A long time ago I did a PhD in surface chemistry at the University of Durham. My professor (Prof Jas Pal Badyal), is still there and still doing great research. Some of my equipment is still in the lab, and I was lucky enough to visit there earlier this year and see some of the work the team are doing. During my PhD, some experiments were done which produced results which I never fully answered. However with my new interest in microscopy it has helped me look at my PhD again with different eyes, and cast light on some of my observations from over 25 years ago. Today I’d like to share these with you.

The first was purely a few nice images from a microscope slide I bought. During my PhD I worked with precious metal compounds. I’d dissolve them up in different solvents, spin coat them on to different substrates and then hit them with a cold gas plasma to reduce them down to metal layers. The goal of this was to find new ways to make catalysts on temperature sensitive substrates. At the time, I’d often get nice complex crystal structures and a few of these even made it into my PhD as rather poor quality photos. The microscope slide I got recently was of potassium platinum cyanide (a platinum compound) on glass. Under crossed polarized light, which is often used for the imaging of crystals, this is what it looks like.

The image above was taken using an 1x Olympus SPlan FL1 objective, on my modified Olympus BHB microscope. The whole circle is about 6mm across. I removed the condenser and just used the field lens to cast light onto the sample given its size. Very often I’d get complex dendritic structures like this with silver and palladium compounds, and I struggled to photograph them with what I had to hand at the time.

Here’s some more images from this slide as it is so beautiful and complex. These were taken with a higher magnification objective (a Nikon 10x NA 0.5 Fluor). If only I’d had this microscope back in my PhD….

The second thing is a bit more of a technical one, and the purchase of a book gave me some answers I could have done with when I was back in the lab. This came about because of a chance observation one day. I used to deal with gold chloride (AuCl₃) solutions. These were spin coated on to Nylon, and the result was the Nylon would slightly dissolve, and the gold salt and Nylon would mix together to form a thin layer. Normally I would then treat this with a hydrogen plasma to make a gold metal layer. To analyse the samples I used a range of different techniques, one of which was X-ray Photoelectron Spectroscopy (XPS) which involves using X-rays to irradiate a sample, kicking out electrons which are then collected and analysed. The energy of the electrons tells you about which elements are present. I noticed that with some of my gold chloride samples, if I analysed them before treating them with the hydrogen plasma, they would come out of the device looking different where the X-rays had hit the sample. They would be a different colour, often red or brown. Sometimes this sample would then change over the next day or so and become more ‘gold’ looking. Essentially, the X-rays were doing something to the sample. Here’s some of my original notes from my lab books from 1996 (and yes I still have my PhD lab books, and no my hand writing has not improved over the years). The little ‘letter box’ shaped things in the text are actually the coated Nylon samples from the experiment – I stuck them in the lab book where possible.

In the second image above, one of the samples looks to be transparent red, and this gives you an idea of what they looked like (this one never changed over time). When I originally did the work I suspected this effect was due to the X-rays breaking down the gold chloride molecule and forming nano gold clusters or colloids suspended in the Nylon polymer matrix. Depending on the precise nature of the X-ray treatment with some of these samples the gold clusters then migrated over time to form larger structures and eventually more coherent gold films. However this was a distraction from the actual work, as my main focus was plasma treatment, and I was never able to prove what was going on.

A few months back, I came across a book called “Colloids and the Ultramicroscope” by Zsigmondy, and translated by Alexander in 1909. This was for sale in the US, and being a bit of a nerd I bought it as ultramicroscopy was a technique I was interested in reading about as it was an approach to produce very high resolution images. Here’s the book.

It turned out that he had used this technique to look at gold colloids, and amazingly the book had some tables and colour plates of gold colloid solutions with different sized particles.

It does indeed look like small gold colloids have that distinct red colour which I observed – the smaller the particle the more intense the red.

Very often as scientists our experiments throw up questions which we cannot answer at the time. Unfortunately, in today’s deadline driven world, these are often seen as problems – things which slow us down and distract from the desired goal. But the key driver for a scientist is exploring and hopefully explaining the unknown. Sometimes this happens quickly, but as in this case it can take years, and inspiration often comes from unexpected sources.

As always, thank you for reading, and if you’d like to know more about the work I do, I can be reached here.

One of the original reasons for moving to UV for microscopy was that of improved resolution – the shorter the wavelength, the better the resolution for a given numerical aperture. I’ve discussed this effect before (for example see here) but in my earlier work, I was limited to 313nm as my lower limit for wavelength. With my new work using 254nm I thought I would do a little test with one of the diatoms on my test slide and see how it looked as the wavelength changed. Today I am sharing those results.

For this test I used a 32x Zeiss Ultrafluar NA 0.4 objective. For now this is highest magnification objective I can use at 254nm. This was used without any glycerin immersion (it’s labelled as being for glycerine immersion, but can also be used without). Camera is a monochrome converted Nikon d850 from MaxMax. Light source for 546nm to 313nm was a mercury xenon lamp, and for 254nm a low pressure mercury lamp. Images are the full size, but reduced in resolution for sharing from the originals (which were nearly 50Mp). All images saved as RAW files and processed in Darktable, and all have had the same degree of sharpening. Simple brightfield images. Wavelength of light used shown in the bottom left of the images.

I’ll show them in the order going from long wavelength to short.

Even with the small images used here, the improvement in resolution as the wavelength is reduced is quite obvious. The diatom (a Pleurosigma angulatum) has features on it which are not visible when viewed with 546nm light with this objective – the dots or punctae. They are too small to be resolved with the NA 0.4 objective at 546nm. At 405nm they start to become faintly visible. They become more obvious at 365nm and even more so at 313nm. At 254nm, they might even be more apparent, but the nature of the image makes it harder to be certain. I also had to up the ISO for the 254nm image from 200 to 400, as the camera has virtually no sensitivity that far down. Even doing this it needed a 15 minute exposure at 254nm vs a 4 second one at 313nm. The silica based structure of the diatom also becomes darker at the shorter wavelengths, as it absorbs more of the light. This helps with contrast and is in itself a good reason for me to keep pursuing this technique.

There is however a big and obvious issue with the 254nm image – the heavy ring artifacts. I saw this to a lesser extent with the 10x objective, and I am not 100% certain as to the reason for it. One possibility is something called window etaloning, where incoming light is reflected from the front of the camera sensor, before being reflected back again from the underside of the sensor coverglass. It bounces back and forth and causes these light and dark bands (there’s more info about it here). This could be more of an obvious issue at 254nm as the reflectance of silicon used for the sensor is quite high down there (higher than the longer wavelengths). However with the 254nm light I do not have a diffuser on there, unlike the the light used for the other wavelengths, and I cannot rule that out as a possible contributing factor.

254nm UVC microscopy continues to present some very strong technical challenges, however the improved contrast and resolution for diatom imaging is obvious, and it is something I shall continue to work on as and when I get the chance. As always, thanks for reading, and if you’d like to know more about this or any other aspect of my work, I can be reached here.

A couple of weeks ago I posted my first microscope image captured at 254nm using my UV modified Olympus BHB microscope (see here). While it showed the potential of the technique, it highlighted a number of issues which still needed addressing – especially the lighting (making it stronger and more collimated) and the filtration of the light to remove unwanted wavelengths. This is a bit of an update on progress since then, mainly with regards to the lighting, but also a mention of a slight change to filtration of the light which is imaged.

The main issue with the lighting was that it is a light bulb, emitting light in all directions. This is very wasteful and I need a focused, collimated light source, wasting as little as possible. In the original lamp housing there was a glass condenser lens which could be moved in an out to provide some focusing. Glass is fine for visible light, and even maybe for 365nm depending on thickness and coatings, but is no good at all for 254nm as it would just absorb everythin. This means needing to use lenses made from UV fused silica, quartz or materials such as calcium fluoride. As you can imagine the availability of lenses in these materials is much more restrictive (and the cost much higher) when compared with glass. As a result I tend to have to choose off the shelf components, which are ‘close’ to the parameters of the original pieces, rather than ones which are exactly the same. The original glass condenser lens was about 42mm diameter, while off the shelf components were available in 25mm or 50mm diameter. I went with a 25mm UV fused silica aspheric condenser lens with a 20mm focal length from Thorlabs (see here). This was mounted in one of their SM1 tubes, which in turn was screwed in to an SM1 to M42 adapter. The M42 adapter was just the right diameter to fit into the lamp housing where the original lens did, and be held in with the original retaining clip. Bonus. This is where I give a ‘shout out’ to Thorlabs. As I’ve mentioned before, they are great to deal with and have an amazing range of optical equipment. Plus they send out Labsnacks with their orders….. This is how it looked once put together (you can see the lens in the middle of the adapter).

When I measured the intensity of the light at 10cm from the bulb with the lens in place I got about 6x more light than without the lens, which is a great improvement and will help with the imaging.

What do the images look like now? In a moment I’ll some examples from the fused silica/quartz diatom test slide, taken using a 10x Zeiss Ultrafluar NA 0.2 objective lens. The full sized images have been reduced is resolution for sharing here.

First though, a visible light (white LED) image of part of the slide with the same objective.

The same region of the slide now imaged at 254nm.

The modified lighting has helped a lot with getting an even light distribution across the image. It’s interesting that some of the diatoms seem to be absorbing quite a lot of light at 254nm while others do not (this is in keeping with some of the literature I’ve read on the subject). I am getting some rings in the lighting, which I think is down to the very reflective surface of the filters. I need to do a bit more work and see if that can be rectified.

A couple of other images of the slide at 254nm.

Finally, some crops from the image above, shown at the original pixel resolution.

Doing some math on the images, in the original un-cropped images, there were 11 pixels per micron. If you look at basic Abbe resolution calculation for this objective (λ / 2xNA) you’d get a theoretical resolution of 635nm for this objective at 254nm. Some of the features in the cropped images above are of the order of 7 or 8 pixels across or about 750nm, so similar to the theoretical resolution limit predicted by Abbe. Now this is a bit of a simplification, as I really should take into account the NA of the condenser underneath the stage (which would push the theoretical resolution to smaller features) and the MTF function of the camera and lens (which would push it back towards bigger features), but I’ve not had enough coffee for that yet.

The images above are a bit noisy. I had to use 30s exposure at ISO1000 as the camera sensitivity this far into the UV is low. I am hoping longer exposures at lower ISO will help with that in future, or perhaps I’ll try stacking multiple exposures if I figure out how to do that.

Before I wrap up this post, I want to briefly mention filtration of the light before it reaches the camera. In my original post I was using a UVC bandpass filter which came from a forensics camera I bought a few years back. While this had good blocking of longer wavelengths, it had relatively low UVC transmission (about 20%). I decided to get another UVC bandpass filter and stack the two together, the aim being to improve blocking of unwanted wavelengths. The new filter is a Semrock 260/16 Brightline one which I bought from Laser2000 here in the UK (another company I’ve had good experience with in the past). This isolates the 250-270nm region very well, and blocks the rest of the UV and visible light. However it does not block the IR. So far, the initial tests with the two stacked together look good, and I’ll talk more about this filter in a future post as the whole area of filtering light for imaging at 254nm warrants a bit more of a discussion than there is time to do here.

As always, thanks for reading, and if you’d like to know more about this or any other aspect of my work, you can reach me here.

I’ll leave you with my standard warning for anyone considering working with UVC – don’t do it unless you know what you are doing and are using the right safety gear and clothing. UVC can be extremely damaging to eyes and skin, and it isn’t worth taking the risk unless you have the correct safety procedures in place.

Bit of a short update today, but there will be more to come on this after I have done some more work on it. Last year I reported being able to photograph in the UVC at 254nm with some of my modified cameras (for example see here). When I was building my UV microscope I wanted to make sure that it would be usable that far into the UV, even though the logistics of actually making it happen would be challenging. Recently I got hold of a small UVC light source which could make this more feasible, and I wanted to share some initial results from it.

Firstly though, a word of warning. UVC at 254nm can be extremely damaging to the eyes and skin. Do not attempt any work like this without the proper safety equipment and knowledge of how to use it. You have been warned……

The camera I used was my MaxMax converted monochrome Nikon d850 which is one I’ve used previously for UVC imaging. The light source was a 3W UVC lamp (a bit more on that later) mounted in a spare Olympus lamp housing. The slide was the diatom slide that I had made using fused silica/quartz instead of glass, and the one I’ve been using for my imaging work at 313nm and 365nm. I used a 10x Zeiss Ultrafluar objective. The light was basically just put in the empty lamp housing and moved around until I could get an image with it. And this is a first image at 254nm….

As images go this perhaps is not as impressive as some others. But the key thing to remember here is that this was done using 254nm and captured on a (modified) high street camera. There are some issues, and these will need to be overcome with future work. The image isn’t pure UVC – the camera has so little sensitivity down there that there is some contamination from other wavelengths. Based on some tests done at the same time, about 90% of this image is UVC related and the rest isn’t. I have another filter on order which will probably be stacked with the one currently being used. Better blocking of unwanted wavelengths is needed. The light source was not focused at all, and a lot was lost. I’ll be building a UVC focusing system for the lamp over the next couple of months. Image acquisition times are long, very long. Part of this is due to lack of focusing on the lamp, but part is just down to lack of camera sensitivity. It’s a pain, as I cannot use live view to optimise the focus, and everything has to be done slowly. This image above was not the best focus, but was a proof of concept one. No simple answer to this, other than hopefully the focusing system that needs to be built.

What about the light and setup? The lamp is a 3W low pressure mercury one, and was driven by a DC powersupply (manufacturer claimed 10VDC, but mine worked between at about 11V). Here’s a couple of pictures which shows the lamp and how it was setup on the microscope.

Also, here’s the emission spectrum from the lamp, showing the strong mercury emission line at just under 254nm (measured using an Ocean Insight FX spectrometer at a distance of 10cm from the side of the lamp).

It’s early days for this experiment and hopefully there will be more news on it later in the year. For now though thanks for reading and if you’d like to know more about my work you can reach me here.

Quick update today. Always nice to see my work getting recognized. One of my UVB (313nm) diatom microscope images has been selected as the July 2022 cover image for the site amateurmicrography.net. Here’s the image.

The Amateur Micrography page links through to the Photomacrography forum which has some of the best macro and micro imagers from around the world, and is well worth checking out if you have a minute.

This is a bit of technical update post, as it covers something I wasn’t aware of before, which may be of interest to other microscopists. A few days ago, I was doing some work with an old 170x Leitz Q NA 0.5 reflecting objective (see here) and noticed that when the aperture of the condenser was closed right down I got a darkfield image, with a black background. This got me wondering whether it would work with other reflecting objectives, and whether it was possible to use this technique to get darkfield images in the UVB region at 313nm (something I previously couldn’t attempt as I don’t have a quartz darkfield condenser). This post gives and update to that work.

As a subject for imaging I chose the diatom slide which I’d had prepared using quartz slide and coverslip which I have been using for imaging at 313nm (examples here). Rather than use the 170x Leitz reflecting objective, which wasn’t designed for a microscope with a tube length like mine, I used a 15x Edmund Optics NA 0.28 reflecting objective. The condenser was my vintage Zeiss quartz one. With the condenser iris closed right down, imaging at 313nm, I got the following image of one of the diatoms on the slide.

By closing the iris on the condenser right down it does indeed produce a nice darkfield image. By using the quartz condenser and slide, and combining this with a mirror objective it was possible to create a darkfield image in the UVB region at 313nm, something which I hadn’t previously thought possible with the equipment I currently have. For the fellow optics geeks out there, here’s a picture of the objective.

Sometimes it is worth doing experiments just to see what will happen, as the results can be unexpected and open up new possibilities. So, don’t be afraid to experiment. As always thanks for reading, and if you’d like to know more, I can be reached here.

EDIT 3/7/23. Turns out this principle of using a smaller aperture in the condenser with a mirror lens to create a darkfield image has been reported before and is called ‘luminance contrast’. I just spotted mention of it in “Piper, J., & Chmela, G. (2011). Advanced Techniques for Observation and Photomicrography of Subcellular Structures in Diatom Shells. Microscopy Today, 19(1), 10-14. doi:10.1017/S1551929510001203“. They refer back to two other papers for more information, “J Piper, Microscopy Today 15(4) (2007) 26–34” and “J Piper, Int J Light Electron Opt 120(18) (2009) 963–75, doi: 10.1016/j.ijleo.2008.03.032“.

Ok, perhaps not the catchiest title for today’s post. However, it’s an unusual post. When you were young, like me, I presume there were things you were told not to do – “don’t grab stuff out of a hot oven without gloves”, “don’t charge up the cat with static electricity by stroking it for 5 minutes”, etc, etc. Of course, I often did things anyway, although to be fair I never did grab stuff from a hot oven. The objective used was most definitely not designed to be used in this way and would have fallen into the ‘don’t try this’ bucket. This builds on the imaging of the Synedra superba diatom slide I share recently (see here).

So, what is this odd objective? It’s a 170x Leitz Q NA 0.5 reflecting objective designed for use with 0.18mm thick coverslips and a 400mm tube length microscope. It’s rare and unusual, and there’s a little bit about it here in one of my previous posts. I’ll show a picture of it at the end of the post, but in summary this was designed for use between 220nm and 700nm and with its own specific reflecting condenser. It wasn’t designed to be used on a 160mm tube length microscope, or with an Olympus Aplanat Achromat condenser. Hence it was done just to see what type of image it would give. Imaging was done with 405nm light and using my monochrome converted Nikon d800.

What do the images from it look like? I first tried closing the condenser right down, and was a bit surprised to see it looked like a ‘dark field’ type of image, with a black background and the diatom showing bright white.

There was a surprising amount of detail here, given that this is being used way outside of the way in which it should be. Why does it look like a dark field image? My guess here is that when the condenser iris is closed right down the spot illuminating the sample becomes very small. There is a mirror in the middle of the objective (it is a reflecting design) and I guess this blocks the small spot of direct light so the only light that can pass through the objective is the indirect light scattered within the sample. However that is a hypothesis, and if anyone has any other ideas, feel free to drop me an email.

I also captured a video of the sample in this setup, where I changed the focus from above the sample to below.

When the iris on the condenser was opened up a bit the image returned to a typical bright field appearance, but with quite low contrast.

I wouldn’t want to use this objective like this, as there are plenty of ways to get better images, but with the iris closed down, the ‘dark field’ image looks quite cool, so I may try it for other samples in the future.

The 170x Leitz objective itself is shown below.

You may well have noticed something odd here. This objective is 170x, but the field of view of the images looks not too dissimilar to those from the 60x objective in my previous post on this slide. Don’t forget that this objective was designed for use with a 400mm tube length microscope, but this one is 160mm. Effectively this reduces the magnification by 2.5x times (400/160), making the magnification just 68x. This is why the images look to be similar magnification to the ones with the 60x objective. This objective does have a big brother – a 300x one – and I may try that as well in the future given the results with this one.

Part of science is doing experiments just to see what will happen, even when you think it may not work. If we knew everything that was going to happen, why do research? With this, I tried imaging using an objective which common sense said wouldn’t work, but I think it’d be interesting to try. It produced an interesting result and I learned something new. Winner. As always, thanks for reading, and if you’d like to know more about this or any other aspect of my work, I can be reached here.

Using shorter wavelength UV light for microscopy can (and does) improve resolution, allowing smaller features to be resolved than with longer wavelength light. There are however times when using UV is impractical, or impossible, and in those circumstances, other approaches to improve the image are required. Today I’ll cover some basics about one of those approaches – the use of oblique lighting to illuminate the subject. With oblique lighting, we get to emphasize certain features on a subject, and this approach was used to capture the image shown below using my custom modified Olympus BHB microscope.

So what is oblique illumination? Basically, it is offsetting the direction of the incoming light to the subject compared with normal bright field microscopy. Changing the angle of light emphasizes some features within a subject, and the direction of the incoming light can be altered to highlight specific features depending on their orientation. As an example let’s take a look at three images of a Synedra superba diatom, captured using an Olympus 40x Dplan Apo UV objective, using 405nm light, on my Olympus BHB. Firstly, normal bright field, with the light coming from directly below.

And now, two images using oblique illumination, one at 90 degrees to the other.

As can be seen from the images, using oblique illumination highlights certain features, and by changing the direction the light comes in from different features can be emphasized.

The images above have been cropped from the original and have not been cleaned up (they still show all the imperfections from the camera sensor – microscopy will show everything on the sensor, even when you think you have cleaned it). Below is a cleaned up image of the Synedra superba diatom slide using 405nm oblique illumination with the Olympus 40x Dplan Apo UV objective.

The oblique illumination gives an almost 3D appearance to the image which I really like, and there is certainly plenty of resolution given the NA of the objective (NA 0.85). Note, I have flipped this horizontally, as I like the composition better this way round. However I still wanted to go further, so I broke out the 60x Olympus SPlan Apo (NA 1.4) oil immersion objective. The objective and condenser were oiled to the slide and this is what an image from that looks like using 405nm oblique illumination.

There is a big step up in resolution with this objective, as expected given its much higher NA. However showing the images at this size online doesn’t really show that, as they have to be reduced in resolution for the website. Below is a crop from the image above, which has also had to be shrunk in resolution).

It is interesting to note in this image the ‘imperfections’ in the ‘perfect’ structure. However as mentioned, even this was reduced in resolution for sharing. Below we go in even tighter, and keeping the original image resolution.

This tight crop shows some really impressive detail, with the small features being about 200nm across.

I’m a bit of an imaging geek, so at this stage I’ll show the objectives and condenser that were used for this work, and the slide itself (which cost me about £30 on eBay).

The slide was made by Arthur Cottam and is over 100 years old (more information can be found out about him here).

I’d also like to give a bit of a ‘shout out’ to a company – J.B Microscopes Ltd. When I bought the 40x Dplan Apo UV objective on eBay, the coverslip thickness correction collar was stuck and wouldn’t move (as far as I know the objective was bought new and then kept in a cupboard for 40 years, so presumably the grease had set). J.B Microscopes were able to fix this for a very good price and send it back to me within a few days of sending it to them, so I would definitely consider using them again in the future.

I hope that has been interesting for you to see. Microscopy continues to amaze me, as do the details I am able to see with my little custom made Olympus BHB UV microscope. Thanks for reading, and if you’d like to know more about this or any other aspect of my work, you can reach me here.

Very often I get caught up in showing my microscope images, and while they are nice to see the magnifications can sometimes be so extreme it is difficult to get a feel for what it is you are actually looking at – for example see below, which shows part of a diatom taken using 313nm light on my UV microscope.

I’m not too good on naming diatoms yest, but I think the one above is an Actinophycus senarius. Today I want to go on a bit of journey with you, from large scale down to the very small. We will go from ‘centimeter’ down to ‘nanometer’ and take a look at the images at those different scales. Note, the images being shared here are low resolution ones of the originals, and and as a result are less sharp than the originals, but ideal for sharing online.

We start on the large scale with the microscope slide itself. This is standard sized one of 7.5cm by 2.5cm and the image was just taken with my camera phone.

This slide and coverslip is quartz rather than glass to allow me to use short wavelength UV light to look at the sample – glass would be essentially opaque down at 313nm. The diatoms are in the middle of the slide surrounded by a metal washer and with a coverslip on top, and in the image above you can just about see it. It becomes easier to see if we zoom in a bit – the image below was from the gentleman who very kindly made the slide for me and you can make out the individual diatoms in the arrangement. This cluster of diatoms is about 2mm across.

Next, I thought it would be good to get an overview of the diatom arrangement using the microscope. I should say at this point that all the microscope images which are shown below were taken using a monochrome converted Nikon d800 (conversion done by MaxMax in the US). This has very good UV sensitivity and tends to be my ‘go to’ camera for the microscope.

The image below was taken using white light bright field imaging using a 4x Zeiss Planapo objective, and with oblique lighting.

With the image above you can see the individual diatoms, but it is difficult to make out structures. As you can see the whole arrangement is about 2mm across. The objective has a relatively good Numerical Aperture (NA) for the magnification – the NA is 0.16 – but it is not going to be enough to show fine structure. Using oblique lighting (I used an Olympus Aplanat condenser which has this ability built in to it) gives an almost 3D appearance to the image.

The next stage is to highlight a specific diatom and look closer with a higher magnification. The one I’ll be looking is highlighted in red in the image below.

To look closer it’s time to go to a higher magnification (a 40x objective) and use 313nm light – the shorter wavelength gives better resolution for a given numerical aperture. The objective of choice was a Leitz 40x UV objective with an NA of 0.65. I used a Zeiss quartz condenser with an NA of 0.85, and the image was straight bright field. First I’ll show two images of the diatom in question.

As can be seen, even at this magnification, the depth of field is very shallow, and the height of the sample needs to be changed to get the high and low parts of the top surface of the diatom in focus. When the top of the sample is in focus, it is very sharp – this part of the diatom is what is fixed to the coverslip using gelatin. It is possible to combine the two images above using stacking software (I use Zerene Stacker), to produce the following.

With this magnification, it is possible to make out some small features on the diatom, and it shows it in its entirety, so it is a good intermediate magnification to use. But I can go to even higher magnifications and with an objective with a larger NA to get even more detail.

Next, we have an image showing part of the diatom using a Leitz 100x UV objective and Zeiss quartz condenser with an NA of 1.25, again at 313nm. The image is now shown as monochrome rather than sepia toned and has been rotated 90 degrees. This is the same image as at the top of this article.

At this magnification and with such a high NA, the depth of field is now tiny, and the best focus is on the parts of the diatom with the gelatin which are attached to the coverslip. The observant amongst you will have noticed that there is any array of dots on the parts of the diatom which are in focus. In the 100x image, these look black, and in the 40x image they look white. I’ve noticed this with diatom images, as you slightly change the focus, what looks to be dark dots go white, and then dark again. I believe this is to do with the size of the dots and diffraction of the light as it comes through the holes – as the light comes through it diffracts and then as you change focus you get constructive and destructive interference, and the features change from light to dark. For the 40x image, it happened to be where thy looked light, and or the 100x image, it was sharpest where they were dark. As they are holes, it makes sense to me for them to look dark.

For a final image the one above can be cropped, focusing in on the top right had part. This is shown below.

In the cropped image the holes themselves become quite clearly resolved. Those ‘holes’ are about 300nm across. Normally to get images with this type of resolution, it’d be easier to use scanning electron microscopy, and while it would be possible to resolve these features with optical microscopy without using 313nm UV light, it becomes complex, relying on either more involved lighting setups, or more processing, or even different ways of preparing the sample. With UV, because of the short wavelength involved the resolution for a given NA increases (as per Abbe’s equation), and in fact this was originally why UV microscopy was invented – the pursuit of resolution. While it is an old technique now (and I have even had some microscopists describe as a ‘dead technique’ than no-one would use these days) it still has its uses in the 21st century. As a technique it is actually relative simple to do (once you have the equipment), and means that simple bright field imaging can be used making image capture and processing less complex.

I’ll be giving a talk on the design and building of my UV microscope at the Royal Photographic Society Imaging Science Group Good Picture Symposium 2022 in London in December (for details see here) where I will go into the challenges with the technique and share more high resolution images taken with it, including some of my work on the imaging of sunscreens, which is what got me started on this journey.

It’s been fun going from something that is large enough to be seen by the naked eye, all the way down to the nanometer level, and I hope you have enjoyed it as well. As always, thanks for reading, and if you’d like to know more about this or any other aspects of my work, I can be reached here.

I’m very happy to be giving a talk to the Royal Photographic Society Imaging Science Group later this year (3rd December 2022, at the University of Westminster in London) as part of their Good Picture Symposium. The day will consist of 7 talks from imaging experts in different fields and the flyer for the event is shown below.

Please contact Dr Mike Christianson for tickets using the details shown in the flyer above, and you don’t need to be an RPS member to attend – these talks are open to all.