All part of a five year £270m government-funded programme, in essence it involves the development of accelerometers, gyros, gravity gradiometers, and clocks based not on conventional electro-mechanics, but on measuring the behaviour of individual atoms or small clouds of atoms using what is becoming known as ‘quantum 2.0’ technology.
“Quantum 1.0 is things like lasers and infra-red night-vision. In some ways it is still a victim of Newtonian thinking,” said Andrew Middleton of DSTL. “Quantum 2.0 views atoms as resonant system. A quantum accelerometer or gyro could be accurate to 10 digits, orders of magnitude better than current thinking.”
In practice, the errors of a quantum 1.0 device depend on all sorts of effects, while the errors of a quantum 2.0 device are limited only by Heisenberg’s uncertainty principle.
NPL is involved because it is one of the few organisations in the world which is already using advanced quantum techniques – in this case in its atomic clocks and other programmes to relate all SI units to fundamental physics.
An example of the usefulness of atoms is in accelerometers.
As far as anyone can tell, every atom of any particular isotope of any particular element has the same mass – not just nearly the same, but exactly the same down to any number of zeros.
This means that, once big science has determined the mass of that atom, anyone making an accelerometer using a known number of those atoms as a proof mass will know exactly how heavy the mass is to the same number of zeros without having to calibrate it.
In general, any quantum sensor needs a cold atom trap where a number of atoms, sometimes just the right number, are kept close to absolute zero – mK or µK – in a hard vacuum.
Then there will be a bunch of lasers to hold the atoms in place, keep them cold, and measure them. Sometimes other techniques replace some of the lasers.
A quantum inertial navigation package would have the same components as a normal inertial navigation package – three accelerometers, three gyros, and a clock, but all of them based on measuring the behaviour of super-cold atoms.
Accuracy in these systems is related to the frequency of the measuring system.
In the case of a caesium atomic clock, the current ‘gold standard’, this 9.2GHz microwaves. If lasers can be used, the frequency rises to around 100THz and accuracy is up to 100x better. The inventors of the optical ‘frequency comb’ – a divider capable of counting THz accurately, were awarded a Nobel Prize in 2005.
At NPL, Professor Kai Bongs described a shoe box-sized ‘optical’ atomic clock his team has created at the University of Birmingham.
Noise in quantum 2.0 systems is related to 1/√n, where n is the number of atoms. If these atoms can be ‘entangled’ – Einstein’s ‘spooky action at a distance’ – noise is then proportional to 1/n, so only, 10 entangle atoms have to be manipulated precisely rather than 100 independent atoms. That said, this programme is already hugely difficult and no one is going to be adding entanglement any time soon.
Other tricks can be played
Aluminium is ideal ion for atomic clocks – or it would be if it didn’t require a deep UV laser if the ions are to be both cooled and their oscillations counted.
Now the US lab NIST has made a ‘quantum logic clock’ which separates clock function from cooling.
In it, beryllium ions (‘logic’ ions) are coupled to aluminium clock ions. Lasers with more reasonable wavelengths cool the Be+ ions, which couple energy from the Al+ ions, cooling them to the point that output frequency is stable to 3×10^-17.
The gravity gradiometer may be an obscure device, but it has a remarkable number of applications.
In essence, it measures the pull of gravity on two separate masses, one above the other. The difference in the two gravity measurements indicates how steeply gravity is changing at that locality.
Proposals for the quantum version involve suspending two atoms one above the other using lasers, and using a single laser passing both of them to measure distance changes between them.
Predicted accuracy is so high that voids in rock such as old mine workings could be detected and mapped in detail and, using side-by-side atoms, an image of what is behind a wall could be built up – with no known way of shielding it – leading to an instrument dubbed the ‘quantum telescope’, where multiple gradiometers form the pixels of an imaging array.
However, the current algorithms used to creating images from gradiometer readings are only suitable for 10x the capability of existing electromechanical devices, so new maths will be needed to go along with quantum versions.
Quantum navigators will also benefit from gradiometers, as knowledge of local gravity gradient removes an instability term from the maths of inertial navigation.
Gathered at NPL were academics and companies describing and demonstrating the components they are developing for quantum 2.0 navigators, telescopes, and atomic clocks:
Professor Tim Spiller at the University of Leeds is investigating quantum-enhanced gyroscopes for navigation – first simulations indicate they could approach the Heisenberg’s uncertainty limits.
In a quantum gyroscope, ultra cold atoms are held in an asymmetric trap of magnetic and electric fields. When the trap is rotated it imparts a superposition on the cold atoms where they now exist in two states – rotating and static – simultaneously. This non-classical quantum set up is highly sensitive to any additional rotation.
Spiller has mapped out the whole measurement process from the beginning to the end to calculate realistic performance estimates. Long term, the aim will be to build accurate gyroscopes on semiconductor chips.
The University of Glasgow and Heriot-Watt University are developing miniature frequency combs, locked to atomic clocks, to bridge > 100THz optical signals to electrically-countable frequencies (100s of GHz).
They combs, which can be as big as a bench, produce a series of well-defined optical frequencies with a very regular spacing, leading to a comb-like appearance.
Once the frequency of one comb tooth is locked to to the reference, the comb of frequencies is locked.
In this research project, the scientists are trying to biuld a chip-scale precision comb.
Dr Alastair Sinclair and colleagues at NPL are working on a chip-scale cold ion trap for atomic clocks.
The trap will use an oscillating electric field to trap ions, either individually or several in a row, then lasers will reduce their temperature and momentum – by prompting them to emit photons and therefore loose energy. Once almost stationary, Doppler shift is all-but eliminated and the abroption and emission spectra of the ions becomes very pure and suited to atomic clock use.
In prototype form, the trap is a slot right through a wafer, with conductors runing along the edges on both sides, forming an X-shaped field to restrain the ions in a row.
In addition to smaller atomic clocks, the ion microchip could potentially be used in quantum information processing, said NPL.
The photo (above) shows ultraviolet cooling laser beams from three directions focussed through the centre of an ion trap to cool single ions or strings of tens of ions confined in an electromagnetic field. Cooled ions are separated by a few microns and each localised to within a few tens of nanometres. The ultra-high vacuum package has straightforward optical and electronic access from air side.
Hollow core fibre atomic clock
NPL has got together with with the Opto-Electronics Research Centre, of the University of Southampton to develop another form of atomic clock – a quantum 1.0 atomic clock at whose centre is a hollow optical fibre. It has a 350µm outside diameter and a 80µm core consisting of multiple parallel honeycomb-like holes running its length – and constituting a photonic crystal. Within the holes, some (non-cooled) caesium vapour is trapped.
Caesium atomic clocks are generally probed by microwaves at 9.2GHz, but this is probed by a laser running along its length modulated at microwave frequencies.
Compared with the US-developed ‘CSAC’ chip-scale atomic clock, which uses a caesium-filled miniature glass cell, the fibre provides increased length and therefore interaction with more atoms, which should give greater signal contrast, as well as being low-power.
An early stage demonstrator is under investigation. The next stage is to seal the fibre and create a small package, eventually note book-sized.
Imperial College is developing a portable quantum accelerometer for navigation, which is expected to be many orders of magnitude more stable than the best mechanical devices of today.
Based around ultra-cold atoms, the accelerometer uses laser and an atom interferometer to track how the atom moves in free-fall relative to the instrument itself.
“Atom interferometers are far more stable because the atoms, and the laser light they use, are intrinsically stable and calibrated. However experimental set ups the size of rooms are normally required to achieve this,” said the university.
The team is working on a one axis set up before moving to a thre axis version. A company in Berlin will provide a demonstrator package.
Graphene atom chips
University of Nottingham researchers are investigating the potential of graphene to improve the performance and power consumption of atom chips for use in gyroscopes and electromagnetic field sensors.
Once again, is will be restraining atoms only a few billionths of a degree above absolute zero, where they lose their particle identity and become a quantum wave up to 1mm long – and highly sensitive to a range of forces including acceleration, gravity and rotation.
Current atomic chips based on metal wires are inefficient, requiring significant power (~1W) to keep the ultra cold atoms in place. “Even with such a power supply, the system is far from perfect and researchers believe optimum performance can only be reached by finding a way to hold the atoms stably 100 times closer to the chip where they can be controlled on a finer spatial scale,” said Nottingham.
Graphene may provide the solution. Simulations suggest that the material properties of graphene will allow a cloud of atoms cooled to µK, to float just above room temperature graphene, while gaining little or no heat – equivalent to snowball existing in environment 5,000x hotter, said the University.
This virtual insulation could reduce energy consumption by a million times, and now a team combining atomic physicists, condensed matter physicists, crystal growers and quantum electronics experts are trying to make a graphene-based trap.
The University of Southampton is investigating magneto optical trap that can hold neutral (non-ion) laser-cooled atoms within an ultra-high vacuum.
“Whilst room temperature vapour atomic clocks lose one second of time every 10^11s (3000 years), current laboratory-based cold clocks can take this precision down by seven further orders of magnitude, a level which would see a clock maintain accuracy to within more than half a second throughout the entire history of the universe,” said Southampton.
These are about a metre across, which the University hopes to shrink using chip-making techniques and new materials – one of these is aluminosilicate glass, which is vastly less leaky than other glasses. The aim is a postage stamp-sized trap.
Portable optical clock
Usually lab-sized, the Birmingham clock is already down to 1m across, and the aim is to reduce this to the size of a suitcase, while retaining as much as possible of the 100x accuracy improvement possible compared to microwave atomic clocks.
According to Einstein’s theory of relativity, the stronger the gravitational field, the slower time should pass. Optical clocks are now accurate enough to detect these differences. Changing their height and measuring time differences should allow gravitational potential to be measured. Such a clock would prediction of things that depend on the Earth’s geoid – the shape of the surface of the seas if they were affected by gravitational force alone – rather than height from sea level, such as paths of missiles and levels of water in a dam. They could also help produce a height reference system between different countries – which are current based on coastal sea levels, which can differ by tens of centimetres.
The University is also applying atom interferometers ot the gravity telescope challenge. “Oil and mineral reservoirs go along with a density anomaly which for detection needs a combination of an acceleration sensor in addition to a good geoid reference,” said the University.
Sharing time improves accuracy
Plextek Consulting is investigating how networked assets – for example a shoal of submarnines – can share inertial navigation information to improve estimates of position and time.
Autonomous submarines use gyroscopes and Doppler velocity logs to estimate their position relative to a known starting point, and regularly need re-set accumulated errors by surfacing for GPS fixes.
According to Plextek, position and time estimates can be significantly improved by combining position and time estimates from a number of nearby submarines using an appropriate Kalman filter – a principle similar to that used by the Bureau of Weights and Measures in Paris to combine the outputs of trusted atomic clocks around the world.
Given this filtering, less frequent GPS fixes would be needed, and only one submarine would have to do it.
Diamond MEMs resonator
Not an atomic technique, but an attempt to move crystal oscillators up to navigation grade.
Astrium is developing a 1GHz diamond MEMs resonator, offering increased performance from a device 20um across, compared to the 1mm 32kHz quartz oscillators using in watches and laptops.
“This is of particular interest to the satellite industry, where savings on mass and size can equate to significant launch cost reduction,” said the firm.
Diamond is far stiffer than quartz – hence the size reduction.
As a research tool, Glasgow-based M Squared was at NPL to promote its SolsTiS phone-book-sized tunable, narrow-line-width continuous wave titanium sapphire laser, claimed to have the lowest noise in the industry. The output can be locked externally to a transition, reference cavity or wavemeter. Integrated accessories such as beam pick-off and fiber coupling are available. It includes instrument control by ethernet (ICE).
Applications include: atom cooling, optical tweezing, holography, high-density optical data storage, and metrology.