Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Paschen and Balmer hydrogen lines correlation in Be stars

  • Please log in to reply
13 replies to this topic

#1 mwr

mwr

    Mariner 2

  • -----
  • topic starter
  • Posts: 295
  • Joined: 12 May 2018

Posted 14 November 2020 - 06:57 AM

Beside gamma Casiopeiae (https://www.cloudyni...uvex-vs-sa-200/), I have now recorded some additional spectra of bright Be stars using the SA-200 Grism:

 

eta Tauri (Alcyone)

phi Persei

 

A comparison of the spectra reveals some striking differences, especially in the NIR region of the spectrum:

 

 

It seems that the presence of Paschen lines in emission is somehow linked to the emission of neutral oxygen (O I) and to the emission of H alpha and beta. Is there a rule or an astrophysical law that correlates the Paschen and Balmer line shapes (emission or absorption, equivalent width) in Be stars? How are these findings interpreted?  A cursory look into the literature (Gray and Corbally) has given no answer so far (six emission classes of Be stars are described, but the astrophysical explanations are not given). 

 

I have also tried to find some NIR spectra of Alcyone and phi Persei in ProAm databases (BAA, AAVSO and BeSS) but so far without success. Most of the ALPY 600 spectra terminate at about 7500 Angström. Is the NIR region not well accessible by the APLY setup or is it just not of interest ?

 

 


  • RobboK and descott12 like this

#2 Organic Astrochemist

Organic Astrochemist

    Apollo

  • *****
  • Posts: 1,074
  • Joined: 10 Jan 2015

Posted 14 November 2020 - 02:35 PM

Great spectra and great questions. I think that as UVEX becomes popular you will see a lot more spectra like yours.
ALPY has a fixed (short) backfocus from the grism to the camera so there definitely is a limit on how deep in the NIR it can go. I am nowhere near that limit because I have a small (inexpensive?) ASI178MM Camera. Here's Gamma Cas.
gammacas_20191111_177_Jim Ley.png

For most chemists, I think the composite nature of these spectra is disorienting. On earth, we put a sample in a machine and obtain a spectrum. We put starlight from Beta Mon A in our telescope and we get a spectrum.
Beta Mon A full.png

On earth, with our sample at equilibrium, it isn't possible to observe both an emission peak and an absorption trough at the same time, from the same transition (at the same wavelength). But look closer at H-Beta in Beta Mon A and we see an emission peak inside of a wider absorption peak. What is going on?
Beta Mon A zoom.png

Even the continuum comes from different places. So the far UV continuum is closer to the core and the visible continuum comes from further out, where it's cooler. Above the visible photosphere the hydrogen is even cooler, so many more electrons are in n=2 level (and cause Balmer absorptions) than in the n=3 level (that cause Paschen absorptions). That's why Balmer absorptions are (usually) deeper in these B-type stars. In Be stars there is also hydrogen in the decretion disk. This hydrogen is moving much slower (has a lower velocity dispersion) than in the star, so this absorption peak is narrower. I think in Beta Mon A you can see this composite in some higher Balmer lines with wide absorption at the top and narrower absorption at the bottom.

From the UV photosphere there is a continuum of light including at 102.5 nm (Ly-ß). These photons could cause an electron to go from n=1 to n=3, but in a normal B star, closer to the star, where the H density is high enough for significant absorption, it's too hot and most H atoms are at n=2. Further out there are neutral H atoms, but the density is too low. Net result is that these Ly-ß photons come streaming out of B stars. However, when they hit the denser hydrogen in the cool Be disk, they excite electrons to n=3, which can decay to n=2 and produce H-α emission (by fluorescence). Other wavelengths (and mechanisms) can also cause H-α. Disk density and inclination angle also affect the observed H-α. More energetic photons are needed for the other emission lines in the Balmer series. Less energetic photons are needed for Paschen emission (these will be more easily absorbed by the star and thus cause less emission).

It just happens that oxygen, O I, has a transition at 102.6 nm, so O I in the disk could absorb Ly-ß and re-emit at 844.6 nm (by fluorescence). If a Be star has the disk density and inclination angle favorable for Ly-ß to produce H-α, it should also produce O I at 844.6 nm. Ly-ß could also cause emission at [O I] 630.0 nm and 634.6 nm (by fluorescence), but the density would have to be much lower to avoid collisions so this doesn't happen in Be disks.

At least this is my understanding so far.

Edited by Organic Astrochemist, 14 November 2020 - 05:34 PM.

  • mwr likes this

#3 Organic Astrochemist

Organic Astrochemist

    Apollo

  • *****
  • Posts: 1,074
  • Joined: 10 Jan 2015

Posted 14 November 2020 - 03:50 PM

OK Here I’m going out on a limb to explain your observation of H-alpha in Alcyone.

See figure 5.
https://arxiv.org/pdf/0905.1295.pdf

(Waving my hands) The general idea to me seems to be that if you have a sufficient density (rho) you will get enough cooling to allow the abovementioned fluorescence from ground state hydrogen. If the density falls below some critical value the temperature would be too hot for the ground state to be significantly populated. H-alpha emission could still be possible but not from those copious Lyman β photons pumping electrons directly to the n=3 level.

Edited by Organic Astrochemist, 15 November 2020 - 01:07 PM.

  • mwr likes this

#4 robin_astro

robin_astro

    Apollo

  • -----
  • Posts: 1,224
  • Joined: 18 Dec 2005

Posted 14 November 2020 - 07:22 PM

Most of the ALPY 600 spectra terminate at about 7500 Angström. Is the NIR region not well accessible by the APLY setup or is it just not of interest ?

With the ALPY the position of the spectrum is fixed so how far you can get at the IR end depends on the width of the camera sensor. With my ATIK428 I can only get up to ~7850A (Also the optics are optimised for the blue end so the focus is not so good in the IR.)

 

 It is impressive that you can get so far with the Star Analyser. Is there any sign of 2nd order contamination with these Be stars which are so bright at the blue end. (Repeats of the Balmer absorption lines at twice the wavelength?) Normally an orange long pass filter is recommended when working in the IR.

 

The LHIRES can be used in the IR with gratings 1200 and below but the achromatic optics limit sharpness. Peter Somogyi works in this area quite a bit, for example recently nova Cas 2020

http://www.spectro-a...&p=14708#p14708

 

The LISA has an IR option but you don't see it used much as you have to swap over various components.

 

Cheers

Robin 


  • mwr likes this

#5 mwr

mwr

    Mariner 2

  • -----
  • topic starter
  • Posts: 295
  • Joined: 12 May 2018

Posted 15 November 2020 - 12:56 AM

From the UV photosphere there is a continuum of light including at 102.5 nm (Ly-ß). 

It just happens that oxygen, O I, has a transition at 102.6 nm, so O I in the disk could absorb Ly-ß and re-emit at 844.6 nm (by fluorescence). If a Be star has the disk density and inclination angle favorable for Ly-ß to produce H-α, it should also produce O I at 844.6 nm. Ly-ß could also cause emission at [O I] 630.0 nm and 634.6 nm (by fluorescence), but the density would have to be much lower to avoid collisions so this doesn't happen in Be disks.
 

Many thanks for your clear explanations! Sometimes a chemist is needed who can explain astrophysics in an understandable way to another chemist ;-) The effects of the Lyman-ß radiation were new to me and I better understand now what is going on in complex Be star systems. Can you recommend a good review article on this topic?



#6 mwr

mwr

    Mariner 2

  • -----
  • topic starter
  • Posts: 295
  • Joined: 12 May 2018

Posted 15 November 2020 - 01:27 AM

Is there any sign of 2nd order contamination with these Be stars which are so bright at the blue end. (Repeats of the Balmer absorption lines at twice the wavelength?) Normally an orange long pass filter is recommended when working in the IR.

 

 

Yes, I can detect a weak 2nd order contamination if the spectrum is deliberately overexposed. I have used a Dulux energy saver lamp (setup is described here https://www.cloudyni...opy/?p=10086818) to verify this:

 

 

The second order contamination lines reveal themselves by their "false" color (violet lines in the pinkish NIR region). However, they are quite weak (blazed grating) and can be neglected in non-overexposed spectra (at least I was not able to detect second order lines in the Paschen region in a spectrum of Vega: https://www.cloudyni...opy/?p=10197694)

 

It would be interesting to combine the SA-200 with a SA-100 in perpendicular orientation as a cross-disperser to get rid of such a contamination (a low cost Echelle). Has somebody tried this before?



#7 Organic Astrochemist

Organic Astrochemist

    Apollo

  • *****
  • Posts: 1,074
  • Joined: 10 Jan 2015

Posted 15 November 2020 - 02:05 AM

The idea of the Balmer continuum (that spans the wavelength range that includes Lyman emission lines like Lyman β) contributing to the Be star emission lines is attested to in the older literature, but when I searched for those terms and not much came up recently. (For or against, really)

This more recent paper suggested
“ Stellar Balmer radiation plays a definitive role in the radiation field of the disk. The conversion of the stellar Balmer radiation into the longer wavelength continuous radiation and the line emission is the main process in the radiation field of the Be star disk. The stellar Lyman continuum contributes only to ionization in the outer envelope of the disk, where the disk is optically thin towards the central star.”

But they also say :”However, we could not obtain the large Hα emission strength observed in Be stars.”

https://academic.oup...0/4/749/1397368

So perhaps this is an active research topic...

Edited by Organic Astrochemist, 15 November 2020 - 12:19 PM.


#8 Organic Astrochemist

Organic Astrochemist

    Apollo

  • *****
  • Posts: 1,074
  • Joined: 10 Jan 2015

Posted 15 November 2020 - 02:23 AM

Here’s an article that suggests than Lyman β fluorescence is important in the IR spectroscopy of Be stars.

https://arxiv.org/pdf/1204.1496.pdf

The concept of optical depth is critical to understanding astronomical spectroscopy. I certainly don’t have the intuition that the professionals do. My introduction was via

https://www.amazon.c...05424953&sr=8-1
  • mwr likes this

#9 mwr

mwr

    Mariner 2

  • -----
  • topic starter
  • Posts: 295
  • Joined: 12 May 2018

Posted 15 November 2020 - 03:32 AM


The concept of optical depth is critical to understanding astronomical spectroscopy. I certainly don’t have the intuition that the professionals do. My introduction was via

https://www.amazon.c...05424953&sr=8-1

Thanks for your recommendations!

 

I have seen that you and a colleague have received some funding a couple of years ago for a interdisciplinary project to get college students into spectroscopy. May be this explains your didactic skills in this field. I would appreciate it if you could give a brief resumée of this project here or in an extra thread. 



#10 robin_astro

robin_astro

    Apollo

  • -----
  • Posts: 1,224
  • Joined: 18 Dec 2005

Posted 15 November 2020 - 07:46 AM

 

 

It would be interesting to combine the SA-200 with a SA-100 in perpendicular orientation as a cross-disperser to get rid of such a contamination (a low cost Echelle). Has somebody tried this before?

Yes but the perpendicular dispersion is  rather extreme at this low resolution.  The wedge prism can be used to separate the orders, rotating it by 90 deg.  Of course you would then lose the advantage of the extra sharpness of the grism setup though. Perhaps a setup using a wedge prism with a larger angle rotated at a suitable angle to to the dispersion direction would be optimum ? 

 

Cheers

Robin


Edited by robin_astro, 15 November 2020 - 07:47 AM.


#11 robin_astro

robin_astro

    Apollo

  • -----
  • Posts: 1,224
  • Joined: 18 Dec 2005

Posted 15 November 2020 - 07:54 AM

The wedge prism can be used to separate the orders, rotating it by 90 deg.  

Buil shows an example of this in tip #3 here

 

http://www.astrosurf...3/userguide.htm

 

Robin



#12 mwr

mwr

    Mariner 2

  • -----
  • topic starter
  • Posts: 295
  • Joined: 12 May 2018

Posted 15 November 2020 - 09:22 AM

Yes but the perpendicular dispersion is  rather extreme at this low resolution.  The wedge prism can be used to separate the orders, rotating it by 90 deg.  Of course you would then lose the advantage of the extra sharpness of the grism setup though. Perhaps a setup using a wedge prism with a larger angle rotated at a suitable angle to to the dispersion direction would be optimum ? 

 

Cheers

Robin

I will try both approaches. Thanks for the hints!


  • B McCandless likes this

#13 robin_astro

robin_astro

    Apollo

  • -----
  • Posts: 1,224
  • Joined: 18 Dec 2005

Posted 15 November 2020 - 09:43 AM

The FLOYDS spectrographs on the Faulkes telescopes are an example of a professional low resolution two order echelle arrangement (Here using a 235 l/mm grating with a prism as a cross disperser.)

https://lco.global/o...ruments/floyds/

I played around with some of the raw images a few years back when they were trying to get them up and running and the data reduction pipeline working. (The curve in the spectrum and severe fringing in the IR were problems I remember) They are working now though and I see their spectra pop up among the supernova classifications I monitor eg

https://wis-tns.weiz.../object/2020tld

 

Cheers

Robin


  • B McCandless and mwr like this

#14 mwr

mwr

    Mariner 2

  • -----
  • topic starter
  • Posts: 295
  • Joined: 12 May 2018

Posted 02 February 2021 - 04:45 PM

Yes but the perpendicular dispersion is  rather extreme at this low resolution.  The wedge prism can be used to separate the orders, rotating it by 90 deg.  Of course you would then lose the advantage of the extra sharpness of the grism setup though. Perhaps a setup using a wedge prism with a larger angle rotated at a suitable angle to to the dispersion direction would be optimum ? 

 

Cheers

Robin

I have tried to use the SA-100 as a "cross disperser" to eliminate the weak 2nd order contamination in the near infrared. The spectrum of a DULUX energy saver lamp was used as a test spectrum:

 

Folie1.JPG

 

The SA-100 was placed behind the SA-200 grism using a filter drawer:

 

Folie2.JPG

 

This setup can be considered as a bizarre "low resolution echelle spectrograph" but it does indeed remove the 2nd order contamination in the NIR:

 

Folie3.JPG

 

Folie4.JPG

 

Replacing the SA-100 with a real transmission echelle grating would be a fun experiment to do. The ESO Faint Object Spectrograph has a similar setup http://adsabs.harvar...A&A...189..353D:

 

efosc.jpg

 

However, the question is: Is there an affordable transmission echelle grating available  on the market ??




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics