudinal term that takes into account whether the system is easy or hard axis. The corresponding term written for the the spin degrees of freedom is:
where the z-axis is chosen along the special axis of the molecule. If D is a positive number, the contribution to the energy will increase if the spin aligns along the axis: the larger the Sz component, the more excited the state is. The ground state will correspond to the smallest component of the spin along the molecular axis. Easy axis corresponds to negative values of D.
This equation contains two important facts. The first one is that the spin excitation does not correspond to a change of the total spin (S^2) of the molecule, but rather to a change of orientation of the spin. This is what we expect from mangetic anisotropy. The second fact is that a spin 1/2 cannot have any anisotropy at all and therefore, no excitations!

The plane perpendicular to the special axis of the molecule can also have an internal anisotropy. This has to be smaller than the longitudinal anisotropy because otherwise we made a mistake asigning the "large anisotropy" axis. The transversal anisotropy contains the "transversal" spin components (Sx and Sy or S+ and S-, where S+ = Sx + i Sy, and S- = Sx - i Sy ):
A thorough account of the "magnetic landscape" due to the coupling of the geometry of the arrangement of atoms and the spoin-orbit interaction can be found in our recent review article.

The magnetic spectra of molecules can strongly depend on their environment. On the one hand, we have the effects of goemtry that we just briefly mention to illustrate the effect of spin excitations due to preferential directions. On the ohter hand, we have the effects of bonding and charge transfer that can lead to different electronic ground states. A negative molecule will show a spin very different from the spin of the molecule in the gas phase simply because its electronic configuration has one more electron. A positive molecule too. Partial charge transfer and charge reorentation will also take a toll on the final spin. We have doing serious research in the last year to obtain as much information as possible between the molecular spin as shown by magnetic IETS and the chemical properties of molecules on surfaces.

A particular example of new ground state due to charge transfer and spin change is the case of the Kondo state. Please refer to the Kondo section below.

4. Manipulation of single molecules.-

In the vibrational section above, we have seen that modifications on the single-molecule level can be produced by injecting electrons that are going to excite certain energetic modes, or repeatedly excite the same mode until the molecule has enough energy to overcome a barrier. This is the inelatic tunneling electron mode for manipulating molecules. It goes beyond IETS because a strong perturbation of the molecule is effected. As a consequence the molecule undertakes an evolution that can be sometimes irreversible.

However, it is interesting to manipulate molecules in the best possible controlled way. The reason to do this is besides exploring the chemical bond and chemistry on the single-molecule scale, to be able to modify condensed matter by acting on the smallest possible building blocks. In an attempt to draw the attention to the enormous interest in this type of research, the Nanocar race is becoming an international event to show the great control that STMs can grant on the basic units of condensed matter; molecules.

Our project MeMo goes a step further, doiong research in the way we can build molecular motors. This is an exciting topic that has seen great progress in chemistry trying to mimick the biological world. Our approach is rather based on the STM. This requires ultra-high vacuum and very low temperatures because initially we try to rationalize the processes that can lead us to obtain motion of molecules or of assembles of atoms in a desired way. Motors will be achieved if we can use molecules to produce a measurable work. An important aspect of this research is qunatum dynamics. Molecules live in the quantum world and we can take advatange of that to search for new stratgies in the way a useful work is performed. Additionally, the use of a holding substrate imposes its own contrains. Substrates are a source of interactions that can damped molecular motions or induce decoherence of the molecular quantum properties. This field of research is starting and is very exciting.