Features RSS

SDO AIA 01/15/13

The Sun often exhibits outbursts, launching material from its surface in powerful releases of energy. Recent analysis of such an outburst — captured on video by several Sun-monitoring spacecraft — may help us understand the mechanisms that launch these eruptions.

Many Outbursts

Solar jets are elongated, transient structures that are thought to regularly release magnetic energy from the Sun, contributing to coronal heating and solar wind acceleration. Coronal mass ejections (CMEs), on the other hand, are enormous blob-like explosions, violently ejecting energy and mass from the Sun at incredible speeds.

But could these two types of events actually be related? According to a team of scientists at the University of Science and Technology of China, they may well be. The team, led by Jiajia Liu, has analyzed observations of a coronal jet that they believe prompted the launch of a powerful CME.

Observing an Explosion

CME gif

Gif of a movie of the CME, taken by the Solar Dynamics Observatory’s Atmospheric Imaging Assembly at a wavelength of 304Å. The original movie can be found in the article. [Liu et al.]

An army of spacecraft was on hand to witness the event on 15 Jan 2013 — including the Solar Dynamics Observatory (SDO), the Solar and Heliospheric Observatory (SOHO), and the Solar Terrestrial Relations Observatory (STEREO). The instruments on board these observatories captured the drama on the northern limb of the Sun as, at 19:32 UT, a coronal jet formed. Just eight minutes later, a powerful CME was released from the same active region.

The fact that the jet and CME occurred in the same place at roughly the same time suggests they’re related. But did the initial motions of the CME blob trigger the jet? Or did the jet trigger the CME?

Tying It All Together

In a recently published study, Liu and collaborators analyzed the multi-wavelength observations of this event to find the heights and positions of the jet and CME. From this analysis, they determined that the coronal jet triggered the release of material to form the CME, which then erupted into space — with the jet at its core — at speeds of over 1000 km/s.

Based on observed clues of the magnetic field configurations, the team has put together a theory for how this event unfolded. They believe that sudden magnetic reconnection in an active region accelerated plasma to form a large-scale coronal jet. This burst of energy also provided a push on a blob of gas, threaded with magnetic field lines, that lay above the jet. The blob then rose, and when the field lines broke, it was released as a CME with the jet at its core.

Citation

Jiajia Liu et al 2015 ApJ 813 115. doi:10.1088/0004-637X/813/2/115

Seth region

High-resolution imagery of comet 67P Churyumov–Gerasimenko has revealed that its surface is covered in active pits — some measuring hundreds of meters both wide and deep! But what processes caused these pits to form?

Pitted Landscape

ESA’s Rosetta mission arrived at comet 67P in August 2014. As the comet continued its journey around the Sun, Rosetta extensively documented 67P’s surface through high-resolution images taken with the on-board instrument NavCam. These images have revealed that active, circular depressions are a common feature on the comet’s surface.

In an attempt to determine how these pits formed, an international team of scientists led by Olivier Mousis (Laboratory of Astrophysics of Marseille) has run a series of simulations of a region of the comet — the “Seth” region — that contains a 200-meter-deep pit. These simulations include the effects of various phase transitions, heat transfer through the matrix of ices and dust, and gas diffusion throughout the porous material.

Escaping Volatiles

pitted regions

Additional examples of pitted areas on 67P’s northern-hemisphere surface include the Ash region and the Ma’at region (both imaged September 2014 by NavCam) [Mousis et al. 2015]

Previous studies have already eliminated two potential formation mechanisms for the pits: impacts (the sizes of the pits weren’t right) and erosion due to sunlight (the pits don’t have the right shape). Mousis and collaborators assume that the pits are instead caused by the depletion of volatile materials — chemical compounds with low boiling points — either via explosive outbursts at the comet’s surface, or via sinkholes opening from below the surface. But what process causes the volatiles to deplete when the comet heats?

The authors’ simulations demonstrate that volatiles trapped beneath the comet’s surface — either in icy structures called “clathrates” or within amorphous ice — can be suddenly released as the comet warms up. The team shows that the release of volatiles from these two structures can create 200-meter-deep pits within ~800 years and ~2,000 years, respectively. Since comet 67P has been around the inner solar system for about 7,000 years, both of these processes are viable explanations for the pits.

The simulations also show that direct sublimation of crystalline ices of water, carbon monoxide, and carbon dioxide can cause deep pits — but only in the absence of a surface layer of dust, known as a “dust mantle,” in that region of the comet. Direct sublimation could be a viable explanation for the pits, then, if dust grains in the area are so small that they are carried away with the released gas, rather than falling back to form a layer on the comet’s surface.

Regardless of the formation mechanism for these pits, the authors conclude that their very existence implies that the physical and chemical properties across the surface and subsurface of the comet cannot be uniform. Further observations from Rosetta will continue to help us understand comet 67P.

Citation

O. Mousis et al 2015 ApJ 814 L5. doi:10.1088/2041-8205/814/1/L5

Procyon

The binary system Procyon, located a mere 11 light-years away, consists of a bright, subgiant star and a faint white dwarf — presenting a distinct challenge for astronomers to observe. But careful analysis of two decades of precise measurements with the Hubble Space Telescope has now finally revealed some of its secrets.

Challenging Observations

Perturbations were detected as early as 1844 in the orbit of Procyon, originally thought to be a single star. Astronomers of the time suspected that this wobbling was due to the pull of a companion orbiting Procyon, but it wasn’t until five decades later that the companion was first detected visually.

Why? Because the subgiant Procyon A is the 8th brightest star in the sky. Its companion, on the other hand, is a white dwarf that’s fainter (in visual wavelengths) by a factor of nearly 16,000! And the two stars are separated by an angular distance of less than 5”.

Due to the difficulty observing the system, the measurements of its motion — and resulting estimates of the masses of the two stars — have been a subject of debate for the better part of the last century.

Led by Howard Bond (Pennsylvania State University and the Space Telescope Science Institute), a team of astronomers has now analyzed two decades of Hubble observations of the system, combined with historical, ground-based observations dating back to the 19th century. Bond and collaborators used these data to precisely measure the orbital elements of Procyon and obtain dynamical masses of the two stars.

Surprising Mixing

Procyon orbit

Relative orbit of Procyon B around Procyon A. The red curve is the authors’ fit to the orbit, and the open blue circles are positions predicted by the orbital elements found. The black dots are the HST observations of Procyon B. The open green and turquoise circles are the (significantly less precise!) historical, ground-based observations. [Bond et al. 2015]

The team reports that this system orbits once every 40.8 years. They find masses for the two stars of 1.48 solar masses for Procyon A, and 0.59 solar masses for Procyon B. Both of these masses produce very satisfying agreement with theoretical predictions of the stars’ masses based on their temperatures, luminosities, and asteroseismology of Procyon A.

But the measurements of this system also have some interesting implications. One example arises when the authors apply the standard model for the evolution of a subgiant star to Procyon A. Using the new parameters, they find that Procyon A may have an unusually high amount of mixing of material beyond its convective core. If this is confirmed, it could mean that our understanding of how stars like Procyon A evolve may need to be updated.

Thus the precise measurements of the Procyon system have allowed the authors to not only pin down the parameters of the system, but also access new information about stellar physics near the main sequence. We’ve come a long way in the 170 years since Procyon was first observed to wobble!

Citation

Howard E. Bond et al 2015 ApJ 813 106. doi:10.1088/0004-637X/813/2/106

active galactic nucleus

Who needs humans? Robotic observations made by telescopes in the Las Cumbres Observatory Global Telescope network (LCOGT) have tracked variability in the active galaxy Arp 151 over 200 days. These observations have proven to be enough information to estimate the mass of the black hole at the galaxy’s center.

Mapping Echoes

Measuring the masses of supermassive black holes is notoriously difficult. Except in the few cases where we’re able to resolve actual objects orbiting around the supermassive black hole (for instance, in the case of the black hole at the center of the Milky Way), our estimates of black-hole mass must come from indirect measurements.

One clever approach is called “reverberation mapping.” In an active galactic nucleus (AGN), continuum emission from the black hole’s accretion disk photoionizes gas clouds in the nearby broad-line region, causing the clouds to emit light. In reverberation mapping, we track the time lag between variability in the disk’s continuum emission and the clouds’ broad-line emission, obtaining a distance scale. Combining this information with a velocity (provided by the broad-line width) allows us to infer the enclosed mass — in this case, that of the black hole.

So what’s the catch? Getting this information requires a lot of man-hours and telescope-hours, because AGN need to be observed over long periods of time to see the variability and the lags needed to make these inferences. This is where LCOGT comes in.

Robotic Network

Arp 151 light curves

Arp 151 light curves. The top panel shows the continuum emission from the disk; the remaining panels show various emission lines from the broad-line-region clouds. The variability of the line emission lags slightly behind that of the continuum emission. [Valenti et al. 2015]

LCOGT is a completely robotic telescope network. Everything from the scheduling to the telescope alignment is done without human involvement. Because of this feature, the LCOGT is an ideal facility for conducting time-intensive observations of AGN.

A team of scientists led by Stefano Valenti (LCOGT, UC Santa Barbara) has published the first results from the AGN Key Project, a project which uses the LCOGT network to conduct several studies of AGN — including reverberation mapping of both local and high-redshift objects. In these first results, the team reports the outcome of a 200-day observing campaign of the galaxy Arp 151, which has a highly variable active nucleus.

The LCOGT observations successfully show a measurable lag between the continuum emission and the broad emission lines for Arp 151. They are also enough-resolved that a velocity can be measured from the broad emission lines, allowing the team to calculate the mass of the black hole enclosed. Valenti and collaborators announce a mass of 6.2 million solar masses — consistent with previously measured masses for this system.

The success of this test demonstrates the viability of this approach, as well as the powerful capabilities of robotic telescope networks for long-term AGN time domain campaigns.

Citation

S. Valenti et al 2015 ApJ 813 L36. doi:10.1088/2041-8205/813/2/L36

Milky Way bulge

In our efforts to map our galaxy’s structure, one region has remained very difficult to probe: the galactic center. A new survey, however, uses infrared light to peer through the gas and dust in the galactic plane, searching for variable stars in the bulge of the galaxy. This study has discovered a population of very young stars in a thin disk in the galactic center, providing clues to the star formation history of the Milky Way over the last 100 million years.

Obscured Center

The center of the Milky Way is dominated by a region known as the galactic bulge. Efforts to better understand this region — in particular, its star formation history — have been hindered by the stars, gas, and dust of the galactic disk, which prevent us from viewing the galactic bulge at low latitudes in visible light.

The positions of the 35 classical Cepheids discovered in VVV data are projected onto an image of the galactic plane. The survey area is outlined by the blue lines. The galactic bar is marked with a red curve, and the bottom panel shows the position of the Cepheids overlaid on the VVV bulge extinction map. Click for a better look! [Dékány et al. 2015]

The positions of the 35 classical Cepheids discovered in VVV data, projected onto an image of the galactic plane. Click for a better look! The survey area is bounded by the blue lines, and the galactic bar is marked with a red curve. The bottom panel shows the position of the Cepheids overlaid on the VVV bulge extinction map. [Dékány et al. 2015]

Infrared light, however, can be used to probe deeper through the dust than visible-light searches. A new survey called VISTA Variables in the Via Lactea (VVV) uses the VISTA telescope in Chile to search, in infrared, for variable stars in the inner part of the galaxy. The VVV survey area spans the Milky Way bulge and an adjacent section of the mid-plane where star formation activity is high.

Led by István Dékány, a researcher at the Millennium Institute of Astrophysics and the Pontifical Catholic University of Chile, a team has now used VVV data to specifically identify classical Cepheid variable stars in the bulge. Why? Cepheids are pulsating stars with a very useful relation between their periods and luminosities that allows them to be used as distance indicators. Moreover, classical Cepheids are indicators of young stellar populations — which can provide information about the star formation history of the region.

New Population

Dékány and collaborators found a total of 35 objects that they believe to be young classical Cepheids. These stars, which were found to have a spread in ages below 100 Myr, span the center of the galaxy, bridging the innermost nuclear region and the larger bulge extent. Despite their distribution across the bulge, however, they make up a very thin disk, all lying close to the mid-plane of the galaxy.

This newly discovered component of the inner galaxy demonstrates that stars have been continuously forming in or near the central region of the galaxy over the last hundred million years. Future investigations of these stars will help determine if the star formation actually occurred in the central bulge, or if the stars originated further out and migrated inwards.

Citation

I. Dékány et al 2015 ApJ 812 L29. doi:10.1088/2041-8205/812/2/L29

Mrk 739

Dual active galactic nuclei (AGN), an intermediary product of galaxy mergers, can give us a better understanding of what happens when two galaxies collide. But because the angular separation of the two galactic nuclei is so small at this stage, identifying these systems is very difficult. In a recent study, a team of authors proposes a new technique for confirming dual AGN candidates.

Signatures in Spectra

Total-intensity VLA image for J1023+3243. This system is confirmed as a dual AGN; the two compact radio cores are separately identifiable here. [Müller-Sánchez et al. 2015]

Total-intensity VLA image for J1023+3243. This system is confirmed as a dual AGN; the two compact radio cores are separately identifiable here. [Müller-Sánchez et al. 2015]

One approach commonly used to identify dual AGN candidates is to look for signatures in the spectra of these galaxies. Light is emitted by ionized gas in the narrow-line region (NLR), the region that extends from a few hundreds of parsecs to ~30kpc from the nuclei. The spatially-averaged spectrum of this region for dual AGN, however, appears double-peaked due to the motion of the two nuclei rotating around each other.

But there’s a problem with using this technique to identify dual AGN: other processes also produce double-peaked narrow-line emission, mimicking the behavior of dual AGN. These processes include the rotation of ionized gas in the galactic disk, and the motion of radio jets emitted from the AGN.

A team of scientists led by Francisco Müller-Sánchez (University of Colorado Boulder) have proposed that the use of a combination of high-resolution radio observations and spatially-resolved spectroscopy could be used to discern between these possible cases.

Dual AGN or Moving Gas?

To test this method, the group examined a sample of 18 active galactic nuclei from the Sloan Digital Sky Survey. These AGN had previously been identified as candidate dual AGN with double-peaked narrow emission lines. The team obtained both optical long-slit spectroscopy and high-resolution Very Large Array observations of these AGN. They then combined this information to identify the cause of the double-peaked lines in each case.

radio jet

Total-intensity VLA image for J0009–0036. This system contains a two-sided radio jet, causing the extended radio emission seen here. [Müller-Sánchez et al. 2015]

Müller-Sánchez and collaborators found the following:

Roughly 15% are confirmed to be dual AGN. In these cases, distinctly separated radio cores are visible in the Very Large Array data, but their spectra are similar to those of single AGN.

Roughly 75% have double-peaked lines due to gas kinematics, instead. These kinematics include jets, rotating NLR regions, and wind-driven outflows. The jets are identifiable by their extended radio emission and steeper spectra, whereas the rotating NLR regions and wind-driven outflows are identifiable by their lack of additional radio cores or extended emission, and the morphology of their spectra.

Only two cases of the 18 were ambiguous and couldn’t be identified. The authors conclude that their method of confirming dual AGN is therefore a powerful means of identifying dual AGN that have very small angular separations.

Citation

F. Müller-Sánchez et al 2015 ApJ 813 103. doi:10.1088/0004-637X/813/2/103

T Tauri star

T Tauri stars are a young class of variable stars. The T Tauri star PTFO 8-8695, which lies around 1,100 light-years away in Orion, has been suspected of harboring a close-in giant planet. A recent study, however, casts doubt on this theory.

Fading Star

Finding a close-in giant planet around a very young (i.e., less than a few million years old) star would be a major discovery. Such a system could tell us about when planets form, how they behave as newly-born planets, and the mechanism that causes planets to migrate inwards to create hot Jupiters. Thus far, the only candidate system of a very young star with a close-in giant planet is PTFO 8-8695.

On top of its normal variability, PTFO 8-8695 was found to exhibit periodic fading events, during which it dims by a few percent for an interval of ~1.8 hours. The leading interpretation of these events was that they are caused by the transit of a planet with a precessing orbit.

Normally we would test this hypothesis by looking for radial-velocity variation of the host star, but stellar activity prevents us from being able to pick out a signal here. So a collaboration of scientists led by Liang Yu (Massachusetts Institute of Technology) set out to systematically test this theory using less conventional approaches.

Failed Tests

Yu and collaborators used three different tests:

  1. They looked for slow changes in the light curve morphology.
    A planet with a precessing orbit would leave a distinctive signature in the appearance of the light curve. The team, however, didn’t find this predicted morphology. Even more importantly, they discovered that the fading events aren’t even strictly periodic — in recent observations, the period has decreased.
  2. They tried to detect the planet’s radiation in infrared.
    Planets emit in infrared, so by observing the system during times when the planet is supposedly passing behind the star, there should be a visible dip in the measured infrared radiation. No dip of the predicted size was found.
  3. They looked for the Rossiter-McLaughlin effect.
    The RM effect is an anomaly that appears in spectroscopic data of a planet’s transit due to stellar rotation. To check for this, the team obtained high-resolution spectroscopy during a fading event. No effect was seen at the predicted level in their data.

Given these three failed tests, and the new evidence that the fading events aren’t strictly periodic, the authors argue that these events aren’t caused by a planet transiting PTFO 8-8695. Instead, the team put forward a few new hypotheses, including starspots, eclipses by circumstellar dust, or occultations of an accretion hotspot. Further observations should help to narrow down the possibilities.

Citation

Liang Yu et al 2015 ApJ 812 48. doi:10.1088/0004-637X/812/1/48

star formation

Extremely high star formation rates have been observed in galaxies at high redshifts, posing somewhat of a mystery: how are these enormous rates achieved? A team of scientists has proposed that these high rates of star formation could be explained by feedback from active nuclei at the centers of the galaxies.

Pressurized Bubble

We believe that star formation occurs in galaxies as a result of gas clumps that collapse under their own gravity, eventually becoming dense enough to launch nuclear fusion. Recently, there’s been mounting evidence that the star formation rate is significantly higher in high-redshift galaxies, particularly those with active galactic nuclei (AGN). Could this simply be caused by a higher gas fraction at higher redshifts? Or is it possible that a different mechanism is at work in these galaxies, producing more efficient star formation?

A team of authors led by Rebekka Bieri (Paris Institute of Astrophysics) has proposed that this enhanced star formation may be caused by positive feedback from the active nucleus of the galaxy. The team suggests that an outflow from the AGN could create an over-pressurized bubble around the galactic disk that pushes back on the disk, leading to a higher rate of star formation.

Simulating a Boost

The authors test this toy model by simulating the scenario. They model a disk galaxy with roughly a tenth of the mass of the Milky Way, which starts in a relaxed state. The galaxy is then evolved either with or without an applied external pressure, representing the isotropic pressure from the bubble created by the AGN outflow. These models are tested in two different scenarios: one where the initial gas fraction is 10%, and one where the initial gas fraction is 50%.

Star formation rates for the low-gas-fraction (left) and high-gas-fraction (right) simulated galaxies. The blue lines show the rates without external pressure; the red lines show the rates with external pressure applied. Click for a better look! [Bieri et al. 2015]

Star formation rates for the low-gas-fraction (left) and high-gas-fraction (right) simulated galaxies. The blue lines show the rates without external pressure; the red lines show the rates with external pressure applied. [Bieri et al. 2015]

The simulations show that in all cases, the presence of external pressure causes the star formation rates to be significantly higher than in the cases with no feedback. For galaxies that contain a high gas fraction, this difference is especially pronounced at early times; the external pressure causes gas to fragment into clumps that are dense enough to form stars much sooner than in the absence of pressure.

Bieri and collaborators also track the mass entering and leaving the galactic disk during the simulation. When external pressure is present, it drives a large net mass inflow at the beginning of the simulation, carrying in material from the halo and outer disk. This early source of mass helps to feed cloud and clump growth in the inner parts of the galaxy — contributing to the increased star formation rate in the pressurized cases.

Thus, this simple model has demonstrated that enhanced pressure due to AGN feedback is an effective means of explaining the large star formation rates we observe in high-redshift galaxies.

Citation

Rebekka Bieri et al 2015 ApJ 812 L36. doi:10.1088/2041-8205/812/2/L36

compact object merger

The merger of two neutron stars (a NS–NS merger) is suspected to be the most likely source of short-duration gamma-ray bursts (GRBs) — powerful explosions that can be seen from billions of light-years away. But whether a GRB is launched is dependent on what remnant is created by the merging NSs. Do they form another NS? Or a black hole (BH)?

Uncertain Remnant

If the NS–NS merger forms a BH remnant, a GRB can be launched during the ensuing accretion. But if it instead forms a NS, a GRB may only be launched if the remnant collapses to a BH within 100 milliseconds; any longer, and theory says that the GRB jet will become loaded with baryons and choke.

Unfortunately, determining whether the merger will produce a NS or a BH is difficult. A major limitation is that we don’t know what equation of state describes the interior of a NS — which means we also don’t know what maximum mass a NS can have before it collapses into a BH.

Led by Chris Fryer of the University of Arizona and the Los Alamos National Laboratory, a group of researchers undertook a highly collaborative study to better understand the fates of NS–NS mergers.

Maximum Mass

fractional outcomes

The fraction of mergers that produce BHs (and, consequently, GRBs) and NSs, as a function of the maximum NS mass allowed by the equation of state. Lines labeled BHAD are mergers that produce BHs (under two different initial conditions); lines labeled NS are those that produce NSs. [Fryer et al. 2015]

The authors used a combination of merger calculations, neutron star equation of state studies, and population synthesis simulations to model the outcome of the merger of two NSs. With this information, they determined the statistical likelihood that the remnant that forms in the merger collapses directly to a BH, collapses to a BH after a delay, or remains a NS.

Fryer and collaborators find that the outcome is highly dependent upon the maximum mass allowed by the uncertain NS equation of state. If this maximum NS mass is below 2.3–2.4 solar masses, most NS mergers will result in a BH within 100 milliseconds of the merger. In this case, most mergers would be capable of producing GRBs. If, on the other hand, the maximum NS mass is above this cutoff, then the majority of NS mergers will form a NS remnant — only rarely launching a GRB.

Since, to match observed GRB rates, the second scenario requires a rate of mergers significantly higher than what theory predicts, it seems more likely that NS masses are limited to 2.3–2.4 solar masses. Upcoming observational projects like advanced LIGO will help to test this theory and place further constraints on our models of NSs.

Citation

Chris L. Fryer et al 2015 ApJ 812 24. doi:10.1088/0004-637X/812/1/24

M31

Usually stars that are born together tend to move together — but sometimes stars can go rogue and “run away” from their original birthplace. A pair of astronomers have now discovered the first runaway red supergiant (RSG) ever identified in another galaxy. With a radial velocity discrepancy of 300 km/s, it’s also the fastest runaway massive star known.

Discrepant Speeds

When massive stars form in giant molecular clouds, they create what are known as OB associations: groups of hot, massive, short-lived stars that have similar velocities because they’re moving through space together. But sometimes stars that appear to be part of an OB association don’t have the same velocity as the rest of the group. These stars are called “runaways.”

What causes an OB star to run away is still debated, but we know that a fairly significant fraction of OB stars are runaways. In spite of this, surprisingly few runaways have been found that are evolved massive stars — i.e., the post-main-sequence state of OB stars. This is presumably because these evolved stars have had more time to move away from their birthplace, and it’s more difficult to identify a runaway without the context of its original group.

An Evolved Runaway

velocity difference

Difference between observed velocity and expected velocity, plotted as a function of expected velocity. The black points are foreground stars. The red points are expected RSGs, clustered around a velocity difference of zero. The green pentagon is the runaway RSG J004330.06+405258.4. [Evans & Massey 2015]

Despite this challenge, a recent survey of RSGs in the galaxy M31 has led to the detection of a massive star on the run! Kate Evans (Lowell Observatory and California Institute of Technology) and Philip Massey (Lowell Observatory and Northern Arizona University) discovered that RSG J004330.06+405258.4 is moving through the Andromeda Galaxy with a radial velocity that’s off by about 300 km/s from the radial velocity expected for its location.

Evans and Massey discovered this rogue star via a photometric survey of RSGs in M31, followed up by spectroscopy with the Multiple Mirror Telescope. They determined that the star is also separated from other massive stars in the disk of the galaxy by about 4.6 kpc — which is roughly the distance it would be expected to travel, given its discrepant motion, in an assumed age of about 10 Myr.

The authors suggest that this star may be a high-mass analog of “hypervelocity stars” — stars within the Milky Way that are moving fast enough to escape the galaxy. The authors demonstrate that the total discrepant speed of RSG J004330.06+405258.4 is probably comparable to the escape velocity of M31’s disk.

But whether or not this star is moving fast enough to escape turns out to be moot: it will only live another million years, which means it won’t have enough time to leave the galaxy before ending its life in a spectacular supernova.

Citation

Kate Anne Evans and Philip Massey 2015 AJ 150 149. doi:10.1088/0004-6256/150/5/149

1 … 108 109 110 111 112 … 114