Bitumen Process

[Nomad Z]

The earliest surviving photograph was made around 1826 by Joseph Nicephore Niepce, using bitumen as the liight sensitive substance. More can be read about him here...

https://en.wikipedia.org/wiki/Nicéphore_Niépce#Photography

I've been interested in the beginnings of photography for some time, and I've decided that I'd like to try Niepce' bitumen process. I plan to document how this proceeds in this thread.

The first thing to do was some research to get a better idea of the materials and method. It seems that the main ingredients are Bitumen of Judea and lavender oil. I've also seen white petroleum and oil of turpentine mentioned. So far as I can tell, white petroleum is what we now call white spirit. I couldn't find anything about oil of turpentine, but I suspect this is just the solvent that's extracted from pine resin, known to artists as distilled turpentine. It may be the case that some stuff referred to as "oil of" isn't really an oil as we might think of it, but closer to a somewhat volatile solvent (another is oil of spike lavender, which isn't lavender oil, but a strong solvent used in oil painting - it dilutes the vegetable oils that are used in the paints in the same way that distilled turpentine does). From what I've read so far, it seems that the bitumen is mixed with lavender oil, although one source says it was mixed with turpentine.

The basic method seems to be pretty consistent in what I've found, namely that the bitumen is dissolved in the lavender oil and then coated onto a substrate. This is dried and then exposed to light. When the exposure is made, the plate is washed in solvent, causing some parts of the coating to dissolve, leaving an image. The key property of the bitumen is that areas that get more exposure harden to the extent that the solvent is less able to dissolve them, resulting in a negative. Descriptions of the solvent vary, Some sources say it's just more lavender oil, some say it's a mix of that and white spirit.

Next, I needed to source the ingredients. The solvents were easy - white spirit from any DIY place, turpentine from an artists supply shop, and lavender oil from anywhere selling essential oils. Of those, the lavender oil was the most expensive at £12 for a 100ml bottle. The hardest thing to find was the Bitumen of Judea. It turns out that this stuff is used in antique restoration and conservation - it's used to add patina to repaired sections of antique furniture. It comes as a liquid and as a powder. I didn't want liquid because it meant I would be stuck with using the maker's chosen solvent (which I don't necessarily know anything about), and is also rather expensive for the amount of actual bitumen it contains. After some searching, I found a conservator place in Portugal that sells the powder in 500g and 1kg bags. It was 6 euro for 500g, and 9 euro for 1kg, and 12 euro for shipping, so I decided to burn the extra 3 euro and get 1kg, even though it's far more than I'll ever need (and so is 500g).

Then it was time to try some experiments to see how the bitumen behaved with the solvents. Something I'd heard mentioned in a video about Niepce' original image was that it's quite grainy when examined up close, suggesting that it's not the case that the powdered bitumen dissolves completely into the solvent to make an amorphous liquid. I mixed a bit of the powder with each of the three solvents to see what happened, and got quite a noticeable variation in how grainy or gritty the result was. Each mix was in a test tube and when each was shaken before being left to settle in a stand, the walls of the tube had a deposit.

These can be seen here...



The leftmost tube is actually a mix of lavender oil and turpentine, but was originally just lavender oil. In terms of grain, the white spirit was the strongest and the turpentine had the least, The lavender oil on its own was somewhere between the two. When I saw that the turpentine seemed to dissolve more of the powder, I added some to the lavender oil tube, and this seemed to make it less grainy. You can see the difference between the eventual three mixtures in the photo above. The white spirit on the right is particularly dark due to the amount of poorly dissolved powder deposited inside the tube.

With this, I decided that I wouldn't try using white spirit to make up the liquid to be coated, but would make one using just lavender oil, and another with a 50/50 mix of lavender oil and turpentine. The mixture is 3g of powder to 10ml of solvent, based on a video where somebody uses the method to make contact prints from large format negatives (he used only lavender oil).

The next thing to do was to try coating some substrates to use for tests. A nice, convenient substrate is microscope slides, available in packs of 50 for not much money. Four slides were numbered with the intention of coating a pair with each mixture. The idea is to have one to play with and a backup in case something goes wrong, without having to wait to coat more.

At this stage, I want to try to establish a few things...

What it's like to coat glass with the mixtures.
How to dry them.
What the grain looks like once they're dry.
What solvent and method works for removing the uncured bitumen.

I tried coating using a fairly soft artists paint brush, but wasn't terribly impressed. It was a mix of grainy particles with extremely thin solution between, with thicker bits where the brushstrokes had overlapped. This was the case with both mixtures. I decided to use the brush to just pick up quantities of the mixture and dollop it onto the glass until there was enough to get it to flow around slowly when the glass was tilted.

Here's a slide coated with the lavender oil mixture...



And a view looking down on it with a small torch behind...



The colour of the mixture, and the graininess become apparent. (The reddish tinge was apparent in the video I mentioned earlier, where the chap was doing contact prints.)

Here's a closer view with some manipulation to bring up some more detail...




There are clearly some large particles that haven't dissolved, which are most noticeable at the left where they make bumps in the surface of the oil. Then there is a general suspension of finer particles which can be seen in the illuminated area. I don't know if the reddish colour is due to even smaller particles suspended in the oil, or whether some of the bitumen has actually dissolved. In the test tubes and the two jars containing the mixtures, there is certainly a bit of sludge at the bottom which can be dissipated into the liquid with a bit of shaking.

Anyway, the next thing is to get the slides dried.

 

[Nomad Z]

From what I've seen, it looks like a bit of heat is used to help get the plates dried. To this end, I'm trying a photographic tray warmer with some extra bits I made from acrylic sheet...



The frame thing is supported at three points to allow for some levelling to help reduce the chance of coatings pooling at one end of the plate. The sheet of glass in the middle is just there to hold up the microscope slides. The bit of wire is a thermocouple that goes to a digital thermometer so that I can measure the temperature. The business end of the thermocouple is in space, just under the sheet of glass.

And with the slides in place and a cover over everything...



Slides 01 and 02 are with the lavender oil mix, while 03 and 04 have the 50/50 lavender oil and turpentine mix. The cover serves both to keep dust out and warmth in. The measured temperature is about 45°C.

The drying seems to be working - here are the slides after a bit under 2 hours...



The two with the turpentine are well on the way, while the nearest corner of slide 01 can be seen to have dried a little. I just had another quick look and the two oil-only ones have some dried more, and 04 is now almost completely dry.

Next step will be to examine these when they're fully dried. I have access to a microscope at work, so I think a close look at the grain will be in order.

 

[sirch]

Very interesting, thanks for sharing. Just one thought, did you consider warming up the bitumen solution to see if more dissolved?

Looking forward to seeing how it turns out

 

[Nomad Z]

Very interesting, thanks for sharing. Just one thought, did you consider warming up the bitumen solution to see if more dissolved?

From my research, there doesn't seem to have been any mention of that, but it might be worth a try. Another possibility is to dissolve in turpentine and then filter it to remove the larger particles, but I think I need to get a close look at the slides under the microscope first to see what the granularity is like. Would also need to work out a suitable filter (lab filters and coffee filters are available, but might be too fine).

 

[Nomad Z]

An update...

After about four hours, all of the slides were dry apart from one bit of slide 01, which was lavender oil only. At this point, I switched off the tray heater and went to bed. The next morning, there wasn't much change to the remaining bit of slide 01, suggesting that the heat helps a lot with the drying time. The temperature with the tray switched off was about 16°C. A bit more heating later that day had the last bit dried.

Here's how the slides looked after drying with lighting from above...



The surface is quite textured, and the ones with only lavender oil look a bit rougher or coarser. The coating feels pretty robust, although I haven't tried damaging it yet - I'll maybe do a bit of scraping later.

On a light table...



Two things stand out here - the lavender oil ones are much more dense, and there has been some pooling of the gloop, which is clearly evident on the two made with the lavender oil and turpentine mix. That said, the pooling doesn't look too bad, and I suspect it could be mitigated by controlling the amount of liquid that's applied - if there's a bit less, the surface tension should help to keep it in place.

Since my light box can barely get through the denser slides when using my auto digicam, I did a balancing act with a little torch clamped in a vice.

Lavender oil only...




Lavender oil and turpentine mix...



Overall exposures are slightly different because the camera is compensating for the relative brightness of the illuminated section in each.

A bit of chopping to make comparison easier...



To give an idea of scale, the lens on the torch is about 18mm diameter.

The coarser surface of the slides with only lavender oil correlates with what looks like larger particles shown on the left. It's also interesting to note that the overall colour is somewhat different - more reddish with lavender oil, and yellowish with the lavender oil and turpentine.

 

[Nomad Z]

I then took one of each slide into work to have a look at them through a microscope, and that's when things started to get very interesting. Unfortunately, there's no camera on that microscope, although there is a machine that does have a camera, but doesn't produce very good images. Here they are anyway...






These are at the same magnification, and have been tweaked a little to show what's going on (the machine's under-stage lighting wasn't terribly strong). Based on what I observed through the optical microscope, I would say these two images show an area that's no more than about 2mm wide, maybe a bit less.

With the optical microscope, I observed the following...

• The particle size in both varies a lot.
• The largest particles in the oil were larger than those in the oil and turpentine. Lavender oil was about 100um, maybe a bit bigger, while the turpentine ones maxed out at about 70um.
• I didn't estimate the size of the smallest particles in the lavender oil, but the ones in the oil & turp mix looked like they were down to about 5um, maybe a bit smaller.
• The particles, especially with above-stage lighting, looked like big, fractured lumps of bitumen or tar - hard edges with flattish clean surfaces.
• In both cases, it looked very much like a load of particles set into a thin layer of some sort of binder, and there were areas of clear binder between the particles.
• The binder was coloured, like stained glass, although both looked closer to reddish under the microscope (I might check this again).

The particle size estimates were based on knowing the width of the field of view in the scope in millimeters for various magnifications, and estimating the particle sizes based on the proportion of the field of view they occupied.

The areas of clear binder were interesting. It's not like the whole space was occupied with ever-smaller particles - there was a distinct difference between areas with particles and unoccupied spaces with just binder. Also interesting is the colouring of the binder. So far as I can tell, this is a change in colour of the oil and not a suspension of particles (which is what a paint is). If it is particles, they are far smaller than even 1 micron across - probably into the range of a few nanometers before the microscope can't resolve shapes or textures. the gaps between the particles just looked like red glass or cellophane - completely amorphous with no hint of texture.

I made my observations using various combinations of lighting: under-stage to look through, above-stage to look down on top, and a bit of both to see where the dried oil was combined with some texture on the surface. The texture and shape of the individual particles was fascinating - I didn't expect them to look like big lumps of bitumen in miniature. What particularly stood out was the shaping of the flattish surfaces - although generally flat, there was some curvature here and there, and then I realised that what I was seeing was a thin coating of oil over the particles, which was clearer on the large particles because they were sticking out of the thin coat that was on the slide.

So, the structure appears to be a thin coating of dried oil with tiny particles suspended in it, and larger particles, in effect glued on and sticking out of the main coating, with some of the dried oil over the top of them making them shiny. This apparent structure gives me cause to wonder about the photographic process involved here. If the oil is dried like a varnish, and the particles are trapped within it, or coated with it, then the idea that the photographic mechanism works because the bitumen hardens and becomes insoluble when exposed to light is open to question.

Consider...

• If the dried oil remains soluble, then putting an exposed plate into a solvent is going to wash away both the dried oil coating and the particles suspended within it, regardless of how soluble or insoluble those particles are.
• If the dried oil is not soluble, then the particles trapped within it are not going to wash away in the solvent because the dried oil is protecting them - it doesn't matter if some of them harden and become insoluble because they're surrounded by stuff that's insoluble anyway.

It should be noted that the particles are already in a solvent, were for several days before coating (with regular shaking to encourage them to dissolve), and plenty of those particles have not dissolved.

I find myself wondering if it's the oil that's the photographically active constituent rather than the bitumen. If the oil progressively cures as a result of exposure to light, then those areas that are still susceptible to being dissolved will wash away to some extent, and so will the bitumen particles suspended in those areas. In other words, it might be the case that it's the binder that is susceptible to becoming insoluble, while the particles serve to simply provide opacity or contrast.

This hypothesis will require some testing, so I have coated another two slides, this time with only the lavender oil - no bitumen or anything else. They are in the drier.

I think the next steps will be something along the lines of finding out what it takes to wash away the coatings on the slides without any overt attempt to make exposures, and, assuming stuff dissolves, to repeat that after trying some exposures to see if there is any sign of the coatings becoming insoluble. The slides with only the lavender oil become controls to try to determine what is photographically active.

 

[Mostly Vintage Cameras]

Good job.

 

[Nod]

Might be worth trying to grind the powder finer before dissolving it in the solvent(s).

 

[Terrywoodenpic]

pestle and mortar.?

 

[Nod]

Yup.

 

[Nomad Z]

I'll maybe try grinding at some point, although it's already pretty fine.

I had another look at the two slides through the microscope. I would say that the one with the turpentine mix had smaller particles than the one with just the lavender oil, although the latter is harder to see through because it's more dense. I would also say that the turpentine one has a lot of smaller particles - there are voids where there's just the coloured varnish, but those are quite small and most of the space between the bigger particles is taken up by smaller particles in a wide range of sizes.


As part of the test to see whether the lavender oil itself can be rendered insoluble when exposed to light, I coated a couple of slides with it and put them into the drier. After several hours of the same temperature as before, I found that the oil hadn't really dried. There was a thin film where it had been applied, plus various droplets...



In this photo, you can see the droplets at the left and along most of the top. These were wet. The smudged area in the middle at the bottom is where I gently rubbed the tip of a finger on the coating to see what the surface was like - it felt draggy and not at all like the other slides with the bitumen mixtures, which had a surface that felt hard and dry. In the above slide, the right-hand end was gently wiped using a soft paintbrush dipped in turpentine and it seemed to remove most of the oil film easily. A hard boundary can be seen at the right-hand end of the smudged area, where I wiped across the slide with the brush.

I also tried scraping at the edge of the coating (there is none near the little label) with the tip of a scalpel, and just got soft stuff or thickish liquid coming off. This was quite different compared to the slides with bitumen in them - on those, scraping at the edge of the coating showed that it was very brittle and scraped away quite easily. I would say both coatings (oil only, and oil & turps) are robust enough for making images, but would be susceptible to damage if mishandled - easily scratched.

I decided to see how the two bitumen slides behaved when treated with turpentine...



(Note the scraped areas at the bottom-right of each slide.)

Both of these were wetted a bit at the left-hand end using a soft brush dipped in turpentine, and the brush used to work the surface a little. Some came off, but not much - dipping the brush in the turpentine indicated that there was only a small amount of staining of the turps coming from the brush. I decided to get a second jar and decant some of the turps into that, and then stick the two slides into it. This let me put a cap on the jar to keep the stench down, and be able to swish the turps around in the bottom of the jar. The washing effect was pretty quick - about 10 minutes, consisting of the jar mostly sitting still with just the occasional swish to agitate the liquid resulted in the above. The top one (oil only) seems to have been more resilient, which figures - the lower one was from a mix with the same amount of liquid, but 50% was turps, which would evaporate as it dried, leaving less oil in the coating for a given amount of bitumen. Note that I can't say whether putting the slides into a bath was better than using a brush - it may have been the case that it takes a bit of time for the solvent to work its way into the varnish and start dissolving it. My initial attempt with the brush might simply have served to get the process started, and would maybe have worked if I had persevered with it.

These two jars show what happened to the turps...



On the left is the amount of staining resulting from cleaning the brush after the original coating of the slides, and on the right is how much darker it got when I washed the ends of the slides.

So, the stuff certainly washes off when the slides are relatively unexposed. (They've been exposed to some daylight, but somewhat weak and with very little UV content.)


The lack of drying of the oil on its own is curious. I can only assume that there is some chemical reaction with the bitumen which allows the drying to happen. The turpentine shouldn't be a factor (or not a requirement, at any rate) since two of the slides were coated in a mix of oil and bitumen, with no turpentine involved, and they dried just fine.

I decided to try another experiment to see if I can get the oil to dry, and also to see what happens when a mix of bitumen and turpentine is filtered to remove the particles. From the initial tests with the various solvents, I took the test tube which had bitumen and turpentine (because that had the least amount of sludge at the bottom) and ran it through a filter paper into another test tube...



You can see above how dark the filtered liquid is. Given how much sludge there was in the source test tube, and how much was on the filter paper afterwards, I think most of the powder has dissolved. There certainly seems to be an element of the stuff dissolving given the stained glass look that the voids had when viewing the slides under a microscope.

When the filtering was done, I doubled the volume of liquid by adding lavender oil to make up a version of the oil & turps coating with the particles filtered out, and arguably with more time given to allowing the bitumen to dissolve. This is also a more dilute version. The resulting liquid was used to coat another slide...



...and that's now in the drier. I had a look after a couple of hours and it does seem to be tacking up. No sign of brittleness with the tip of a scalpel yet, but not noticeably draggy under a finger tip.

The idea with this step is to try a couple of things...

• Test whether the dissolved bitumen reacts with the oil such that the coating dries when heated.
• See what happens by starting with bitumen powder that has been dissolved in turpentine as an alternative to grinding, sifting, shaking, etc.

The idea here is still to test the hypothesis that the oil is a binder and that the particles really just add opacity in the same way that silver particles do in silver halide film. The oil has changed to something that's a mix of oil and 'essence of bitumen'. Basically, I'm trying to make the stained glass stuff seen in the voids of the earlier slides. If this dries and can be washed away, then the first two requirements are satisfied. It then remains to try and make it insoluble by exposing it. So, the test has shifted slightly - no longer just the oil, because that's hard to test if it doesn't dry, but the oil/bitumen compound that results from adding powdered bitumen.

While the idea might be to have some sort of binder that can be cured, and to use the undissolved particles to give opacity, it might also be possible to use just the oil/bitumen compound without particles as the image-forming substance. I suspect slide 07 is maybe a bit too pale for imaging, so it would be interesting to see what happens with a more concentrated solution, and also to see if the turpentine saturates when a larger proportion of bitumen is added (the solvent tests were just a small amount of powder to the amount of liquid). I'll see how this dries, and if it works out okay, I 'll have a look under the microscope.

 

[Nomad Z]

Quick update on the slide with the filtered turpentine mix...

It's not really drying as well as the first ones. The coating is more coherent than the pure lavender oil. In particular, there are no droplets. There's a noticeable but slight tackniness under the finger, and scraping with the tip of a scalpel doesn't show any sign of brittleness or wetness. It's more like something slightly sticky. And this is with substantially more hours in the drier than the first slides (probably about 10, compared to 4, which was enough to dry the original oil & turps slides).

If the bitumen is acting as a siccative, I suspect it's probably the case that there just isn't enough. As mentioned previously, the ratio of bitumen to turpentine was pretty dilute - probably about 0.2g to 3ml of turps. I've now made up another batch based on the 3g/10ml ratio, where I put 3g of bitumen into a test tube and added 6ml of turps (about 7.5x more concentrated). The idea is that some of the liquid will be lost into the filter paper when filtered, so I should hopefully end up with about 5ml of filtered bitumen/turps solution, to which I can add 5ml of oil.

It'll probably take a day or three for the bitumen powder to dissolve fully into the turps, so I'll keep an eye on it. What I noticed with the initial turps/bitumen mix was that the amount of sludge at the bottom gradually reduced until there was hardly any. So, it's a case of checking a couple of times a day to see how much sludge there is, and then giving it a good shake before leaving it again.