SHG slits from China: chrome on borosilicate glass, 7 micron by 13mm

[thesmiths]

I have been working with Maurice in Australia on various mounting designs for spectroheliograph slits. We began with the chrome on fused quartz design I had made some time ago. It was mentioned in a post a few months ago that chrome on borosilicate glass slits were available in China viewtopic.php?p=411342#p411342. I ordered a pair of them from Taobao, as did Maurice. The quality looked quite good under the microscope, only perhaps a little "rougher" and with a bit more debris than the ones I had made (this could have come from the transportation).

The design I have lately been using for my 15mm x 15mm quartz chips we adapted to the 10mm x 16mm glass chips. This helped to eliminate stray light that can come from the edges, particularly because the Chinese slit is quite long (13mm) and there is not much room at the ends. Also, because the chrome layer has no antireflection coating, there is likely to be some internal reflections inside the glass, which could eventually emerge from the edges.

The chrome on glass slits were attached with Torr Seal epoxy onto a 27mm diameter aluminium disk, which then fit into a 1.25-inch filter cell (as I have lately been doing with my chrome on quartz slits). This design reflects a large amount of sunlight before it ever hits the glass/quartz, reducing the energy load. Borosilicate glass is much more robust that soda lime glass and we found this design could be used with no ND filter to attenuate the sunlight. We just used a 2-inch UV/IR filter at the telescope focuser (the main purpose being to extend the life of the chrome).

The last two images below were taken by Maurice from Australia at the Ca-K and H-alpha wavelengths. The quality is very good. Since the width is 7 micron (vs 9 micron for my quartz slits and 10 micron for the standard Shelyak slit), the spectral resolution seems a little higher, although the light throughput is, as expected, somewhat lower. So there are trade-offs with a reduced slit width.

The chrome layer on the Chinese slits is said to be 400nm, which is quite thick. I noticed this led to better opacity in the red and infrared, which is an advantage. But using a thicker metal layer will typically lead to rougher lines during the photolithographic lift-off process. The chrome/chrome oxide layer that I had used was around 200nm thick, which is actually optimised from opacity near the UV, but which has extremely low line edge roughness. So there are trade offs on this parameter as well.

All in all, I think the 7 micron chrome on borosilicate glass slits are a very viable alternative for SHG use and narrowband spectroscopy in general. I spent some time looking at these slits before placing my recent order for a new batch of chrome/chrome oxide on fused quartz slits (the same material I used before). These will be 8 micron wide and 13mm long (vs 9 micron by 12mm previously) and should be ready in a few weeks time.

7 micron slit 16 x 10mm x 2mm thick glass slit length 13mm.jpg (425.97 KiB) Viewed 744 times

7 micron slit 16 x 10mm x 2mm thick glass.jpg (64.28 KiB) Viewed 744 times

narrow-slit-1.jpg (113.73 KiB) Viewed 744 times

2091.jpg (92.81 KiB) Viewed 744 times

2092.jpg (87.65 KiB) Viewed 744 times

2095.jpg (124.2 KiB) Viewed 744 times

2094.jpg (128.61 KiB) Viewed 744 times

Ca-K SHG image with Chinese slit

260124_CaK_6frames_full.jpg (1.04 MiB) Viewed 744 times

H-alpha SHG image with Chinese slit

280324_HA_9frames small.jpg (761.92 KiB) Viewed 241 times

[marktownley]

Looks promising!

[thesmiths]

I actually got these slits in order to try the infrared Calcium lines again. I found my slits gave too much background light at wavelengths above H-alpha since the transmission of thin metal layers increases with wavelength.

[minhlead]

I have been working with Maurice in Australia on various mounting designs for spectroheliograph slits. We began with the chrome on fused quartz design I had made some time ago. It was mentioned in a post a few months ago that chrome on borosilicate glass slits were available in China viewtopic.php?p=411342#p411342. I ordered a pair of them from Taobao, as did Maurice. The quality looked quite good under the microscope, only perhaps a little "rougher" and with a bit more debris than the ones I had made (this could have come from the transportation).

The design I have lately been using for my 15mm x 15mm quartz chips we adapted to the 10mm x 16mm glass chips. This helped to eliminate stray light that can come from the edges, particularly because the Chinese slit is quite long (13mm) and there is not much room at the ends. Also, because the chrome layer has no antireflection coating, there is likely to be some internal reflections inside the glass, which could eventually emerge from the edges.

The chrome on glass slits were attached with Torr Seal epoxy onto a 27mm diameter aluminium disk, which then fit into a 1.25-inch filter cell (as I have lately been doing with my chrome on quartz slits). This design reflects a large amount of sunlight before it ever hits the glass/quartz, reducing the energy load. Borosilicate glass is much more robust that soda lime glass and we found this design could be used with no ND filter to attenuate the sunlight. We just used a 2-inch UV/IR filter at the telescope focuser (the main purpose being to extend the life of the chrome).

The last image below was taken by Maurice from Australia on Jan 26 at the Ca-K wavelength. The quality is very good. Since the width is 7 micron (vs 9 micron for my quartz slits and 10 micron for the standard Shelyak slit), the spectral resolution seems a little higher, although the light throughput is, as expected, somewhat lower. So there are trade-offs with a reduced slit width.

The chrome layer on the Chinese slits is said to be 400nm, which is quite thick. I noticed this led to better opacity in the red and infrared, which is an advantage. But using a thicker metal layer will typically lead to rougher lines during the photolithographic lift-off process. The chrome/chrome oxide layer that I had used was around 200nm thick, which is actually optimised from opacity near the UV, but which has extremely low line edge roughness. So there are trade offs on this parameter as well.

All in all, I think the 7 micron chrome on borosilicate glass slits are a very viable alternative for SHG use and narrowband spectroscopy in general. I spent some time looking at these slits before placing my recent order for a new batch of chrome/chrome oxide on fused quartz slits (the same material I used before). These will be 8 micron wide and 13mm long (vs 9 micron by 12mm previously) and should be ready in a few weeks time.


7 micron slit 16 x 10mm x 2mm thick glass slit length 13mm.jpg
7 micron slit 16 x 10mm x 2mm thick glass.jpg
narrow-slit-1.jpg
2091.jpg
2092.jpg
2095.jpg
2094.jpg
260124_CaK_6frames_full.jpg

Hi. I am getting the same slit and trying to adapt it to my solex build. I have a few questions:
1. How long is the light mask you used? I can see that the opening is a bit shorter than the slit ?
2. How do you mount the slit, with the face with the chrome coat (and the slit) facing the telescope or the face with the naked glass facing the scope?

[thesmiths]

I am getting the same slit and trying to adapt it to my solex build. I have a few questions:
1. How long is the light mask you used? I can see that the opening is a bit shorter than the slit ?
2. How do you mount the slit, with the face with the chrome coat (and the slit) facing the telescope or the face with the naked glass facing the scope?

I think I used a slot of 13mm x 2mm. I thought it was important to use a fairly wide slot to avoid the problem of the edges potential causing a shadow (the light does not come exactly perpendicular but rather with a light cone). 13mm is already much longer than can really be used and is already quite close to the edges of the glass.

The standard SHG configuration is with the chrome layer towards the grating (on the opposite side to the telescope).

I would order at least 2 pieces as I noticed that quality control was not the highest. There is also a fairly high postage charge for shipping outside of China so may as well order a few at once.

[minhlead]

I am getting the same slit and trying to adapt it to my solex build. I have a few questions:
1. How long is the light mask you used? I can see that the opening is a bit shorter than the slit ?
2. How do you mount the slit, with the face with the chrome coat (and the slit) facing the telescope or the face with the naked glass facing the scope?

I think I used a slot of 13mm x 2mm. I thought it was important to use a fairly wide slot to avoid the problem of the edges potential causing a shadow (the light does not come exactly perpendicular but rather with a light cone). 13mm is already much longer than can really be used and is already quite close to the edges of the glass.

The standard SHG configuration is with the chrome layer towards the grating (on the opposite side to the telescope).

I would order at least 2 pieces as I noticed that quality control was not the highest. There is also a fairly high postage charge for shipping outside of China so may as well order a few at once.

Thanks. I can have the slits and grating/lens delivered to VN
(Where I am) relatively cheap. I find several lens/grating/slit provider with quite cheap prices in China and intending testing them out. I also thinking about building a solex/starex optical kit that is much cheaper than what provided by Shelyak (the solex optical kit is currently about 300$ and StarEx is about 450$).

[thesmiths]

I took some microscope images of a Chinese borosilicate slit and one of my quartz slits. I used a metallurgical microscope, which images with a combination of transmission (green light) and reflection (white light). The very small circular spots are dust on the camera sensor and the faint "doughnuts" are dust somewhere in the microscope system.

As I mentioned above, the Chinese slits, although of good quality, do appear to have rougher edges and the surfaces are in general not as clean. The reflected light is a gold/yellow colour.

The rougher edge the Chinese-made slit probably has to do with the lift-off process used. The chrome layer is fairly thick, which has an advantage with regards to opacity, particularly at longer wavelengths. But thicker films will tend to "tear" during lift-off, resulting in a rougher edge.

The chrome / chrome-oxide layer that I used is much thinner and therefore less opaque (particularly in the IR). But the thinner layer will give better lift-off and therefore less edge roughness. The reflected light is a very dark blue.

I'm guessing that the Chinese slits were designed for use with lasers, which often have wavelengths in the red and infrared. For lasers applications, edge roughness is typically not a critical parameter.

The slits I had designed were made using a technique to make semiconductor wafer masks, in which line edge roughness is critically important (as is cleanliness). Also, semiconductor masks are generally used near the UV so they are intentionally opaque at short wavelengths (and less so at longer wavelengths).

For SHG, line edge quality is fairly important, although many defects can be removed with an appropriate filter (transversalium removal). It would be desirable to have more opacity in the red and infrared, but there is always a trade-off in these matters.

Chrome slit on borosilicate glass (from China), 7 micron width.

0872.jpg (133.53 KiB) Viewed 345 times

Chrome / chrome oxide slit on fused quartz (from the UK), 9 micron width.

0865.jpg (83.82 KiB) Viewed 345 times

[thesmiths]

1. How long is the light mask you used? I can see that the opening is a bit shorter than the slit ?
2. How do you mount the slit, with the face with the chrome coat (and the slit) facing the telescope or the face with the naked glass facing the scope?

I looked back at the specifications for the aluminium holder that I used for the Chinese slit. In fact, the slot was 2.5mm wide and 12mm long. I normally would have used 2mm wide but I made it wider since I had them made before I actually received the slits, therefore I didn't know how well centred they would be. I also had the slot shorter than the length of the slit in order to stay away from the ends of the rectangular glass piece. For my own slits, I use a square substrate and put the slit diagonally.

I got around to using one of the Chinese slits myself this week. I didn't find any major problems with them. They had more transversalium than the ones I had made, but nothing too serious. They were definitely more opaque at longer wavelengths due to the thicker chrome layer (see viewtopic.php?t=45048 for some measurements in the infrared).

I also did a quick test with H-alpha (see image below). I used a 1800 l/mm grating so the contrast is not as high as it would be with 2400 l/mm. I actually used no UV/IR filter or ND filter for attenuation while imaging at H-alpha and the slit survived. I did this test in the afternoon when seeing conditions were not ideal so the quality is not that great; I also only stacked 8 frames. With no attenuation, I was able to use a 1ms exposure at zero gain and could achieve 484fps (the ROI was 3096x100).

H-alpha, 7 micron by 13mm slit, 1800 l/mm grating, Skywatcher 80ED doublet telescope. Stack of 8 frames.

Ha-80mm-1800g-070524.jpg (852.79 KiB) Viewed 218 times

[minhlead]

1. How long is the light mask you used? I can see that the opening is a bit shorter than the slit ?
2. How do you mount the slit, with the face with the chrome coat (and the slit) facing the telescope or the face with the naked glass facing the scope?

I looked back at the specifications for the aluminium holder that I used for the Chinese slit. In fact, the slot was 2.5mm wide and 12mm long. I normally would have used 2mm wide but I made it wider since I had them made before I actually received the slits, therefore I didn't know how well centred they would be. I also had the slot shorter than the length of the slit in order to stay away from the ends of the rectangular glass piece. For my own slits, I use a square substrate and put the slit diagonally.

I got around to using one of the Chinese slits myself this week. I didn't find any major problems with them. They had more transversalium than the ones I had made, but nothing too serious. They were definitely more opaque at longer wavelengths due to the thicker chrome layer (see viewtopic.php?t=45048 for some measurements in the infrared).

I also did a quick test with H-alpha (see image below). I used a 1800 l/mm grating so the contrast is not as high as it would be with 2400 l/mm. I actually used no UV/IR filter or ND filter for attenuation while imaging at H-alpha and the slit survived. I did this test in the afternoon when seeing conditions were not ideal so the quality is not that great; I also only stacked 8 frames. With no attenuation, I was able to use a 1ms exposure at zero gain and could achieve 484fps (the ROI was 3096x100).


Ha-80mm-1800g-070524.jpg

If you were using 1ms exposure, how fast did you scan? Wouldn't that take something like 64x scan speed?

[thesmiths]

With no attenuation, I was able to use a 1ms exposure at zero gain and could achieve 484fps (the ROI was 3096x100).

If you were using 1ms exposure, how fast did you scan? Wouldn't that take something like 64x scan speed?

It is the fps that is the important number, not the exposure. Of course 1/exposure will set the maximum fps, but the ultimate limit is determined by the USB3 interface bandwidth. The pixel radius of the Sun was approximately 1160 and the Y/X ratio was 1.0. I believe the scan speed was around 26x.

[stefanoaz]

Douglas: Did you see significant transversalium artifacts in the raw image from this slit?

[minhlead]

Douglas: Did you see significant transversalium artifacts in the raw image from this slit?

I have none on mine.

With no attenuation, I was able to use a 1ms exposure at zero gain and could achieve 484fps (the ROI was 3096x100).

If you were using 1ms exposure, how fast did you scan? Wouldn't that take something like 64x scan speed?

It is the fps that is the important number, not the exposure. Of course 1/exposure will set the maximum fps, but the ultimate limit is determined by the USB3 interface bandwidth. The pixel radius of the Sun was approximately 1160 and the Y/X ratio was 1.0. I believe the scan speed was around 26x.

I got it. Your image also has very little "shearing artifact". Wonder if the fast exposure speed has that effect? Great capture.

[thesmiths]

Your image also has very little "shearing artifact". Wonder if the fast exposure speed has that effect?

By "shearing artefact" I think you mean the jagged edges of the Sun? This tends to be caused by atmospheric turbulence. You're right, there is not so much evidence of this, even though it was taken around 12 noon, which tends to have a lot of turbulence. Also, there were not so many frames stacked (only 8); more frames also tend to reduce the jaggedness.

So, yes, high fps and high scan rate will tend to minimise this artefact by pushing the data rate to higher frequencies (since the turbulence is at low frequencies). The fast exposure is a necessary parameter to increase the data rate but there are the other electronic and mechanical requirements to achieve high data rate acquisition.

Maybe you can also do some more tests with the Chinese slits to see if they can be used long-term with no attenuation.

By the way, if you put a second slotted mask (maybe made of black anodised aluminium) on the side facing the grating, you can probably reduce a lot of the internal reflections.

[thesmiths]

Did you see significant transversalium artifacts in the raw image from this slit?

I did not see much more transversalium in the raw data. I can present some quantitative information on this, but I would avoid trying to read too much into it. The charts below show the "transversalium correction" factor that is applied by our SHG program. It gives some idea of the defects on the slit. The first one shows the data from a Chinese 7 micron slit using my 80ED. The second one shows H-alpha with one of my 9 micron slits, using the same 80ED telescope but some of the other optical components were different (the grating and the focal length of the collimator). The ability to resolve the transversalium also depends on the focusing of the optics. Both slits had one single "strong" defect and then a number of very small ones. The "strong one" you could see by eye in the video image on the screen if you know what to look for. The smaller ones could be seen in the final raw image if there were no correction.

H-alpha, 7 micron slit.

111911_070524_shift=0_transversalium_correction.png (140.2 KiB) Viewed 180 times

H-alpha, 9 micron slit.

073610_300424_shift=0_transversalium_correction.png (132.66 KiB) Viewed 180 times

[y3000]

the Chinese seller released fused quartz slits now , same size, OD4 , double price .

https://item.taobao.com/item.htm?id=798031165940

2024-05-20 123235.png (196.36 KiB) Viewed 49 times

[thesmiths]

If it's only OD4, it might end up being too transparent near H-alpha. The previous slits were very opaque. Also, the blue colour means it will probably absorb a lot of infrared light. The previous ones were highly reflective. Someone would need to test them to see if they are suitable. I found the borosilicate glass was quite ok. Fused quartz is actually much easier to chip.