[Back to the SQUIDs Home Page]
The Josephson Junction
Materials Used in the Construction of SQUIDs
Making Measurements using SQUIDs
Scanning SQUID Microscopy
Geophysical Applications of SQUIDs
High Temperature SQUIDs
Appendix I: Further Reading
Appendix II: Table Of Figures
In 1911 the Dutch physicist H. K. Onnes discovered that when he cooled mercury to below 4.15K (4.15oC above absolute zero) its electrical resistance suddenly disappeared. Since 1911 many other materials have been found that display this remarkable drop in resistance below a certain temperature. These materials have been given the name superconductors and the temperature at which the change occurs has been named the critical temperature.
In 1933 W. Meissner and R. Ochsenfeld discovered that superconductors did not allow magnetic fields to penetrate below a certain critical field. This property of superconductors gives them many more uses than one finds for materials that were simply perfect conductors.
Before 1986 the highest temperature at which any material had been known to superconduct was 23K. However in that year Alex Mueller and Greog Bednorz found that some ceramics could be made to superconduct at far higher temperatures. Within a year of their discovery the critical threshold of 77K had been crossed. Liquid nitrogen, which costs less than milk, boils at 77K. This meant that at last superconductors could be used for many more applications, because they no longer needed to be cooled by expensive and difficult to handle cryogenic liquids, such as liquid helium. However current theory does not explain these high temperature superconductors.
If two superconductors are separated by a small insulating gap, remarkable things can happen. In 1962 B.D. Josephson first noticed these effects whilst studying for a PhD. Josephson was awarded a Nobel prize in 1977 for the device that now bears his name, the Josephson junction.
If the insulator between two superconductors is made thin enough then the superconducting electrons can tunnel through the insulator. This leads to some interesting and useful effects, many of which were studied by Josephson.
These phenomena led to the invention of the Superconducting Quantum Interference Device (SQUID). There are two main types of SQUID, DC and RF (or AC). RF SQUIDs have only one Josephson junction whereas DC SQUIDs have two or more junctions. This makes DC SQUIDs more difficult and expensive to produce, but DC SQUIDs are much more sensitive.
There are many types of Josephson junction, the most common being the tunnel junction (which has an oxide barrier), the normal semiconductor junction (which has a non-superconducting material barrier) and the Dayem bridge junction (which is a constriction-type junction). Other types of junction were more popular in the early days of SQUID construction, but are not used so much now due to their irreproducibility and unreliability.
Josephson junctions are made using three techniques: evaporation through a metal mask, etching with chemicals or ions and photo- or electron-lithography (similar to the process used for integrated circuits).
Evaporation is used when the features of the circuit are large, i.e. when the mask can be made mechanically. Etching is a more complex process, and the size of the features depends very much on the actual reaction. Lithography can be used to make features as small as a millionth of metre.
Most SQUIDs are fabricated from lead or pure niobium. The lead is usually in the form of an alloy with 10% gold or indium, as pure lead is found to be unstable when its temperature is repeatedly changed. The base electrode of the SQUID is made of a very thin niobium layer, formed by deposition and the tunnel barrier is oxidised onto this niobium surface. The top electrode is a layer of lead alloy deposited on top of the other two, forming a sandwich arrangement.
This two-metal structure has both advantages and disadvantages over all lead-alloy structures. Niobium is very hard and has high tensile strength, and its pentoxide is very stable. However, it is difficult to fabricate pure niobium alloy structures with niobium oxide barriers that have good tunnelling characteristics. In particular, current often leaks at the junction, which can be countered by applying a thin copper or gold layer on the top of the oxide. Recent advances utilising ’artificial’ tunnel barriers such as aluminium and magnesium oxides or amorphous silicon between all niobium or niobium nitride electrodes have produced devices, which are more stable and have less leakage current. A particular advantage of the niobium nitride electrodes is that the operating temperature is higher.
Much effort is currently going into researching superconducting ceramic compounds, such as yttrium-barium-copper-oxide, bismuth-strontium-calcium-copper oxide. Due to their unfavourable mechanical chemical and crystalline properties, it has been difficult to develop high quality films and wires. Their importance lies in their very high critical temperatures, from 90K to over 125K.
There are a small number of companies that manufacture and sell SQUIDs while several more produce medical equipment, which utilise SQUIDs. Prices are relatively high, Oxford Instruments a company based in the UK, prices it’s low temperature DC SQUID at £2000 and its High Temperature SQUID at £4500. These prices do not include the control electronics, which can cost over £3000.
If two Josephson junctions are connected in parallel then electrons, which tunnel through the junctions, interfere with one another. This caused by a phase difference between the Quantum Mechanical wavefunctions of the electrons, which is dependent upon the strength of the magnetic field through the loop. The resultant supercurrent varies with any externally applied magnetic field. Hence, the two junctions in parallel can detect variations in a magnetic field very sensitively. This device is known as a SQUID.
DC SQUIDs resist any change in magnetic field. Thus, if the magnetic field is changed there will be a modulation of the supercurrent through the loop, which can be measured.
Figure 1 illustrates the typical DC SQUID circuitry. The circuitry modulates the magnetic flux through the loop. This allows the amplification circuitry to detect fluxes that are fractions of the flux quantum. The loop is subjected to a constant biasing current, which leads to the average output voltage being modulated by the flux through the loop.
A Radio Frequency (RF) SQUID consists of a single Josephson junction mounted on a superconducting ring. A radio frequency oscillating current is then passed through an external circuit. The magnetic flux through the ring can be calculated by measuring the voltage in the external circuit, which is dependent on the interaction between the circuit and the superconducting ring.
SQUIDs are often used to measure very small field magnitudes, so background magnetic fields are a real problem. Often measurements are taken in a shielded room, however this is expensive and cumbersome, so often gradiometers are used. These measure the gradient of the applied field. Assuming that the other sources of magnetic fields are much further away than the source you wish to measure, one can easily differentiate between the two. Other methods that have been used include damping coils and measuring the ambient magnetic field.
The SQUID, used in conjunction with suitable external circuits, is currently the most sensitive device for measuring magnetic fields. SQUIDs can ‘transform’ magnetic flux and hence measure other electrical parameters.
Johnson noise is the background magnetic field created by thermal motions in a given environment. Great care must be taken to design experiments or apparatus using SQUIDs as the Johnson noise can nullify their extreme sensitivity.
The ‘DC current comparator’ is a device that can compare two supercurrents, achieving results accurate to 1 part in a thousand billion. External circuits have also been designed to deal with alternating currents.
The design of flux transformer circuits has been widely used to measure the magnetic susceptibility of a sample. Much work has been carried out in the fields of static nuclear susceptibility and Nuclear Magnetic Resonance (NMR).
Some processes in animals produce very small magnetic fields (typically sized between a billionth of a Tesla and a thousand billionth of a Tesla - a typical fridge magnet is a tenth of a Tesla). The only type of detector sensitive enough to measure such a field is a SQUID. In the human body, studies have taken place measuring the fields arising from the susceptibility of tissue to applied magnetic fields, ionic healing currents and currents associated with neural or muscular activity.
By far the biggest area of research is Magnetoencephalography (MEG) the imaging of the human brain from magnetic fields. This involves measuring the magnetic field produced by the currents due to neural activity. Unlike other methods which image the structure of the brain MEG images can be acquired every millisecond, allowing real-time imaging systems.
In most available systems, arrays of gradiometer DC SQUID detectors are contained within a helmet surrounded by a liquid helium reservoir for cooling. The Neuromag Ltd. 122 is such a system and is shown in figure 2, with the sensor array in figure 3 and a schematic of the detector in figure 4.
Figure 2: Neuromag Ltd. 122 MEG system
Figure 3: Neuromag Ltd. 122 sensor array
Figure 4: Neuromag Ltd. 122 probe
SQUIDs have also been used to measure the magnetic fields from a heartbeat. This is known as a magnetocardiogram. In such systems magnetometers are normally used, measuring the magnitude of the field present. Some systems with gradiometers have been used enabling measurements in an unshielded environment.
The scanning SQUID microscope is an instrument for measuring local magnetic fields. It uses a SQUID immersed in liquid helium as the probe. Being very sensitive to changes in magnetic field the SQUID can produce unprecedented pictures of magnetic fields at the surface of many samples and therefore has many practical applications.
A few of the images it can produce are shown below. Figures 5 and 6 show (in false colour) the magnetic field above a relatively new superconductor (yttrium barium copper oxide). It shows four thin rings of the substance and the strength of the magnetic field due to them.
Figure 5: Yttrium barium copper oxide
Figure 6: Yttrium barium copper oxide
Figure 7 shows the magnetic field present about 15 microns above the surface of a commercial floppy disc. Figure 8 is a scanning SQUID microscope with the SQUID housed at the bottom.
Figure 7: A floppy disk as seen by a scanning SQUID microscope
Figure 8: A scanning SQUID microscope
Geophysicists need to be able to measure weak magnetic fields for studies into such things as the movement of the magnetic poles over time and measurement of the thickness of the Earth’s crust. Until recently the use of SQUIDs in the field was not very common because their extreme sensitivity made it impossible to use them in an unshielded environment and they needed to be cooled by cryogenic liquids with all the attendant problems. Also, there are already a number of suitable devices in existence that have none of these problems (such as search coils and optical pumping magnetometers).
Many of the above problems have been overcome. For example newly invented two walled vacuum flasks can contain liquid helium for a couple of weeks at room temperature, and ways of reducing the effects of magnetic noise have been found (see Biomagnetism section).
The use of SQUIDs in oil prospecting, earthquake prediction and geothermal energy surveying is becoming more widespread as superconductor technology develops.
The development of SQUIDs constructed from superconductors with high critical temperatures has brought about much simplification and reduced cost while increasing the mobility and flexibility of SQUIDs. However modern cooling systems are beginning to give this to Low Temperature Superconductor (LTS) applications as well.
The traditional Josephson Junctions of the type used in LTS SQUIDs could not be simply converted into a High Temperature Superconductor (HTS) equivalent. In addition new HTS junction types are difficult to manufacture to predictable and reproducible standards. Many of these problems have been overcome and commercial HTS SQUIDs were realised some 5 to 6 years after the initial discovery of high temperature superconductors and they are now reaching the performance of LTS SQUIDs.
With modern research enabling uses of SQUIDs outside well controlled shielded environments many additional applications are currently being developed. A promising application involves the non-destructive testing of materials. Such an application has much commercial interest due to increased demands for quality and safety. In many fields, such as aircraft components, subsurface flaws can have disastrous consequences. In such a situation it is often advantageous to take the testing equipment where it is needed and HTS SQUIDs have much greater mobility than their low temperature cousins.
Our web site, this contains full versions of our technical report (see above), popular account (this document) and our progress report (what we said we would do before we started) as well as links to other related web sites.
University Physics: H. Benson
Chapter 42 gives a good, but basic, general description of superconductivity, with a short section on SQUIDs.
SQUIDS, the Josephson Effect and Superconducting Electronics: J. C. Gallop
A more detailed look at the subject. However this book is primarily aimed at physicists.
Foundations of Applied Superconductivity: T. P. Orlando and K. A. Delh
An excellent introduction to applied superconductivity suited to advanced university students as well as scientists and engineers in industry.
Lectures in Physics (Vol. III): R. Feynman
A classic text on many matters in physics. Chapter 21 covers superconductivity and the Josephson junction.
Physics Education 31 (1996), Imaging the Brain: S. J. Swithenby
A magazine article describing MEG imaging of the brain.
Figure 1: Drawn by R. Dawe
Figure 2: The Neuromag Ltd. 122 MEG system.: Neuromag Web site, http://www.neuromag.com
Figure 3: The Neuromag Ltd. 122 probe: Neuromag Web site, http://www.neuromag.com
Figure 4: The Neuromag Ltd. 122 sensor array: Imaging the working brain, S. J. Swithenby, Physics Education 31 No.2, 108-112 (1996).
Figure 5: Taken from an IBM research article on Scanning SQUID Microscopes (Click to view the original image).
Figure 6: Mislaid source - it may be from the same source as Figures 5 and 7, however.
Figure 7: Taken from an IBM research article on Scanning SQUID Microscopes (Click to view the original image).
Figure 8: Taken from an introductory article on Scanning SQUID Microscopes (Click to view the original image).
[Back to the SQUIDs Home Page]
This document was saved as HTML from a Word 97 document and then labouriously converted from its nasty output. This document was last updated on Wednesday 28th October 1998.