The Low Temperature Laboratory of the UAM (LBTUAM) was established in 1971by Sebastian Vieira, in the Department of Physics, which was headed by Professor Nicolás Cabrera. Sebastian Vieira had as collaborator Mariano Hortal, orienting initially their research to the study of thermal properties, specific heat and thermal expansion of disordered solids for T> 1K. The LBTUAM is characterized, since its very beginnings, by a continued effort in the development of scientific instrumentation, training technicians in materials, welding and cryogenics. Raul Villar joined the LBTUAM as a PhD student a few years after its foundation. In the seventies the liquid helium was brought by a company from France in cryogenic containers of 50 l. These were difficult times. Every order of liquid caused a chain of problems which inevitably lead to containers with a strongly decreased volume of liquid helium when they finally arrived at the laboratory. In the late seventies the laboratory, which was in the C-4 building, was moved to its current location. An important step was the acquisition at a very reasonable price of a TBT liquefier, producing two liters of liquid He per hour. This equipment was unused at the University of Essex. Thanks to the expertise of M.Hortal, who led the operation, and R.Villar, the liquefier was moved from Essex to the LBTUAM. Throughout this process we had the technical assistance of the technical staff of the Department, in particular of Andres Buendia who, from the very first moment, took over the management and maintenance of the liquefier. This step allowed a supply of liquid He covering the needs of the laboratory and even meet the needs that arose in other departments of the UAM. The creation in the eighties of SEGAINVEX with an excellent cryogenic service, allowed the use of helium to be extended considerably to reach the current situation. The SEGAINVEX has become, under the direction of M. Pazos, an invaluable support for the experimental investigation of quality. Among its excellent technicians, three, M.Pazos, M.Zayas and J.del Val, have been essential for many early LBTUAM developments. Over the last thirty years there have been two other important steps towards lower temperature limits. The first was made possible by the donation by the Humboldt foundation, as a research aid to R.Villar, of a 3He cryostat, which enabled to reach 0.3 K. The second step was the commissioning of several dilution refrigerators 3He in 4He, which can reach temperatures in the range of thousandths of a Kelvin. This step was made possible by the expertise of H. Suderow. The LBTUAM has actively participated in some of the most interesting condensed matter physics achievements in the last quarter of the twentieth century and so far this century. The birth in the early eighties of the world of nanoprobes with the scanning tunneling microscope (STM), produced a switch from an important part of the laboratory activity towards the development of these techniques at low temperatures. These developments were important for research on high temperature superconductivity in oxides, and, subsequently, for several studies in multiband superconductivity. Also, with mixed techniques based on STM and atomic force microscope (AFM) members of the LBTUAM made outstanding contributions to the knowledge of the mechanical and electronic transport properties, of atomic-size contacts and chains of atoms. Remarkable and pioneering progress has been made on the understanding and applications of the technique of tunneling microscopy and spectroscopy with superconducting tips as well as in the physics of nanoscale superconducting junctions. In the last ten years there have been advances in the study of two-dimensional vortices in superconductors and in transport through molecules. Although the main current effort is related to the precedeeing lines, the LBTUAM has kept interest in the measurement of thermal properties of solids at low temperatures, and new experimental techniques for studying the transport in tunnel junctions, including electronic noise have been developed with success. At the present several groups have been formed within the LBTUAM through the efforts of students who were trained in it, technicians and those who are his current leaders: N.Agräit, F. Aliev, M.A. Ramos, J.G.Rodrigo, G.Rubio-Bollinger, H. Suderow. The LBTUAM has maintained a close relationship with the Institute of Materials Science, CSIC Madrid, forming the GIBT.
Scanning Tunneling Microscopy (SPM) below 1 K offers important surface characterization possibilities in, e.g. quantum nanostructures in semiconductors, or in superconductors. We have succesfully designed and build systems to make STM experiments from 100 mK on and under high magnetic fields.
When the spatial resolution approaches atomic sizes in STM, it becomes particularly important to measure at cryogenic temperatures. Thermal motion of atoms is strongly reduced or blocked, and new phenomena such as magnetic or superconducting order, or other quantum effects, appear. The cryogenic environment provides considerable gain in energy resolution when making tunneling spectroscopy. At 100 mK, the local density of states of a metallic sample can be measured with an energy resolution of some tens of microeV. This makes such machines a unique probe of matter at the local scale. Design of a SPM at low temperatures requires carefully examining the problem of reducing its sensitivity to vibrations.
A Scanning Probe Microscope is usually based on a positioning system, whose location on top of a holder is controlled via coarse approach motors. The positioning system (see figure left, x,y and z in a, most often it is a piezotube) is used to scan the microscopic probe at a given distance over the sample, and the coarse approach (X,Y,Z) is used to reach conditions which give the desired signal, as the tunneling current in STM, and allows for changes in the position of the tip over the sample. Both positioning systems are firmly fixed to the same frame, which maintains the distance between tip and sample constant.
In a tunneling microscope, the tunneling current It(z) depends exponentially on the distance between tip and sample. The tip-sample system should be built as rigid as possible to avoid vibrations of the relative distance between tip and sample, and it should be suspended on an appropriate system with low resonance frequency (figure left).
The limitations due to the support systems are particularly important in a dilution refrigerator, especially when a superconducting magnet is required for magnetic field measurements. The STM is usually hanging on the bottom of the long pumping lines, inserted at the centre of the magnet inside a helium vessel. Vibrations of the STM support are produced by the building, by acoustical noise, by the pumps, and by the circulating cryogenic fluids (helium-4 and helium-3). Vibrations from pumps, mostly transmitted through the pumping lines, can be reduced by using long pumping lines with large diameters.
The microscope head is located on the bottom of a copper support which is attached to the cold point. The microscope head generally consists of a piezotube inserted inside a prism, which is held to a piezoelectric motor using a spring. The microscope is controlled via a home made electronics, with DACs and ADCs, which are suitable chosen between home-made system and commercial data acquisition cards. RF filters reduce noise prior to entering low temperatures.
We have built several models of Scanning Probe Microscopy electronics, from large old very low noise electronics with many amplifiers and entries (right), PID and PLL systems, to new compactly housed amplifier box.
Dilution refrigeration provides for a cold point stable in time. It’s operating principle is explained in many books and web pages (see e.g. here, here and here). It’s working principle was discovered by Heinz London, who noted the striking ressemblance between the osmotic pressure versus volume isotherms of a helium-4 helium-3 solution, and the pressure versus volume isotherms of a condensible gas. When introducing helium-3 from a pure into a diluted phase at very low temperatures, the helium-3 atoms come from a liquid with a Fermi temperature of some Kelvin (pure phase) into a liquid with a Fermi temperature of only a few hundreds of millikelvin (dilute phase). This requires entropy, and thus produces cooling when made adiabatically. Driving the helium-3 through the interface into the diluted phase is not trivial though. An osmotic pressure gradient is set-up between the mixing chamber (where pure helium-3 lies on top of a mixture with 6.4%of helium-3 and the rest of helium-4) and the still. The liquid connects the still to the mixing chamber, and is at a higher temperature than the mixing chamber. The still serves to pump out helium-3 into the gas phase. The evaporation disturbes the equilibrium osmotic pressure and produces a gradient which drives the helium-3 at the mixing chamber from the concentrated (pure) into the diluted phase. The helium-3 which has been pumped is inserted again. As in all cryogenic devices, it is important to use the enthalpy of the outgoing helium-3 to cool the incoming helium-3. This is achieved through a series of heat exchangers, whose construction is delicate. To reach the lowest temperatures, cables must be well thermalized at each cold point. We use different kinds of cabling and thermalization. A gas handling system (we show one of our home made gas handling systems in the figure at the left) is used to store and transfer the helium-3 into the cryostat. Part of the small presentation of dilution refrigeration delivered by I. Guillamón as undergraduate student is given here Dilution_1.
We operate several cryogenic STM systems, as shown above. We use long flexible stainless steel or PVC metal enforced vacuum tubing. Pumps ar located in another adjacent building (figure left), which additionally reduces acoustical noise in the experimental hall. In addition, all flexible stainless steel lines are covered with flexible polyethylene insulation shells, and, close to the cryostat, pumping lines are inserted into sand boxes. Acoustical noise is transmitted through the cryostat walls, and can be reduced by gluing acoustic insulation foam (such as e.g. Copopren®) to the outer part of the helium vessel. Other useful arrangements are the use of vessels without nitrogen shields, which strongly reduces vibrations from evaporating nitrogen. Building vibrations are reduced by hanging the whole cryostat assembly on long flexible ropes.
Continuous indentation-retraction cycles of the STM tip against the sample result in the creation of a nanostructure between tip and sample, the connective neck. In particular using superconducting (SC) materials, sharp nanotips result after the rupture of a lead nanobridge (S-S). The creation of these nanotips is done in-situ at low temperature.
The X-Y positioning device in our LT-STM/S unit allows a complete characterization of the nanotip (both in S-S and N-S tunneling configurations) and its use for the topographic and spectroscopic study of other superconducting materials. These capabilities are shown in the following figures, where a composite sample made of three different samples (Au, Pb and NbSe2) is studied. The tip DOS is checked on gold (N-S) and lead (S-S) regions.
The superconducting nanotips can also be used to investigate the behaviour of the superconducting state at these extreme confinement geometries. We have created Pb nanotips that present superconducting properties (detected measuring the DOS) under magnetic fields as high as 2 Tesla, more than 30 times higher than the bulk critical field for lead.
During November 2012, Paul Canfield’s visited us and we set-up a new work bench to synthesize simple single crystals of binary and ternary alloys form excess of metallic fluxes. Pure elements are put inside alumina crucibles and sealed in silica ampoules. This ampoules are placed in bigger alumina crucibles (for mechanical support), heated in a furnace and slowly cooled to let the nucleation of the single crystal occur. Once the appropriate temperature is reached, the excess liquid is spun off with a centrifugue. Furnace programs are designed with the help of the phase diagrams.
Glass is a well-known and widely used material by mankind since thousands of years ago. Nevertheless, the very nature of the glassy state and its physical properties remain an issue of vivid debate within the scientific community. Indeed, when some prominent scientists analyzed the main challenges to be addressed in the 21st century, it was stated that “The deepest and most interesting unsolved problem in solid state theory is probably the theory of the nature of the glass and the glass transition“. The main reason for this problem to remain unsolved is that two different, contradictory aspects play an important role to understand the glass-transition phenomenon: kinetics and thermodynamics. Although many experiments on any kind of glasses have been performed all along past 20th century, the answer to many key questions is still to be found. On the other hand, the anomalous behaviour of glassy or non-crystalline solids at low temperature remains an interesting and much debated issue. It was showed clearly 40 years ago, that low-temperature thermal properties of non-crystalline solids exhibit a universal behaviour, that differs significantly from that observed in fully-ordered crystals obeying Debye’s theory. In particular, specific heat of glasses depends linearly on temperature below 1 K, what was soon ascribed to the existence of tunnelling states, whereas, above 1 K, Cp still strongly deviates from the expected Debye-like behaviour due to lattice vibrations, exhibiting a characteristic broad maximum in a Cp/T^3 vs T plot. This maximum in Cp/T^3 is closely related to the so-called boson peak, observed by low-frequency vibrational spectroscopies such as Raman- or neutron-scattering, which arises from a broad maximum in the vibrational density of states normalized to the quadratic Debye one, occurring around 1 THz. Above mentioned glassy anomalies for T < 1 K were soon accounted for, at least phenomenologically, by the Tunnelling Model, that postulates the universal existence of atoms or groups of atoms in any amorphous solid, which can tunnel between two configurations of similar energy. Nonetheless, the also rich and universal glassy behaviour for T > 1 K (the maximum in Cp/T^3, a universal “plateau” in thermal conductivity within 2—20 K, the above mentioned boson peak in vibrational spectroscopies…) remained unexplained. The main research line of our group aims to deepen in our understanding of abovementioned properties universally exhibited by non-crystalline solids at low temperatures, in conjunction with the more general problem of the very nature of the glass-transition phenomenon and the glassy state itself. In particular, we have implemented and used a low-temperature, home-made calorimetric experimental system, either in cryogenic environments at liquid helium temperatures, or at temperatures above that of liquid nitrogen, in order to study and characterize substances that are liquid at room temperature and can solidify into both glassy and crystalline phases. A special attention has been paid to pure ethanol. Our main result has been to find out that the “orientational glass” (i.e. a crystal with orientational disorder) possesses qualitatively, and even quantitatively, the same “excess” of specific heat as the structural glass (an amorphous solid), so demonstrating that the lack of long-range translational order is not an essential requisite for the existence of the typical “glassy properties”. Furthermore, their very similar “glass transitions” (that is, the freezing-in from an ergodic state into a non-ergodic one) show that the very phenomenon of the glass transition can be a more general process than just the kinetic arrest experienced by a supercooled liquid when its viscosity dramatically increased when cooling, and it becomes a non-crystalline solid.