Frequency comb, deformable mirrors,, 90 Mpix detectors, all this batch of new technology going into new generation exoplanet instruments right now at ESO!

Exciting times, big challenges

By Luca Pasquini, European Southern Observatory

What incredible times we are living in — searching for an Earth around another star! Not much more than 20 years ago, I looked at the few astronomers hunting for planets, and I shook my head and thought “they are crazy”. Similar to many other occasions throughout my life, I was to be proven wrong.

Nature, of course, helped those early planet hunters quite a lot, by making Hot Jupiters. These are systems inducing radial velocity variations 10 times larger, and with much shorter periods, than the giant planets of the Solar System, which were the only ones known at that time. I believe that the influence of the discovery of the first confirmed exoplanet 51 Pegasi b by Mayor and Queloz, will be very long lasting indeed. By proving the existence of what was before then only considered likely to some of us, or an obvious hypothesis for others, their work unleashed a wealth of energy to the subject that had been unthinkable before. We are now able to analyse exoplanet atmospheres, and to search for small mass exoplanets in the habitable zones of other stars, planets that can harbour life in the way we know it.

Such a tremendous explosion of results has only been possible because fundamental technological advancements have taken place in the past 20 years, advancements that now allow us to search for solar systems and Earth analogues.

Previous expert opinions have shown us that, in order to detect the planetary radial velocity signals, a lot of effort is needed to filter the stellar ‘noise’ (activity, oscillation, variability), and in order to best do that, one needs very precise spectroscopy and long observing campaigns. This requires proper instrumentation and dedicated telescopes. Several observatories such as the European Southern Observatory (ESO) are well equipped with both.

A HARPS Laser Frequency Comb generated spectrum. The horizontal lines are spectral orders (each has two separated fibres, one fed by starlight, the other by A frequency comb combs). The colour reflects the covered spectral range, from short wavelengths (blue) to long ones (red). Each horizontal line corresponds to a diffraction order and is generated and it is composed of thousands of separated, single emission peaks, whose frequency is known with very high accuracy, and are stable over long times. LFC are the next generation of ‘rulers’ to measure precise radial velocities.
A HARPS Laser Frequency Comb generated spectrum. The horizontal lines are spectral orders (each has two separated fibres, one fed by starlight, the other by A frequency comb combs). The colour reflects the covered spectral range, from short wavelengths (blue) to long ones (red). Each horizontal line corresponds to a diffraction order and is generated and it is composed of thousands of separated, single emission peaks, whose frequency is known with very high accuracy, and are stable over long times. LFC are the next generation of ‘rulers’ to measure precise radial velocities.

Precision is the key word. The radial velocity signal induced by an Earth orbiting around a solar mass star with a period of 1 year, is less than 10 cm s-1. Just to provide some comparison, this implies that we must be able to measure the periodic shift of the stellar radial velocity in the focal plane of a typical high-resolution spectrograph for several orbital periods, over several years with a peak shift of just 2 nanometers (10-9 meters or 0.000000001 meters). In addition to requiring the utmost care in stabilising the instruments in temperature and pressure, reaching such a precision requires a very precise ruler, and for this reason several groups in the world have been engaged for almost a decade in work developing the perfect ruler. Currently the most precise rulers are based on Laser Frequency Combs (LFCs), a technique that led to the Nobel Prize for physics being awarded to T. Hänsch and J. Hall in 2005. The LFC can create a series of precisely equally spaced and stable emission lines for spectrograph calibration, whose frequency is known with high accuracy. The worldwide leader instrument in radial velocity precision is probably HARPS at ESO’s La Silla Observatory in Chile. Up to now it has been using the emission line Thorium-Argon lamps as a ruler, but it has been recently equipped with a prototype Laser Frequency Comb system. The short-term tests of this system indicate that a precision of better than 2 cm s-1 can be reached. Advances in understanding optical fibres and their technology, and getting bigger and better optical detectors have also been vital in obtaining the best performance. Optical CCDs are now very clean devices that can be accurately calibrated.

The Laser Frequency Comb hardware in the HARPS room at the ESO 3.6m telescope at La Silla, it is 'a bit' more complex system than a lamp.
The Laser Frequency Comb hardware in the HARPS room at the ESO 3.6m telescope at La Silla, it is ‘a bit’ more complex system than a lamp.

Great expectations are placed on ESPRESSO, the ‘big brother’ instrument to HARPS, that will be hosted at the ESO Very Large Telescope at Paranal before the end of the year. This instrument will boost the original HARPS precision by one order of magnitude and, in addition, will be used by any of the VLT telescopes, or by the four 8-metre Unit Telescopes together, with a 16 metre telescope equivalent diameter. Often the stars observed are relatively bright, so one could question why large telescopes are needed, but we must realise that high precision requires also a lot of stellar photons, or particles of light, hence the quest for large telescopes. A further step in precision is expected from the high-resolution spectrograph at the 39-meter E-ELT. The CODEX concept was originally conceived to measure Doppler shifts so precise such that we would be able to directly observe the expansion of the Universe and Earth-like planets around solar type stars in their habitable zones. The 25-metre Giant Magellan Telescope is planned to have a similar instrument for its first light.

espresso
The ESPRESSO CCD. With 81 million pixels, this device is the largest monolithic detector ever used in astronomy. Each side is 9 cm long.

Planets are cool compared to stars, and emit most of their light as infrared (IR) radiation, which is invisible to the human eye. Expanding observations to this spectral range is therefore essential, and the development of large IR detectors played a fundamental role: progressing in a few years from arrays of a few thousand pixels, to the most recent 16 million pixel devices, which enabled the construction of efficient high resolution infrared (IR) spectrographs. These spectrographs, such as CRIRES at the VLT, have been used to hunt for planets and to observe exoplanet atmospheres. In addition, the radial velocity signal produced by the rotation of inhomogeneities on the stellar surface can mimic the RV periodic variations induced by a planet, but , while the variations induced by the planet are the same for any measured wavelength of light, they are different in the optical and in the IR for the stellar spots that rotate around the star’s surface. That is why the newest generation of spectrographs have a great interest in the IR, and NIRPS, which will be hosted at the ESO 3.6-metre telescope at La Silla, CARMENES at the Calar Alto Observatory, Spirou at CHFT, and the high resolution spectrograph for the E-ELT, will all have an IR arm.

Even if this contribution focuses on radial velocities and spectroscopy, we should not underestimate the power of imaging. NACO at the VLT has imaged the first exoplanet around a very low mass star, similar to Proxima Centauri, and it is beyond any doubt that images, like the one shown below, have been transformational. More powerful high contrast imagers recently became available, imagers such as SPHERE at the VLT and GPI at GEMINI. These create superb images of planets and proto-planetary discs and are able to detect objects more than one million times fainter than the host star. And the next generation of instruments that will be able to exploit the tremendous potential of the ELTs is already taking shape.

The NACO image of an exoplanet (red dot) close to a very low, cool mass star. Possibly the first image of an exoplanet ever. Credits: ESO
The NACO image of an exoplanet (red dot) close to a very low, cool mass star. Possibly the first image of an exoplanet ever. Credits: ESO

Large surveys are now carried out and images of exoplanets are becoming more and more common. The enabling technology has been adaptive optics, a technique that deforms mirrors to compensate for the atmospheric turbulence and therefore recovers the cleanest image of the telescope, as if it was in space, observing with no turbulent atmosphere to look through. Deformable mirrors with more than 1000 actuators, applying corrections faster than 1000 times per second are needed, and they have been developed either of small size, for instance for SPHERE and GPI, or of large size, and these are directly replacing the secondary mirrors of the telescopes, as has happened at the Large Binocular Telescope and soon will occur at the VLT.

The Adaptive Secondary Mirror of the VLT is 1.12 m in diameter and extremely thin: the glass shell is only 2 mm thick. Being so thin, it can be easily deformed, and 1170 actuators act on the back of the glass to correct in real time the distortions induced by the Earth's atmosphere in the observed images.
The Adaptive Secondary Mirror of the VLT is 1.12 m in diameter and extremely thin: the glass shell is only 2 mm thick. Being so thin, it can be easily deformed, and 1170 actuators act on the back of the glass to correct in real time the distortions induced by the Earth’s atmosphere in the observed images.

In a nutshell, thanks to all these great instruments, exoplanet science has a bright present, and even more promising future.

 

lucaAbout the author. Luca Pasquini is an astronomer, working at ESO, Garching, and since 2013 he has been managing the Paranal Instrumentation Programme. After completing his studies in Firenze (Italy), he moved to become an ESA postdoc at MPE (Germany) in 1986, and then went on to ESO La Silla (Chile), where he was in charge of high resolution spectroscopy and of the 3.6m telescope upgrade. In 1997 he moved to ESO Garching, to work in the instrumentation group there. Before his present position, he has been instrument scientist for the FEROS, HARPS, FLAMES, MUSE, and ESPRESSO spectrographs. His scientific work and interests range from stellar activity and stellar abundances, to search for planets around giant stars and around stars in open clusters, as well as to different applications of precision spectroscopy.