We perform electromigration experiments on this self-organized silver nanowires in-situ in a scanning electron microscope (SEM). Therefore, the wires have to be electrically connected to our Keithley 2400 Source-Measure-Unit. Since the wires are randomly distributed over the silicon samples a contact structure has to be attached to each wire separately. In a first step the position of some silver wires and silver islands is determined and a contact layout is designed. Thereafter, a standard electron beam lithography (EBL) process is carried out and a four point contact structure (gold or silver contacts) is attached to the silver wire with macroscopic bond pads. Using commercially available chip carrier and a standard bonding technique electrical contact is established and the electromigration experiments can be performed.

 

This SEM image shows a silver nanowire after the gold contact structure is applied by means of EBL. The wire has a width of 390 nm and a length of approximately 10 µm. One can see that the gold covers the silver structures completely. The smaller gold stripes are the voltage leads whereas the bigger gold pads are the current leads.

 

The series of SEM images above shows a silver nanowire during different stages of an electromigration experiment. The values of the applied current and the time during which this current was applied are given at the right side of the image. The direction was reversed several times during this experiment. We observe the formation of voids at the anode of the wire. Two things are remarkable in this experiment:

First: The electromigration behaviour is reversible when the current is reversed. Voids, which have been developed during one part of the experiment, are closed and new voids form at the other side of the wire.

Second: The observation of voids in electromigration experiments usually takes place at the cathode. The atoms in the wire move to the anode due to the so called electron wind force; vacancies move to the opposite direction which is the cathode. In our experiment we observe the void formation at the anode. Therefore, the wind force cannot be the dominating driving force in our electromigration experiments

This unusual electromigration behaviour will be studied in the future in more detail. Special interest will be paid on the role of the contact material, since the theory predicts reversed electromigration behaviour for small contaminations and special experimental conditions. Also the reversibility of the mass transport has to be investigated in more detail.