Mechanical control of spin states and the underscreened Kondo effect

We have attached single cobalt complexes within mechanically controllable break-junction devices that allow for controlled stretching of the molecule and simultaneous conductance measurements [1]. As we stretch individual spin-1 molecules to alter their symmetry, we find that the molecule’s spin states and magnetic anisotropy can be manipulated in the absence of a magnetic field. This control enables studies of the underscreened Kondo effect, in which conduction electrons only partially compensate the molecular spin. Our findings demonstrate a new mechanism of spin control and establish that single-molecule devices can serve as model systems for precision tests of correlated-electron theories.

People Involved

Joshua Parks, Alexandre Champagne, Theo Costi, Will Shum, Abhay Pasupathy, Eric Neuscamman, Samuel Flores-Torres, Pablo Cornaglia, Armando Aligia, Carlos Balseiro, Garnet Chan, Héctor Abruña, and Dan Ralph


Our device fabrication begins with a growth of 200 nm SiO2 on 200 µm-thick degenerately doped silicon wafers, which are sufficiently flexible to allow for mechanical bending. We then strip the oxide within a 75 µm × 75 µm square window, in which we grow a thinner 40 nm-thick oxide. Using a series of photolithography and electron-beam lithography steps, we define 32 nm-thick Au lines with a 50 nm constriction centered in the window, connected to thicker Au bonding pads lying on the thicker oxide. We then etch the 40 nm oxide from under the Au wires to leave Au bridges suspended 40 nm above the silicon substrate (Fig. 1).

To incorporate molecules in our devices, we clean unbroken wires in an oxygen plasma to remove organic contaminants and then immerse the samples into a <0.1 mM solution of Co(terpyridine-SH)2 in acetonitrile, allowing the thiol end groups of the molecules to attach to the gold. After the sample chip is removed from the molecular solution, excess solution is blown off with nitrogen gas and the samples are cooled to low temperature. We then use electromigration [2] to create a molecular-scale break in the wires. Following the break, the gap spacing between electrodes can be varied by bending the substrate [3].

A measurement of differential conductance (dI/dV) as a function of increasing electrode spacing at a temperature T = 1.6 K is shown in Fig. 2. At the initial position of the electrodes, dI/dV has a single peak centered at zero bias voltage, a signature of Kondo-assisted tunneling through the molecule. As we stretch the molecule, the single conductance peak splits into two beyond a value for the change in electrode spacing that varies from device to device. The observed splitting indicates that the spin degeneracy required for the Kondo effect is lifted, which cannot occur at zero magnetic field in traditional spin-1/2 Kondo systems. In our present work, the peak splitting is instead due to a higher spin, S = 1 Kondo effect together with the breaking of degeneracy within the S = 1 triplet ground state caused by molecular distortion [4]. For an unstretched S = 1 ion in a ligand field with octahedral symmetry, the triplet states are strictly degenerate. However, if the molecule is stretched axially (the z-axis), the Sz = 0 state will be lowered by a zero-field splitting energy D below the Sz = ±1 states, giving rise to magnetic anisotropy. This broken degeneracy quenches the Kondo resonance near zero bias and causes conductance peaks at V = ±D/e due to inelastic tunneling (Fig. 2, inset). Our measurements of the Kondo peak evolution in a magnetic field show good agreement with this anisotropy model.

The temperature dependence of the Kondo signal for the unstretched molecule provides further support that the spin of the molecule is S = 1. For a molecule with S > 1/2 attached to two electrodes with one dominant screening channel, the result is an underscreened Kondo effect, in which the molecular spin is only partially screened to a value S - 1/2. The underscreened Kondo effect is predicted to produce a much slower rise in the conductance as the temperature is lowered, relative to the fully screened Kondo effect [6]. In Figure 3, we plot the normalized conductance G(T)/G(0) versus the scaled temperature T/TK for seven different Co(tpy-SH)2 devices based on separate fits to numerical renormalization group predictions for the fully screened S = 1/2 and for the underscreened S = 1 Kondo models. We find that G(T)/G(0) deviates strongly from the form for the S = 1/2 Kondo effect, and instead agrees quantitatively with the prediction for the S = 1 underscreened case. Our results demonstrate that single-molecule electrical devices can provide well-controlled model systems for studying S = 1 underscreened Kondo effects not previously realizable in experiment. Our work further demonstrates that mechanical control can be a realistic strategy for manipulating molecular spin states, to supplement or replace the use of magnetic fields in proposed applications such as quantum manipulation or information storage [7].

Figure 1
Figure 1: Scanning electron micrograph of a Au bridge suspended 40 nm above a Si substrate.

Figure 2
Figure 2: Differential conductance as a function of bias voltage and electrode spacing at T = 1.6 K. As a metal complex is stretched, a single Kondo peak splits into two separate peaks. Inset: Axial distortion from octahedral symmetry combined with spin-orbit coupling breaks the degeneracy among the S = 1 triplet states.

Figure 3

Figure 3: Normalized conductance G(T)/G(0) versus scaled temperature T/TK for 7 different devices.


  1. J. J. Parks et al., Science 328 1370 (2010).
  2. H. Park et al., Appl. Phys. Lett. 75, 301 (1999).
  3. N. Agraït, A. L. Yeyati, J. M. van Ruitenbeek, Phys. Rep. 377 81 (2003).
  4. R. Boca, Coord. Chem. Rev. 248 757 (2004).
  5. P. Nozières, A. Blandin, J. Phys. (Paris) 41 193 (1980).
  6. F. Mallet et al., Phys. Rev. Lett. 97 226804 (2006).
  7. L. Bogani, W. Wernsdorfer, Nature Materials 7 179 (2008).

Last updated: 19-June-2010

contact info