[Cardiff, Wales – 6:43AM, 12th of March, 1928]
God, it was good to be home.
A year spent travelling to the Orient negotiating with creditors in Hong Kong, bribing officials, hiring surveyors and setting out industrial plans had exhausted me. Kitty hadn't been exactly thrilled that I'd left soon after Charlotte's birth, but she at least understood.
And even though I was now back home, most of my time was spent assuring creditors that their investments in Indochina were not going to be burnt down by communists, meeting with licensees or law firms over negotiations for my patents and simply doing the obscene amount of paperwork that was required of me.
So times like this where I could work in my laboratory were precious.
Of course I was currently working one-handed since with the other arm I was holding Charlotte, who had just turned one. Honestly, I didn't want to leave her with her grandparents, but until she was old enough, I couldn't exactly take her with me.
She was very clingy since my return, though, so it appears my absence was noticed to some degree. And to be honest, I enjoyed just spending time with her when I wasn't doing the more dangerous experiments.
For my current project, I was currently working on producing a Kodachrome film.
Kodachrome was based on the subtractive colour model, using cyan, magenta, and yellow dye layers. The steps I had needed to perfect to do this were as follows: Create three separate emulsion layers sensitive to blue, green, and red light, respectively. Use sensitising dyes to make each layer sensitive to a specific wavelength. Find a way to produce dye molecules during development that correspond to each primary colour layer. And finally, ensure the dyes are stable and do not bleed into other layers.
The materials I needed in bulk were several bags of cotton linters, which were the short fibres left on the seed after a gin had removed the longer staple cotton fibres, along with drums of acetic anhydride, sulphuric acid, sodium hydroxide, acetone, and triacetin.
The triacetin was a polymer coating and was a pain since most polymer coatings around were camphor, which was basically an acid to acetate film, so I had to shill out like £50 a barrel for the stuff since very few places made it, and I had to get it shipped from the states, which delayed my progress by two months.
The process started by washing the cotton linters in water and then bathing them in sodium hydroxide to remove further impurities.
It's then rinsed once again in water and then dried. We then add it to a vat along with acetic acid to partially swell it and continue to add more acetic acid to the mixture until it is fully absorbed. Small amounts of sulphuric acid are added to catalyse the reaction, converting cellulose into cellulose triacetate.
It took a few weeks of trial and error to adjust the reaction time or add water during the reaction to hydrolyse excess acetyl groups, yielding cellulose diacetate, which was less brittle and more suitable for film.
I then had to neutralise residual acid by washing with a dilute sodium bicarbonate solution, which removed unreacted acetic anhydride or acid, which could lead to acetic acid production later. Big no-no.
Then hydroquinone was added to minimise oxidative degradation and magnesium oxide to neutralise any further acetic acid that might form over time and prolong shelf life and film stability.
I then had to dissolve the cellulose mixture into acetone to start the casting process. Next the triacetin polymer is added, which will give the film more flexibility and durability. It had to sit in an airtight container for a few hours to allow air bubbles to rise to the surface. If I just poured it now, I risked the acetone evaporating before Sir bubbles escaped. Once I was sure it was properly mixed and had no microscopic air pockets, the process was continued by spreading the solution thinly on a flat surface and smoothing it out with a spreader and allowing the solvent to evaporate, leaving a flexible film behind after a few hours.
Heat treatment is the next critical process used to relieve internal stresses, improve dimensional stability, and enhance the overall mechanical properties of the film. This process helps the cellulose acetate film achieve the desired balance of flexibility, durability, and optical clarity.
I cried at the amount of losses I had to burning, warping, and melting.
It starts by baking the film in an oven at around 40 and 50°C for a time to make sure all moisture and solvent had been evaporated, and then we had to slowly raise the heat over the course of a few hours to somewhere between 90 and 130°C.
After heat treatment, it needs to rest at room temperature for a few hours to reset, and then the film may be trimmed, slit, or rolled into the final desired shape.
I've probably gone through close to a thousand batches before nailing it. Too brittle, too soft, too misty, it has air bubbles, it's too thick, it's too thin.
I cried this night; I had made five batches in a row, and they had all come out perfect.
Step One Complete.
Colour mixing with subtractive colours makes my head hurt but works like this.
Cyan + Magenta = Blue (Red and green wavelengths are absorbed, leaving blue.)
Cyan + Yellow = Green (Red wavelengths are absorbed, leaving green and blue; together, they appear green.)
Magenta + Yellow = Red (Green wavelengths are absorbed, leaving red and blue; together, they appear red.)
Cyan + Magenta + Yellow = Black (All wavelengths are absorbed, resulting in no reflected light.)
I would need to 'invent' synthetic dye couplers, so I just decided to steal Kodak's work again.
Their yellow coupler was called dichlorophen-chlorophenyl-methyl-ethyl-triazine.
Their Magenta Coupler was called Dichloro-phenyl-methane-sulfonyl-phenyl-phenyl-oxadiazole.
Their cyan coupler was called chlorophenyl-chlorophenyl-ethyl-triazine.
Okay, so for the yellow dye, the challenge started with getting this triazine structure just right. You know, a triazine ring is like a puzzle, three nitrogen atoms in a cycle, and we needed to build it so it would react with an aromatic compound to make the colour happen. Now, we didn't get it right the first time... or the second. It took months of trying different combinations of chemicals to get that perfect balance.
First, I started with dichlorophenol, which has two chlorine atoms on the ring. Then, I tried attaching this triazine ring that was also chlorinated and had an ethyl group sticking out. The trick was getting the right reaction between the triazine and phenol so we could get those two to bond, but in the right way. Sometimes I'd get something close, but not quite right, and I had to adjust temperature, solvents, and timing until we nailed it. The trick was making sure that everything lined up to form that strong bond between the two rings.
After many, many failed attempts and tweaking the recipe, I finally hit on a mix that gave us the right yellow! So yeah, it took forever, but it was totally worth it in the end.
Now, for the magenta coupler, that one had its own challenges. This time, I was working with an oxadiazole ring. It sounds fancy, right? But the real challenge was getting two aromatic rings to bond with the oxadiazole ring in just the right way.
I started with dichlorophenyl and methane-sulfonyl-phenyl. The goal was to get these two groups to connect through a ring structure: the oxadiazole. But let me tell you, those aromatic groups didn't want to cooperate! I had to carefully choose conditions to make the cyclisation work so that the two phenyl groups could lock in with the oxadiazole.
I also had to figure out how to introduce the sulfonyl group. Too much, and it wouldn't form the right structure; too little, and I couldn't get the colour I needed. There were a lot of trial runs where things either didn't connect, or they connected in the wrong positions.
But with months of adjustments, different temperatures, solvents, and times, we finally got the right product. The trick was finding that sweet spot, where everything worked perfectly, and the result was this vibrant magenta dye. It was a real breakthrough when we finally cracked it.
Ah, the cyan coupler was a bit like the yellow one, but with some new twists. It also used a triazine ring, but this time we had to add an ethyl group and use a different positioning of the chlorine atoms.
It wasn't as straightforward as it sounds. I started with chlorophenyl and the triazine compound, but the trick was getting the ethyl group attached to the right spot. If I got the ethyl group in the wrong position, the colour didn't work out. But if I got it right, it would produce a sharp cyan, which was the whole goal.
Like with the yellow coupler, it took a lot of trial and error, adjusting the alkylation process and playing around with temperatures, solvents, and times. I also had to be super careful with the chlorination, because too much or too little would change the final product. After months of tweaks and tests, we finally got the ethylation just right, and bam, cyan. I couldn't have been happier."
Now to actually turn our lovely, precious cellulose acetate film into colour film. The sheer amount of failures I got made my heart bleed. Which is incredibly funny since I was sure it turned to coal years ago like all Englishmen.
I start by creating an emulsion that contains silver nitrate mixed with halide salts like potassium bromide, potassium chloride, and potassium iodide in a certain mix that I had to fine-tune specifically for each dye coupler. The specific mix of halides determines the sensitivity and the colour characteristics of the final film, along with gelatine to hold it into place and water to act as a solvent. The emulsion is what reacts upon being exposed to light. This is then coated on one side of the film, allowed to set, and then repeated again for the next two layers.
Cyan for red, yellow for blue, and magenta for green.
At this stage, the film is still in its unexposed state, but it's now light-sensitive. To expose the film, expose the film to light using a camera, and the silver halides in each emulsion will form a latent image based on the amount of light they're exposed to. The blue layer will record the blue light, the green layer will record the green light, and the red layer will record the red light.
We continue by coating on top of the now-set emulsion with a developer made of phenidone and one of the dye couplers of our choosing. It doesn't really matter which order. This is then coated onto the film, converting exposed silver halide crystals into silver, setting up the dye image. Once done, distilled water is run over to stop the development and then repeated for the other two dye images. The specific mix of salts in each emulsion layer is what ensures the dye couplers attach to the right layer. That was a nightmare to perfect.
The bleach process is especially important because it removes the metallic silver that forms the negative image without damaging the dye image. Potassium dichromate and sulphuric acid are what make up the bleaching agents.
It takes about 5 minutes at 24°C. Bleaching agents require careful control since they can affect the dye layers if overexposed.
Afterward we use ammonium thiosulphate. This step fixes the image and prevents further chemical reactions. The film must be immersed for about 5 minutes.
And finally we add a mixture of citric acid, formaldehyde, and tannic acid, which stabilises the film and helps to ensure no more chemical reactions take place and makes sure the film can last years in proper storage.
Watching the vivid images I had snapped of some of the local trees, I couldn't help but smile; sure, I had cheated, but still, it took teams of people years of trial and error to do this, and by myself, I had done it in a few months.