Online material

Appendix A: Detailed description of the stars with companion candidates

Appendix A.1: LRa01_E1_2101

Seeing-limited images obtained with the 3.6 m Canada France Hawaii Telescope (CFHT; located at Mauna Kea, Hawaii, USA), the IAC 80 cm (located at Izaña Tenerife, Spain), and the 1.2 m robotic MONET telescope (located at McDonald Observatory, USA) did not show any stars in the field that became significantly fainter during transit. The NaCo image shows a previously unrecognized star with J = 16.3 ± 0.1 at a distance of ( E and N) from the primary (Fig. 1). If it were an unrelated star to the target and using the average colour of stars in the field, we estimate that it would be about V_UCAC ~ 17.9. If it were a physical companion, its brightness would correspond to that of an M5V-star, which would be V ~ 21.5 and R ~ 19.7. Thus, no matter if the CC were related or unrelated to the primary, it is in both cases bright enough to be an FP. If the CC were a physical companion, it would be at a distance of ~300 AU from the primary.

We obtained a spectrum with CRIRES of LRa01_E1_2101 in the range of the CO bands (2284.1 to 2322.9 nm). To search for additional companions, we kept the CC outside of the slit. Figure 2 shows part of the CRIRES spectrum together with a spectrum of a sunspot (Wallace & Livingston 1992). The spectrum of LRa01_E1_2101 shows only the CO lines of the primary but not of any other additional star. Using the cross-correlation function, we derived an upper limit for possible additional companions. Figure 3 shows the cross-correlation function of LRa01_E1_2101, together with that of a hypothetical M3V companion star. We can thus exclude that there is any additional CC with a spectral type of M3V or earlier with a separation of 03 or less.

Appendix A.2: LRa01_E1_4719

Seeing-limited images obtained with the MONET telescope did not show any contaminants. According to EXODAT, the spectral type of the object is F8IV. However, in the NaCo images, the target is nicely resolved into two stars with a separation of 08. This is shown in Fig. 1 where the fainter component can be found 071 W and 044 S of the brighter one. To assess the nature of the companion, we obtained J- and K-band images. From the images, we derive J = 14.8 ± 0.1 and K = 14.5 ± 0.2 for the primary and J = 15.8 ± 0.1, K = 15.4 ± 0.2 for the secondary.

Figure 4 shows the absolute brightness in the J-band (M_J), and the J − K colours, assuming that the two objects are at the same distance, together with the absolute brightness and colours of stars taken from Lépine et al. (2009), Henry et al. (2006), and Bilir et al. (2009).The brightness as well as the J − K-colours of the secondary is in agreement with being a physical G9V-companion at a projected distance of ~1900 AU. If the CC were either bluer or redder, we would know that it is an unrelated star. Table 4 gives the brightness of the companion in the optical regime for the case that it is a physical companion, as well as for the case that it is an unrelated background star. In both cases the CC is bright enough to be an FP.

The transit could thus either be on the primary, or on the secondary. If it is on the primary, the size of the occulting object would be ~0.4 R_Jup, and if it is on the secondary, the size would be ~0.7 R_Jup. Thus, in both cases the transiting object could be a planet.

Appendix A.3: LRa02_E1_1715

The spectral type of this object is G2IV/V, as determined from a HIRES spectrum taken with the Keck telescope. Seeing-limited images with CFHT, the 1.2 m Leonard Euler telescope at ESO, La Silla, and the IAC 80 cm telescope gave us the result that the transit is on target. However, NaCo resolves the target into two components, which have a separation of 152 (companion 140 W, 060 N of the primary and J = 18.8    ±    0.1, K = 17.7 ± 0.1; see Fig. 1). In order to constrain the nature of the companion, we took J- and a K-band one. Figure 5 shows the absolute brightness in the J-band (M_J), assuming that the two objects are at the same distance, together with the J − K colours of the two stars. Both stars are slightly redder than the standard stars. Since they are reddened by the same amount, the data are fully consistent, with the hypothesis that their two stars are at the same distance. Thus, the colour and the brightness difference between the primary and secondary is consistent with an M4V star companion at a projected distance of about 1400 AU. If the transit were on the primary, the occulting object would have ~0.2 R_Jup and where thus be in the planetary regime. In the case the transit is on the secondary, the occulting object would have a radius between ~1.5 and ~3 R_Jup, corresponding to a low-mass star.

Appendix A.4: LRa02_E2_1136

Images taken with the 1m ESA OGS telescope (Izaña Tenerife, Spain) allow exclusion of FPs with V < 21 at distances larger than three arcsec from the target. We obtained NaCo images in the J- and in the K-band of this object. In both images the object is resolved into two stars with a separation of 04 (primary: J = 12.6 ± 0.1, K = 12.2 ± 0.1, secondary: 034 ± 003 E, 016 ± 003 S, J = 14.5 ± 0.1, K = 13.6 ± 0.1; Fig. 1). The primary is a G0V star. Since not only the brightness difference but also the J − K colour matches that of a K4V to K5V companion, it is likely that this star is a physical companion at a projected distance of ~800 AU (Fig. 6). Because the transit is only 0.3% deep, it could be either on the primary or the secondary. If it is on the primary, the transiting object would be ~0.6 R_Jup, and if it is on the the secondary, it would be due to a ~2.4 R_Jup object. Thus it could either be a planet or a low-mass star. Interestingly, the NaCo images show another binary (Corot-ID 110676867, 2MASS06515971-0536425) in the same field of view, but the distance to LRa02_E2_1136 is quite large, 1144 (956 E and 656 N). The distance between the two stars is 027, and the two components are J = 15.5 ± 0.2 and J = 15.6 ± 0.2.

Appendix A.5: LRa03_E2_0861

When making the acquisition image of LRa03_E2_0861 with CRIRES, we recognized a CC. The CC is at a distance of 108 from the primary (companion 078 E, 075 N of the primary; Fig. 1). A subsequent image taken with NaCo showed that the CC is J = 16.4 ± 0.1. The star is thus bright enough to be a FP (Table 4). If this star were a physical companion, it would be an M4V star at a 600 AU. Seeing-limited images taken during transit with the Euler telescope and the CFHT telescope show the transit, but the faint star detected with NaCo and CRIRES can not be fully excluded as an FP. The situation is the same as for LRa02_E2_1136: If the transit is on the primary, the object has the size of a planet. If it is on the companion, it has the size of a star.

Appendix A.6: LRa04_E2_0626

The NaCo image of this star shows a previously unrecognized star, which is J = 16.8 ± 0.1, at a distance of 09 from the primary (08 E, 04 S; Fig. 1). If it were a physical companion, it would be an M3V star at a 600 AU. Images taken during transit with the 0.6 m TRAPPIST telescope at ESO, La Silla (Gillon et al. 2011), show part of the transit, but the faint star detected with NaCo is not resolved and thus cannot be fully excluded as an FP. The CRIRES spectrum does not show any additional CC. Also for this object we have the same sitation as above: The transiting object can have either the size of a planet or a star. That depends on whether the transit is on the primary or the secondary.

Appendix A.7: LRc07_E2_0158

The NaCo image of this star shows a previously unrecognized star, which is J = 14.6    ±    0.1, at a distance of 09 from the primary (00 E, 09 N; Fig. 1). A deep image taken with the EULER telescope, in fact, shows no sign of this star. NaCo was thus required to detect it. A spectrum taken with the TLS-NASMYTH spectrograph mounted on the 2 m Alfred-Jensch telescope at Tautenburg (Germany) shows that this star has a spectral type F9IV or F9V. If it were a physical companion, it would be an M1.5V star at 400 AU. As shown in Table 4, the star is bright enough to be an FP.

Appendix B: Detailed description of the stars without companion candidates

Appendix B.1: LRa01_E1_0286

We observed LRa01_E1_0286 with NaCo and found six faint stars: The closest star has a brightness of J = 15.7    ±    0.1 and a distance of 50 (10 E, 49 S). This star has already been detected in the optical. The next closest is at a distance of 86 (80 W, 30 S) and is only J = 17.4. No additional star closer to the target was found.

We also obtained a CRIRES spectrum of this star. The spectrum shows the CO lines with a strength corresponding to an M0V star that is about 2−3 mag fainter. However, a TLS-NASMYTH spectrum shows that the star has a spectral-type G0V star. Since the strength of the CO lines is what is expected for an early G star, the CO lines are thus presumably the lines of the primary.

Appendix B.2: LRa01_E1_2240

LRa01_E1_2240 was observed with NaCo but not with CRIRES. We found three nearby stars that are at 56 (36 W, 43 N), 64 (39 W, 51 N), and 72 (46 W, 55 N) distance. The stars have a brightness of J = 17.8 ± 0.1, J = 16.4 ± 0.1, and J = 17.9 ± 0.1, respectively.

Appendix B.3: LRa01_E1_4667

We obtained a spectrum with CRIRES of this star. This spectrum allows exclusion of a companion star with a spectral type M3.5V or earlier. Since CRIRES is also an AO-instrument, we can also use the aquisition image to search for previously unrecognized CCs close to the star. From the aquisition image taken in the K-band we can exclude a companion of roughly the same brightness as the target within two arcsec.

Appendix B.4: LRa01_E2_0165, or CoRoT-7

As part of this programme, we obtained deep images with NaCo and CRIRES of the field surrounding the CoRoT-7 object. The results are discussed in detail in Léger et al. (2009), where we report that we did not find any CCs. We do not discuss this object further here but refer instead to the above-mentioned paper.

Appendix B.5: LRa02_E1_1475

This star is an A4V star with a transit that is 0.3% deep. This means that the eclipsing object could have a size of about 0.1 R_sun, corresponding to the size of Jupiter. From the depth of the transit and the spectral type of the primary, we find that an FP would have to be a star with a spectral type of K7V, or earlier that is being eclipsed. We thus have to exclude stars with spectral type K7V to prove that the transit is on target. The aquisition image shows the star 2MASS06512856-0348468 which has K = 15.8 and thus 3.2 mag fainter than the target in K. The images show three additional stars. The first one is at a distance of 28 south-east of the target, the second 48 (43 W, 23 N), and the third one at a distance of 65 (64 W, 07 S). All four stars had already been observed with the CFHT and in all cases an FP was ruled out. The CRIRES aquistion image alone already allows physical companions with spectral types earlier than K3V to be ruled out. Using the Mg 4383.23 cm^-1, and Fe 4396.25 cm^-1 line, we can rule out companions with spectral type F6V stars or earlier. The observation thus does not fully rule out all possible FPs but still a large number of them.

Appendix B.6: LRa02_E1_4601

For this object we carried out the most comprehensive study by taking NaCo images in J and K and CRIRES spectra in both settings. We can thus use these data to assess which setting is the most sensitive.

The NaCo J-band image would have allowed us to detect a star of J = 17 at a distance of 018 from the primary, a star of J = 18 at 023 distance, and a star with J = 19 at 058 distance. The limits for the K-band image are 016 for a star of K = 16, 030 for K = 17, and 051 for K = 18. If we assume that we want to detect a physical binary, than the limit in both filters is almost the same, although the J-band image is deeper. In both cases, we can exclude companions with spectral types earlier than M2.5V for distances ≥02 from the primary. For distances ≥025, the limit is an M3.5V star and for distances ≥05 a M4.5V star.

Figure B.1 shows part of the spectrum taken with the first setting which contains prominent Ca I lines. This spectrum allows exclusions of companion stars that are earlier than M1V. We note that the Ca-lines at 2261.4 nm and 2263.1 nm (vacuum) are single in the spectrum of the Sun and LRa01_E1_2101 but double in the spectrum of the sunspot due to the Zeeman effect (the magnetic field strength of the spot is 3360 Gauss). Figure B.2 shows a part of CRIRES spectrum taken with the second setting containing the CO lines. Although this spectral region is close to the edge of the K-band and thus the sensitivity of CRIRES is lower, the number of CO lines is so large that the sensitivity for detecting CCs is an higher. Using this spectral range, we exclude companions earlier the M2.5V. Thus, the second setting is more sensitive for detecting stars with CO lines.

As a third approach we also derived the spectral energy distribution, that shows no excess which would indicate a late-type companion (Fig. B.3). Altough this results conforms to the previous ones, the SED method is significantly less sensitive than the observations with NaCo and CRIRES.

	Fig. B.1 Part of the CRIRES spectrum of LRa02_E1_4601 taken with the first setting, together with a spectrum of the Sun and a sunspot.
Open with DEXTER

	Fig. B.2 Part of the CRIRES spectrum of LRa02_E1_4601 taken with the second setting, together with a spectrum of the Sun and a sunspot.
Open with DEXTER

	Fig. B.3 Spectral energy distribution of LRa02_E1_4601. There is no excess in the IR. This is consistent with the NaCo and CRIRES results that the third object of this system is not a star.
Open with DEXTER

Appendix B.7: LRa02_E2_2057

The closest star to LRa02_E2_2057 seen in the NaCo image is at a distance of 51 (22 E, 28 N), and it is J = 19.1 ± 0.1. There is another star with J = 21.1 ± 0.1 at a distance of 74 (74 E, 00 N). Since an FP could be as faint as J = 20.9 ± 1.0, both stars are bright enough to be FPs. Observations with the IAC 80 cm telescope and the CFHT show that neither star is an FP. The CRIRES spectrum of LRa02_E2_2057 excludes companions with spectral types M0V or earlier. This means that the CRIRES spectrum allows companions with R ≤ 18.6 ± 0.2 to be excluded. However, an FP could be as faint as R = 22.5 (corresponding to a companion with spectral-type M3V). There is thus still the possibility for an FP but this companion would have to be much closer than one arcsec to the target and it would have to have a spectral type in the range between M1V and M3.5V. A companion with exactly these properties is not very likely.

Appendix B.8: LRa02_E2_3804

The observations of LRa02_E2_3804 with NaCo did not show any CC. The closest star to LRa02_E2_3804 is already at a distance of 100 (52 W, 86 N), and it is J = 18.3 ± 0.1.

Appendix B.9: LRa03_E2_0678

The expected brightness of an FP is J ≤ 18.9 ± 1.0. The closest star found is at a distance of 98 (21 E, 96 N) and has J = 19.4 ± 0.1. Observation with the CFHT and with the IAC 80 cm telescope rules out that this star is an FP. The CRIRES observations exclude companions earlier than M1V, corresponding to a star with R = 13.9. Also in this case, the CRIRES observations are not deep enough to exclude close companions of very-late spectral type, but they exclude at least all companions earlier than M1V.

Appendix B.10: LRa03_E2_1326

In the case of LRa03_E2_1326, the closest stars found in the NaCo images are already at distances of 83 (64 E, 52 N) and 105 (01 E, 104 S) and have a brightness of J = 16.9 ± 0.1 and J = 17.9 ± 0.1.

Appendix B.11: LRa06_E2_5287

The NaCo image obtained of LRa06_E2_5287 shows ten stars, in addition to the target. The closest one is at a distance of 36 (13 E, 34 S) but it is J = 20.3. The next closest star is at a distance of 50 (37 W, 34 N), but it is even J = 21.4. These stars should thus are too faint to be FPs.

CRIRES spectrum was taken that shows weak CO lines. A subsequent spectrum taken with the TLS-NASMYTH spectrograph shows that the primary is a G0V star. Since such a star has weak CO lines, we interpret the lines as coming from the target itself.

Appendix B.12: LRa07_E2_3354

Like LRa02_E1_1475, LRa07_E2_3354 an early-type star. The aquisition image taken with CRIRES shows no star within 100 arsec of the target. For the CRIRES observation, we used the first setting and the Mg 4383.23 cm^-1, and Fe 4396.25 cm^-1 to exclude companions. Given that a TLS-NASMYTH spectrum shows that it is a B9 star, we can rule out companions earlier than A6V. Since the transit is 3.9% deep the eclipsing object could still well be a late K or early M star.

Appendix B.13: LRc02_E1_0591

This object is is located in the “galactic center” eye (LRc-fields) of CoRoT. The density of stars in this region is much higher than in the “galactic anti-center” eye (LRa-fields). The large number of stars in this field made it necessary to obtain two images with NaCo. One was taken during transit and the other out of transit. Figure B.4 shows the image taken out of transit. Table B.1 gives the brightness measurements and their differences obtained in the two images. Three stars, labeled PVP1, PVP2, and PVP3 in Fig. B.4, and Table B.1, were a little bit fainter during transit than out of transit. Since potential FPs of LRc02_E1_0591 have to be J ≤ 17.6 ± 1.6 (see Table 2), these stars could potentially be FPs. However, images taken with the IAC 80 cm and the Euler telescope show that all of them are too faint in the optical and thus can not be FPs.

	Fig. B.4 NaCo image of LRc02_E1_0591 taken in the J-band. Marked as PFP 1, 2, 3 are stars that became fainter during the transit (“potential FP”). See also Table B.1. North is up and east is left.
Open with DEXTER

Appendix B.14: SRa01_E1_0770

The two closest stars to SRa01_E1_0770 are at a distance of 62 (52 E, 35 S) and 110 (12 E, 109 N) and have J = 18.8    ±    0.1, and J = 17.8 ± 0.1. Since we have to remove FPs with J ≤ 17.5 ± 1.2, the second star is just bright enough that it could be an FP. However, seeing-limited observations are sifficient to find out if this star is an FP or not. The situation with the CRIRES observations is the same as for the other stars. The spectra allow exclusion of companions earlier than M0V, corresponding to stars of R = 17.0, but this is not sufficient to fully exclude all types of FPs.

Appendix B.15: SRa02_E1_1011

Using a spectrum obtained with the TLS-NASMYTH spectrograph, we derived the spectral type of this star, which is F6V. The NaCo image shows two additional stars. They are J = 18.9 ± 0.1 and J = 16.2 ± 0.1 and are at distances of 89 (56 E, 69 N), and 98 (78 E, 58 S), respectively. Both stars are thus well separated from the primary, and observations with seeing-limited telescopes could show if these are FPs or not. The CRIRES observations permit us to exclude companions earlier than M0V corresponding to stars with R ≥ 17.7. As before, the CRIRES observations are not deep enough to exclude companions of very late-type, but they do allow the exclusion of most of them.

Appendix B.16: SRa03_E2_1073

A TLS-NASMYTH spectrum shows that this star has a spectral type F3V. The CRIRES spectrum allows exclusions of a companion that is 1.7 mag fainter in K and has CO lines like an M0V star. However, this does not exclude stars with spectral type K2V because the CO lines in these stars are weaker. Since the CO lines become progressively shallower for earlier type stars, an eclipsing binary with spectral type G cannot be fully excluded.

Appendix B.17: SRa03_E2_2355

The three closest stars detected in the NaCo images of SRa03_E2_2355 are J = 17.0 ± 0.1 (distance of 30; 08 E, 252 S), J = 20.2 ± 0.1 (distance of 42; 40 E, 12 S), and J = 18.1 ± 0.1 (distance of 92; 40 E, 83 S).

Appendix B.18: SRa04_E2_0106, CoRoT-32

This object is CoRoT-32. Since it is discussed in Gandolfi et al. (2012), we will just briefly mention the results here. We obtained a spectrum and an aquisition image with CRIRES. The limiting magnitude of the aquistion image is K ≥ 13.9. No additional stars were detected within 124 of the target, which means that physical companions with a spectral type M2V or earlier can be excluded. The CRIRES spectrum itself allows exclusion of physical companions with a spectral type M3V or later.

© ESO, 2013