This page describes the specifications of the volume-phase holographic grisms (VPHGs) procured for the ULTRASPEC run on EFOSC2 in January 2008. The VPHGs have been procured by myself at the University of Sheffield, but will be payed for by ESO. Therefore, once the 17-night ULTRASPEC run is over, the VPHGs will remain on La Silla and will become available to all EFOSC2 users.
The table below summarises the specifications of the two VPHGs for EFOSC2. The resolution depends on the slit used. Those quoted here are for a 1 arcsecond slit and have been calculated using equation A8 of Baldry et al (2004, PASP, 116, 403). Note that the parameter alpha_one in this equation refers to the angle of the beam within the prism, which is straightforward to derive from Snell's law (see figure 9 of Baldry et al). It is different to the "Angle of incidence" (or "Bragg angle") given in the quotation from Kaiser Optical, which refers to the angle within the fused-silica substrate of the VPH grating. Note that the quotation has an error, as it quotes an angle of incidence of 14.6o for the 656nm VPHG, whereas it should read 13.95o.
|VPHG #1||VPHG #2|
|Central wavelength (Angstroms)||6563||4750|
|Grating frequency (lines/mm)||1070||1557|
|Angle of beam within prism (alpha_one)||10.59o||10.85o|
|Dispersion (Angstroms/pixel, ULTRASPEC/CCD#40)||0.514/0.593||0.354/0.408|
|Wavelength range (Angstroms, ULTRASPEC/CCD#40)||6300-6826/5956-7170||4569-4931/4332-5168|
For comparison, the above specifications are similar to VPHG #1 and #4 proposed by the Golem team, including the prism material, prism angle and the grating frequency. Since our design was arrived at independently of Golem's, it lends additional confidence that we have arrived at an optimal solution. Note that our predicted resolutions are very slightly lower than those of the Golem team, but we believe this is probably due to adopting different values for alpha_one.
For those interested in Hbeta and CIII/NIII+HeII broad emission-line work, e.g. interacting binary star researchers, the wavelength range with the 475nm VPHG and ULTRASPEC ensures +/- 4500 km/s of spectrum either side of these lines, which should be more than adequate for most scientific programmes.
As well as far surpassing the existing EFOSC2 grisms in efficiency, the VPHGs described above deliver substantially higher resolution than is currently available. Grism #7 is closest in specification to VPHG#1, with a wavelength range on CCD #40 of 3270-5240 Angstroms, a resolution of 6.2 Angstroms, and a peak efficiency of 30%. Grism #18 is closest in specification to VPHG#2, with a wavelength range on CCD #40 of 4700-6770 Angstroms, a resolution of 6.3 Angstroms, and a peak efficiency of 35%.
Some useful information on the EFOSC2 spectrograph: collimator focal length = 322mm, with a focal ratio of 8.05. Camera focal length = 200mm (178.7mm), with a focal ratio of 4.96 (4.45), where the figures in parentheses do not include the field lens. The pupil diameter is 40mm.
Some useful information on the CCDs: The pixel sizes of the ULTRASPEC and ESO CCD #40 are 13 microns and 15 microns, corresponding to 0.136 and 0.157 arcseconds on the sky, respectively. The dispersion axes of the ULTRASPEC and ESO CCD #40 contain 1024 and 2048 pixels, respectively.
The VPH gratings designed by Kaiser Optical require an angle of incidence of approximately 14 degrees. To achieve this, but still work in a straight-through optical beam, it is necessary to place a prism above the VPH grating to deviate the light to the correct angle of incidence, and then another prism below the VPH grating to straighten the dispersed beam.
Using a low refractive-index glass would require a long prism to deviate the light by the required amount. Unfortunately, the space envelope available to the EFOSC2 grisms restricts the total length of the VPHG to 45mm (as shown by the mounting drawings), which means a high refractive-index glass is required. We selected Schott N-SF66, which turns out to be identical to the Ohara PBH71 glass selected by the Golem team. This is the highest refractive index glass currently available from Schott, and hence the VPHGs we have designed have the maximum resolution obtainable without resorting to crystalline prisms. Even with N-SF66, it was still necessary to truncate the edge of the prism to fit within the space envelope, but since the beam size of EFOSC2 is approximately 40mm and the clear aperture of the truncated prism is 45mm, this will not affect the performance of the VPHG.
To save money and time in the manufacture and AR-coating of the prisms, it was decided to use the same glass and the same angle for all four prisms. This compromised the throughput somewhat at the blue end of the 475nm VPHG, as shown in the transmission curves of N-SF66, although the efficiency of the resulting VPHG will still be far higher than the current EFOSC2 grisms in this wavelength range (and the resolution is much higher).
Note that, although a higher resolution could be obtained using crystalline prisms, such as ZnS (which has a refractive index of approximately 2.5 compared to the 1.9 of N-SF66), it is more difficult to polish such prisms and the resulting surface irregularities could be a source of scattered light.
Kaiser Optical kindly provided ICOS with a surplus VPH grating with which to test their alignment procedures prior to assembly of the final VPHG. Kaiser Optical have also agreed to place a fiducial mark on the edge of the VPH grating indicating the orientation of the "rulings" to aid alignment. I visited ICOS on January 4th 2008 to agree upon an alignment procedure, which is identical to the one described in this paper by some of the Golem team.
Once aligned, the prisms will be bonded to the VPH grating using Norland Optical Adhesive 61 (NOA61), details of which can be found here. We decided not to use the EpoTek 301-2, as recommended by Kaiser Optical in this email, as it takes a full working day to set and might become slightly soft if the dome interior becomes too warm, compromising the alignment of the VPHG assembly. One problem of using NOA61, however, is that it requires UV illumination to set, and N-SF66 is opaque to these wavelengths. ICOS experimented with UV illumination from the side, so it does not have to shine through the prisms, and this seemed to set the NOA61 satisfactorily on a test piece. Note that it is not possible to AR coat the bonded faces of the prisms, as it would compromise the performance of the adhesive, so only the outer faces of the prisms are AR coated.
The links below show technical drawings of the prisms and the complete 656nm VPHG assembly. Note that the details for the 475nm VPHG are identical.
Note that it is critical to get the grating "rulings" the correct way round. They should be aligned as shown by the dashed lines in the VPH grism assembly drawing above, i.e. coming out of the plane of the paper. This causes the prisms to act as beam steerers, as required. Mounting the grating so that the rulings lie parallel to the plane of the paper would cause the prisms to act as cross dispersers.
The completed VPHG assembly will be mounted in an EFOSC2 grism barrel manufactured by ICOS. A thin layer of silicone sealant will be injected into the small gap between the inner wall of the barrel and the VPHG assembly, which in conjunction with locking sleeves in the barrel will hold the VPHG in place without placing undue stress on the optics due to thermal variations. Fine adjustment of the tilt and rotation of the VPHG can then be achieved by adjusting the position of the barrel in the EFOSC2 grism wheel. The links below show a photograph and technical drawing of the barrel:
The space above and below the grism wheel in EFOSC is severely restricted, limiting the length of the VPHG assembly. The links below show a photograph and schematic of the space around the EFOSC2 grism wheel:
This email contains advice from Hans Dekker on VPHG procurement. Following this advice, we procured the VPH gratings from Kaiser Optical Systems Inc. in the US. My contact person there is Jim Arns (email@example.com, tel: (734) 665-8083 x339). The completed gratings, which are sandwiched between fused-silica windows, will then be sent to IC Optical Systems Ltd. in the UK. My contact person there is Chris Pietraszewski (firstname.lastname@example.org, tel: +44 (0) 20 8778 5094). ICOS will be making the prisms, aligning and bonding them to the VPHG gratings, AR-coating the outer faces, making the barrels, and then mounting the VPHG assemblies in the barrels.
To complete the VPHGs in time for our run, which begins on 27 January 2008, I hand-delivered the VPH gratings to ICOS as soon as I received them from Kaiser Optical just before Christmas 2007. In the meantime, ICOS had manufactured the prisms and AR-coated them in the Isle of Man. Final alignment, bonding and mounting will take place in the second week of January, and the expected delivery date is in the week ending January 18th. ULTRASPEC is shipped to Chile on January 11th, so I will have to take the completed VPHGs in my hand luggage when I leave for Chile on January 22nd.
Below are copies of the quotations:
Below are various datasheets and performance curves for the components in the VPHG: