The Xinglong 2.16-m Telescope: Current Instruments and Scientific Projects

The BFOSC is one of the primary instruments of the telescope, for which the design and processing of mechanism and electronics control; the design of grims and assembling and debugging of the whole system were done by the University of Copenhagen, and ESO provided consult and optical design of the focal reducer. It is available for the f/9 Cassegrain focus. The scale of the focal plane is 1061 mm⁻¹. As discussed in Section 1, it can be used for both imaging mode and spectroscopy mode, which is switchable. Figure 3 shows the optical layout of the BFOSC instrument, including the aperture wheel, filter wheel, grism/echelle wheel, calibration device, guiding device, collimator, shutter, camera, and the CCD detector.

Zoom In Zoom Out Reset image size

There are eight positions on the aperture wheel: three of them for direct imaging, coronagraph mask, and focal adjusting plate, respectively, and the other five for long/short slit plates. For the long slits, the lengths are all 94, and there are nine options for the slit widths (06, 07, 11, 14, 18, 23, 36, 70, and 140). While for the short slits, the slit lengths are various for different slit widths: 35 for slit width of 06, 40 for slit width of 10, 36 for slit width of 16, and 37 for both slit widths of 23 and 32. The coronagraph mask is used to block out the intense light from bright sources near the observing targets, with circular spots of which the diameters are 20, 30, 40, 60, 90, and 120.

For the filter wheel, there are several sets of filters available. The Johnson-Cousins UBVRI filters are for the broadband photometry. In addition, there are special filters used for the spectroscopic observing: the Z band for the transmitting spectral region λ ≥ 910 nm, and the 385LP band for removing the second-order spectrum in the wavelength region λ ≤ 385 nm. The transmission curves are shown in Figure 4.

Zoom In Zoom Out Reset image size

Besides, a series of interference narrowband filters covering [O iii] (rest-frame wavelength of λ5007A), Hα, and He ii(rest-frame wavelength λ4686A) is also available for the BFSOC imaging mode. For series of [O iii] band observations, there are eight filters of which the central wavelengths are between 500.9 and 536.0 nm with FWHM of 6 nm, corresponding to redshifts of 0–20000 km s⁻¹. While for the series of H_α bands, there are 11 filters of which the central wavelengths are between 656.2 and 706.0 nm with an FWHM of 7 nm, corresponding to redshifts of 0–22000 km s⁻¹. For the He bands, there are three filters of which the central wavelengths are between 447.1 and 468.6 nm with an FWHM of 6 nm. In addition, an [O iii] filter with an FWHM of 13 nm and an H_α filter with FWHM of 14 nm are also provided for a redshift of z = 0. These filters can be used for the study of star-forming regions of nearby galaxies at different redshifts.

Table 1 presents the parameters of the grisms/prisms/echelle for the BFOSC instrument. From left to right in the columns are names, working spectral orders, reciprocal linear dispersions, dispersions, and wavelength coverages of the various grisms/prisms/echelle, respectively. The configurations can be chosen by the users depending on the different requirements of the projects. For grisms G3/G6/G10, the blue ends of wavelength coverages are limited by the cutoff of the atmospheric window, while for grisms G5/G8, the red ends are limited by the size of the CCD. The low dispersion grisms G10/G11/G12 are also used for the cross-disperser when mounted on the filter wheel. The echelle E13 is only used for measuring the velocity field of extended sources with the third-order spectrum, and the V band filter is recommended to be applied together with it to remove the other order of the spectrum.

Table 2 presents the spectral resolutions of some frequently using BFSOC grisms at a minimum slit width 06 and a slit width of ∼23 used in typical seeing conditions, which are estimated with the emission lines of the planetary nebula (PN) IC4997. The observations were taken on 2014 November 24 and 2016 February 26, with gratings of G4, G6, G7, and G8 on BFOSC. Throughout the nights of the observations, the weather was clear, and the seeing was between ∼20 and 24. The exposure time was 0.2–10 s depending on the dispersions of the different grisms. The shortest exposure of 0.2 s is considerably longer than the shutter speed of 1.5 ms.

Table 2.  Spectral Resolutions of Several Frequently used Grisms of BFSOC at a Slit Width of 06 and 23, which is around the Typical Seeing

Wavelength	G4	G6	G7	G8	G4	G6	G7	G8
(Å)	06				23
4341	531	1423	1204	⋯	246	463	478	⋯
4363	557	1503	1252	⋯	⋯	⋯	⋯	⋯
4861	⋯	⋯	⋯	⋯	262	538	521	⋯
4959					265	540	524	⋯
5007	620	1413	1328	⋯	265	555	532	⋯
6563	824	⋯	⋯	2245	321	⋯	⋯	820

Download table as:  ASCIITypeset image

For the calibration device, four lamps can be mounted at most. One is for the flat-fielding correction, and the other lamps can be used for wavelength calibration. The lights of the lamps illuminate the integrating sphere at first, then are modified to be f/9 light beam, and finally are reflected to the focal plane of the BFOSC instrument. The Fe lamp and Ne lamp are frequently used for the wavelength calibration.

On the guiding device, a mirror reflects the view outside of the observing field of view (FOV) to an ICCD, which is movable in 3D directions to find a proper guide star and also to adjust the focus. The FOV of the guiding device is 10' × 20', which is large enough to find a suitable guide star.

In 2010, an E2V 55–30–1–348 back-illuminated CCD, AIMO, was installed on the spectrograph, and the CCD controller was made by the Lick Observatory. The size of the CCD is 1242 × 1152 pixels with a pixel size of 22.5 μm. The pixel scale is 0457, and the FOV is 946 × 877 according to the size of the CCD. Figure 5 shows the quantum efficiency (QE) of the CCD. It can be seen that the maximum QE is higher than 90% around 5700 Å of the wavelength.

Zoom In Zoom Out Reset image size

Table 3 lists the gains, readout noises, and readout speed of the BFOSC CCD at various readout times. Observers can choose different options for the specific observations, and the slow readout speed is applied in most of the time.

There are six observing modes for the BFOSC instrument: (1) direct imaging, (2) long-slit spectroscopy, (3) slitless spectroscopy, (4) echelle grism spectroscopy, (5) coronograph mask, and (6) multiple-object spectroscopy. Figure 6 shows the optical layout of the lens and light path for the direct imaging mode and the spectroscopy observing mode, which could be switched in a few minutes.

Zoom In Zoom Out Reset image size

Since 2012, a multi-object spectroscopy (MOS) observing mode of BFSOC is available for observers by placing a multiple-aperture mask on the aperture wheel (Zhou et al. 2014). In this mode, 10–20 objects can be observed simultaneously, depending on the spatial distribution density of the targets. The MOS improves the observing efficiency of multiple-object observations, such as star-forming activities of H ii regions of nearby galaxies, star clusters, or groups and clusters of galaxies.

The total efficiency of the observing system, including the atmosphere, telescope, and its instruments, is what the observers are mostly concerned about. In fact, there are a number of factors to be considered for calculating the total efficiency of the system, such as atmospheric extinction, reflectivity of primary and secondary mirror, transmissions of filters, and the quantum efficiency of the CCD. It can be estimated through observing standard stars in Equation (1):

$$\begin{eqnarray}&&\eta (\lambda )=\displaystyle \frac{{F}_{{\rm{ADU}}}\cdot G}{{F}_{\lambda }\cdot \delta \lambda \cdot {S}_{{\rm{tel}}}}\mathrm{.}\end{eqnarray} \tag{ 1 }$$

In this formula, F_ADU is the observed number counts of a standard star per second (ADU s⁻¹), G is the gain of the CCD (e⁻¹ ADU⁻¹), F_λ is theoretical photon flux of a standard star derived from its AB mag (photon s⁻¹ cm⁻² Å⁻¹), δλ is the effective bandwidth of the filter in imaging observations or the dispersion of the grating for spectroscopic observations (Å), S_tel is the effective area of the primary mirror of the telescope (cm²), and λ is the effective wavelength of the filter or the wavelength at which the efficiency is to be computed for the spectroscopy (Å).

In order to estimate the total efficiency of the 2.16-m telescope, a Landolt standard star PG2336+004B was observed on 2014 November 21 in the broadband UBVRI bands of the BFOSC photometric system by using Equation (1). The seeing was ∼20–30 throughout the whole night of observations, and the airmass is ∼1.3. Figure 7 is a plot of the total efficiency in different bands as a function of central wavelength. The atmospheric extinction, reflectivity of primary and secondary mirrors, transmissions of filters, QE of CCD, as well as other factors, were not corrected in the calculation. It can be seen that the total efficiency is relatively low in the U band (∼2%) and B band (∼7%), but it is relatively high in the VRI bands (∼15%).

Zoom In Zoom Out Reset image size

The limiting magnitude is also an important quantity to evaluate the observing ability of a telescope. Actually, at the same night on 2014 November 21, a Landolt standard star PG2336+004 was observed in the UBVRI bands. The airmass was ∼1.3. The limiting magnitude (signal-to-noise ratio, S/N = 5) as a function of exposure time in the UBVRI bands are shown in Figure 8, and the same relations but for the limiting magnitudes of S/N = 10 are shown in Figure 9. It is noted that in the 640 s exposure the limiting magnitudes of the B/V/R bands could reach ∼22 mag in S/N = 5 and ∼21 mag in S/N = 10. The limiting magnitude in the U band is the lowest because the sensitivity of the observing system in the blue band is relatively low, further affected by the effect of significant atmospheric extinction in the blue band. However, although the efficiency in the I band is similar to those of the VR bands, the night-sky background is much brighter (∼2 times) than in the VR bands. In addition, the fringe of the CCD in the I band contributes significant noise. Although we have tried the de-fringing method introduced by Chen et al. (1987), it has not been improved significantly. Therefore, these factors together make the limiting magnitudes ∼1 mag shallower than those in the BVR bands.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In order to estimate the efficiency of the spectroscopic system of BFOSC, the two ESO standard stars HR153 and HR9087 were observed on 2014 November 24, with several frequently used grisms (G4, G5, G6, G7, and G8) of the BFOSC instrument. The weather was clear, and the seeing was ∼24. The exposure time was 0.2–10 s according to the dispersions of the grisms. To make sure that most of the flux can be obtained and measured for estimating the efficiency, the slit was configured as 70. The total efficiency, which is defined above, including atmospheric extinction and instruments, was calculated with Equation (1) and shown in Figure 10. For the grisms G4 and G5, the peak efficiencies are ∼15% and ∼10%, respectively, while for the other three grisms, the total efficiencies of the peak are ∼5%. The filter 385LP was used for removing the second-order spectrum of wavelength λ ≥ 385 nm.

Zoom In Zoom Out Reset image size

For the limiting magnitude in the spectroscopic observations, the previous observations show that typically it could reach V = 20 mag of S/N = 5 in the grism G6, with an exposure of 1 hour, given by Huang et al. (2012).

The OMR is another low-resolution spectrograph available on the Cassegrain focus of the 2.16-m telescope. It was made by Optomechanics Research Inc. (Arizona, USA) at the end of 1994 and was tested on the 2.1-m telescope at Kitt Peak. Then, it was installed on the 2.16-m telescope of Xinglong Observatory in 1995 and was available to astronomers in 1996.

The optical layout of the OMR is shown in Figure 11. The system is composed of slit and decker assembly, filters, spectrograph, calibration system, collimator, gratings, CCD camera, guiding CCD camera, console, and data-collecting system.

Zoom In Zoom Out Reset image size

The slit and decker assembly is composed of decker, the slit width adjuster, calibration reflector, and driving device. The top surface of the slit is aluminized stainless steel with high reflecting power, while the slit edge was polished to be very sensitive and should not be touched. The slit is tilted by 20° with respect to telescope optical axis, and the projected slit width can be adjusted remotely through the main console in the range of 0.05–1.0 mm (corresponding to 05–106), which is displayed on the main console via the voltage value. The size of the slit jaws is 32.8 mm × 38 mm, and the effective length of the slit is 28.8 mm (corresponding to 51). The slit jaw reflects the light to the guiding system.

There are six positions on the filter wheel for the filters of Clear, Corning 4–71, Schott BNG–37, BG–39, GG–475, and RG–695. All the filters have the same size of 25 mm × 25 mm. The central wavelengths of filters could be matched with the first- or second-order of the spectrum by the softwares automatically.

There are three wavelength calibration lamps (Fe-Ne, Fe-Ar, and He-Ar) and a flat-fielding lamp for the calibration device. All of lamps can be controlled remotely. For the Fe-Ne lamp, a standard 1.5-inch ISTC Model #WL-22810 is supplied, and other types of lamps actually can be used as well, such as Hamamatsu and Starna. The maximum working current is 20 mA, and the normal working current is 10–12 mA. The Fe-Ar lamp ISTC Model is #WL-22611. For the safety of He-Ar lamp, the DC power should be used for the supply. As for the flat-fielding lamp, the power is the standard 12 V DC, 1.5A halogen tungsten lamp (Sylvania #808-301550), and it usually can be used for 1000–2000 hours. Meanwhile, there are four condensers for the calibration lamps, which are made of quartz or pyrex glass. The diameters are 25 mm, and the focal lengths are f = 38 mm. The field lens is made of quartz, with a diameter of 38 mm and a focal length of 64 mm.

The mirror of the collimator is an off-axis parabolic aluminized reflector, made of pyrex glass. The aperture of the mirror is D = 110 mm, and the focal length is f = 674 mm. The off-axis angle is 8.1° and the focus can be adjusted remotely.

At present, there are six reflecting blazed gratings mounted on the OMR spectrograph. The gratings are made of aluminized pyrex glass and can be switched manually. The parameters of the blazed gratings are given in Table 4, including the number of the gratings, the grooves, groove areas, reciprocal linear dispersions, dispersions, blaze wavelengths of the first-order spectra, and the blaze angles. The overall working wavelength coverage of the gratings is ∼3700–10000 Å. For the specific grating, the wavelength coverage is adjustable, and it can be estimated through parameters of Table 4 and the size of the CCD. For instance, the wavelength coverage of gratings of 1200 lines/mm is ∼1380 Å, while for the grating of 300 line/mm, it is ∼5420 Å. The SPEC software, which is the camera-controlling software designed and installed with the PI Spec-10 CCD camera, can recognize the gratings, and the wavelength coverage can be adjusted via SPEC according to the requirements of the observers.

The camera is a Schmidt-Cassegrain system. The useful aperture is D = 100 mm, and the focal length is f = 150 mm. The CCD mounted is a PI Spec-10 PIXIS 1340 × 400 scientific CCD detector, which delivers the highest sensitivity possible and >16–bit dynamic range for spectroscopy applications. The pixel size is 20 × 20 μm pixels, and the CCD size is 26.8 × 8.0 mm². The deepest cooling temperature is −75° C. The QE of the CCD is shown in Figure 12, which is midband.

Zoom In Zoom Out Reset image size

The guiding system of the OMR is composed of a reflecting system and a CCD Camera. In the reflecting system, the slit reflecting mirror is aluminized and the rms of the surface flatness is 1/4 λ. For the transfer lens, the diameter is 50 mm, with a focal length of f = 260 mm. While for the focal lens, the diameter is 36 mm, with a focal length of f = 85 mm. Both of the lenses have been coated. The CCD camera is an Alta U42 CCD, made by Apogee Imaging Systems, Inc. The sensor is E2V CCD42–40, and the size is 2048 × 2048 pixels. The Gain is 1.3 e⁻ count⁻¹ and the maximum digitized well capacity is 82 k e⁻. The dark current is 0.39 e⁻ pixel⁻¹ s⁻¹.

Similarly, the Landolt standard star Feige 66 was observed with two blazing gratings 1200B and 1200R of the OMR instrument and the total efficiencies (including the atmosphere extinction, the reflecting rate of mirrors, QE of CCD, etc.) have been computed through Equation (1). As described above, the central wavelength and coverage is adjustable; Figure 13 shows the total efficiencies of two gratings 1200B and 1200R, in two different central wavelength and wavelength coverages. It can be seen that the efficiency ranges from ∼1% to ∼2%, which is relatively lower than that of the BFOSC gratings.

Zoom In Zoom Out Reset image size

Typically, for a star of V = 17.3 mag in an exposure time of 1800s, the signal-to-noise ratio is S/N = 12 at wavelength of λ = 5500 Å; for a star of SDSS g = 19.8 mag in an exposure time of 3600 s, the signal-to-noise ratio is S/N = 5 on average for the whole band.

Previously, an echelle spectrograph was mounted on the Coudé focus of the telescope for high-resolution spectroscopic observations. The special design of the optical system is to allow the Cassegrain system and the Coudé system to share the same secondary mirror and there is a relay mirror in the Coudé system. When switching between the two systems, the Coudé system can eliminate spherical aberration and coma aberration sufficiently just by slightly moving the secondary mirror (Su et al. 1989). The resolution power was between R = 16,000 and 170,000 (79 gr mm⁻¹) in the blue beam from 330 nm to 580 nm and R = 13,000 and 170,000 (31.6 gr mm⁻¹) in the red beam from 520 nm to 1100 nm. For an exposure time of 1 hour, the limiting magnitude could reach V = 9.5 in the red band and V = 7.2 mag in the blue band with a signal-to-noise ratio S/N = 100 (please see Zhao et al. 2000; Zhao & Li 2001).

Since 2010, a new fiber-fed High Resolution Spectrograph (HRS) has been developed by the Nanjing Institute of Astronomical Optics & Technology (NIAOT) to satisfy the scientific requirement of, for instance, exoplanet surveys, the study of stellar abundances, and stellar magnetic activities. The optical layout and light path are shown in Figure 14. The HRS system is available for the Cassegrain focus and composed of four parts: 1) the Cassegrain connector of the telescope; 2) the fiber connector and the micro-optical module for focal ratio changing; 3) the main body of the spectrograph; 4) the data-collecting part of the CCD camera.

Zoom In Zoom Out Reset image size

The fiber-fed HRS observing mode could be switched to other instruments (such as BFOSC or OMR) conveniently through the Cassegrain focal interface of the telescope within only a few minutes. The calibration system of HRS, the I₂ cell and its heating system, as well as the tip/tilt system are mounted on the interface of the spectrograph. The whole main body is sealed in a protective box and the temperature variation is less than rms = 0.5°C for a week. At present, a 24 aperture fiber is configured and the tip/tilt system is working to improve the efficiency of the system when the seeing is ideal. The fiber of 16 is at the commissioning stage and it could be available very soon. An environmental controlling system can accurately keep the stability of both temperature and humidity for the HRS system. Furthermore, a sub-controlling system of pressure will be installed at this system in the future.

The basic parameters of HRS are listed in Table 5. The working wavelength coverage is 360–1000 nm, and the instrumental efficiency of the spectrograph is ≥34% for the peak at a wavelength coverage of 640–790 nm (RI band), and ≥10% for λ > 4500 Å (the whole working band), based on the tests of 2010 November 23. The spectral resolution is 32,000–106,000 for the spectrograph, and it is R = 49,800 at a fixed slit width of 0.19 mm (corresponding to 18) based on the test of 2011 April 12. The stability of the instrument for the velocity measurement is rms = ±6 m s⁻¹, and the temperature is quite stable even for two weeks. The CCD camera is a back-illuminated first-order red-sensitive E2V CCD 203-82. The size is 4096 × 4096 pixels with a pixel size of 12 μm. The typical QE of the CCD is >90% under the temperature of −100°C in the wavelength coverage of ∼500–650 nm, which is shown as the solid line in Figure 15.

Zoom In Zoom Out Reset image size

Table 5.  The Current Instrumental Parameters of the HRS

Tech. Index	Parameters	Notes
Wave. Cov.	3650–10,000 Å	ThAr lamp
Aper. of Fiber	24/16
Spec. Res.	32,000–106,000	tested on 2011 Apr 12
Stab. of Inst.	±6 m s−1	Zhao et al. 2014
Effi.	34% at peak value
	≥10% for λ > 4500 Å	tested on 2010 Nov 23
	E2V 4k × 4k 12 μm
	Scientific chip, class 1
	On-chip Binning
CCD Camera	50–200 KHz	2 × 1 binning
	LN holding time: 20 hours	mode suggested
	(w/LN auto-filling)
	LN cooling
FOV of Guid. Cam.	33 × 33
Temp. var./day	±005
Temp. var./week	±034

Download table as:  ASCIITypeset image

The liquid nitrogen (LN) holding time of the system is ∼20 hours, and the cooling temperature is −106°C. For Gain = 1.01 e⁻/ADU, the readout noise (RON) is 2.84 e⁻ at the readout speed of 50 k, 4.29 e⁻ at readout speed of 100 k, and 7.88 e⁻ at readout speed of 200 k. The distance between the nearby orders of the spectrum is ≥20 pixels, and the two fibers (24 aperture one and 16 aperture one) can work at the same time. The aperture of the fiber is 100 μm (24), and the fiber length is ∼19 m. The FOV for the guiding camera is 3' × 3', and in the guiding plate the aperture of pupil is 40 in front of the fiber. The temperature variation of the system is rms = ±0.05°C for a whole night and rms = ±0.34°C for a week. The tip/tilt system is working normally on the Cassegrain focus. The system, including the astro-frequence comb instrument, is currently in its commissioning phase, and its radial velocity precision dose not yet reach its design goal of a few ∼cm s⁻¹. It is expected that this goal will be achieved once commissioning is complete.

The guiding CCD camera is a GC1380, and the software AVTUniCamViewer is used for CCD controlling, exposure time configuration, and data collection. The telescope controlling software can monitor the guiding uncertainty of the guiding system.

An astro-frequency comb calibration system (Zhao et al. 2014, 2015), which is developed by Peking University (PKU), also was installed and is being tested on the 2.16-m telescope. The full spectral wavelength range of the astro-frequency comb is 160 nm and the central wavelength is 640 ± 20 nm. The observed spacing of the comb teeth is 29.01 GHz. Once working normally, it can greatly improve the measuring accuracy of the stellar radial velocity. The system has significant advantages over the I₂ cell (Wilken et al. 2010). It consists of a series of discrete, equally spaced spectral lines with equal intensity, and it is repeatable in a long timescale. Right now, the system has been set up, and the first spectrum simultaneously from the comb and the flat-fielding lamp has been obtained during the engineering run. Meanwhile, a 25 GHz AstroComb Optical Frequency calibration system manufactured by the MenloSystems company, also has been installed on the HRS instrument system recently. The spectral coverage is ∼450–720 nm, with the central wavelength of 540 ± 30 nm. The flatness of the spectrum is rms < 5 dB, and the observed spacing of the comb teeth is 25 GHz. The luminous power of the system for the full spectral wavelength range is >10 μW. The two new calibration systems are at the commissioning stage. The two new systems are supposed to be applied in the end of this year.

In order to estimate the total efficiency (including the atmospheric extinctions and reflectivity of the primary/secondary mirrors, among others), a number of standard stars were observed with HRS on 2015 June 30. By using Equation (1), the final result shows that it is >2% at 5500 Å.

In addition, the limiting magnitudes are estimated. Figure 16 is a plot of the signal-to-noise ratio (S/N) at λ = 5500 Å as a function of exposure time for the stars from V = 5 mag to V = 9 mag for the HRS system, based on the observing data. For a V = 9 mag star, the typical signal-to-noise ratio S/N = 100 for an exposure time of 3000–3600 s. A few observing tests give the S/N in the typical observing conditions of 2012, for a star of V = 8.85 mag with an exposure time of 3600 s, the signal-to-noise ratio is S/N = 80 at wavelength of λ = 6000 Å, and for a star of V = 7.83 mag in 2400 s exposure, the signal-to-noise ratio is S/N = 150 at wavelength of λ = 6000 Å (provided by Xiaoling Yang and Yuqin Chen).

Zoom In Zoom Out Reset image size