Swelling of Dichromated Gelatin Film

[Joe Farina]

The attached paper is called "Swelling of Dichromated Gelatin Film," and appeared in a 1984 issue of Photographic Science and Engineering. The author is Sven Sjolinder.

This paper is excellent. At first glance, it seems quite dry and unexciting. However, there is a huge quantity of carefully-observed tests and supporting data. I think it's one of the best papers on DCG ever published. I've referred to it many times over the years, especially recently.

For example, figure 5 on page 182 graphically shows the amount of swelling which takes place during the typical processing sequence.

Also, there was some recent discussion on another thread about a supposed "hardening" effect when gelatin is treated with cold water, of which I was skeptical. However, on page 182, a paragraph is titled in bold letters: "Effect of Cold Water Treatment", and relates: "Pretreatment in cold water decreases the thermal movements in the molecules and permits the gelatin molecules to form hydrogen bonds (reference 5) which are sufficiently strong to resist further swelling even if the temperature is subsequently raised above 20C. This annealing effect decreases swelling as is confirmed by experiments, with results shown in Table II."

Superb.

[Steven]

Many thanks for posting that Joe.
I agree, it's an excellent paper.

I notice that they were not using any form of chemical hardening in their processing.

I'm trying to get a colour shift from green (532nm) to yellow, but I want it to be narrow band as the depth of the scene is 100mm.
I need to make some more DCG plates and have another play.

[Joe Farina]

Thank you Steven.

That's true, no chemical hardening. As you probably noticed, he is pre-hardening the gelatin substantially by baking at 150C for 2 hours. I too am trying to get a shift to a longer wavelength. Ordinary DCG does tend to expand after processing, but in my case the "washing out" of the solid TMG most likely causes a substantial shift in the opposite direction, towards blue. But of course, other (overriding) factors may be involved.

Sjolinder's processing is simple and consists of three steps:

1. Tap water rinse/bath
2. 50% IPA
3. 100% IPA

He does make it clear that the second development bath (50%) is important with regards to final replay wavelength. I think this may correspond to the "tuning bath" so often mentioned by Rallison. I've had some trouble in finding out how Rallison used his tuning bath (0.86 specific gravity, which I believe corresponds to ~70% IPA) to shift the replay wavelength towards the red (I'm searching my papers now). But Sjolinder makes the clear statement on page 184: "The resulting swelling after processing is ruled by the swelling in bath 2. If a large swelling is achieved in this bath then the resulting swelling after drying also is proportionately large."

Figure 9 on page 183 shows how swelling varies in different IPA concentrations. I was surprised at the difference between 50% IPA (~300% swelling at 20C) and 70% IPA (~170% swelling). From the data supplied by Sjolinder, it seems that 50% IPA would cause more of a red-shift, in comparison to 70%. Can anyone confirm this?

[Joe Farina]

From the Rallison section of the wiki: "Control of DCG and nonsilver holographic materials":

"The direction of the shift can be controlled by the ratio of alcohol and water in the first bath and the amount of shift can be controlled by the time in the same bath. A near ideal tuning bath has a specific gravity of .86 when it is warmed to about 55 degrees C. This process can be repeated many times if necessary, especially if the last hot bath is not hot enough to cause excessive scattering center buildup. Multiple buffer baths between the first color control bath and the last dehydration bath help to keep the last bath clean."

[Steven]

Thank you Steven.
Figure 9 on page 183 shows how swelling varies in different IPA concentrations. I was surprised at the difference between 50% IPA (~300% swelling at 20C) and 70% IPA (~170% swelling). From the data supplied by Sjolinder, it seems that 50% IPA would cause more of a red-shift, in comparison to 70%. Can anyone confirm this?

As I'm a beginner, I have played with (and still do) different water/IPA concentrations in the first alcohol bath.
Yes, more water in the first alcohol bath tends to push the resulting image towards the red end of the spectrum, but I think that the temperature of this bath is key too.
If the temperature of the first bath is not warm/hot, but cold, you can end up with very little red shift depending on the concentration and temperatures of the following baths.

I also get the impression that the temperature of the first alcohol bath determines the rate of dehydration in that bath.
A hotter bath producing a faster rate of dehydration and hence, a more broadband result.

It's interesting that the paper calls the first alcohol bath the swelling bath.
I always thought that the swelling bath was the first warm water bath after the chemical fixing bath, the first alcohol bath being called the colour tuning bath.
The temperature of the warm water bath determining the amount of swelling.
I suppose that if the temperature of the first alcohol bath was higher than the first warm water bath, then more swelling would occur.

I think the difficult one is to get a red shift and narrow band.
I made a single beam Denisyuk couple of days ago that came out shifted towards yellow but it was not narrow band.
It had very good DE (for me), and it was so bright that it made my eyes water when I viewed it under sun light.
The hologram cast a very dense magenta shadow.

I'm just starting to get a feeling for the DCG development process, so I'm still learning.
I think that it is probably the most difficult part of DCG to get right.

[Joe Farina]

Thanks Steven. That's interesting about the need for warmer alcohol/water in the tuning bath to get a decent red-shift, but the side result being a more broadband image. It's looks like we have the same objective at the moment: to get a narrowband red shift.

Years ago, there was a long thread in the old forum about red-shifting DCG, which was started by John Pecora. Some of the discussion included the idea of hardening/stiffening/etc. the layer while it was still swollen. The plan, as I recall, was to stiffen the layer in a water solution (i.e., while swollen), so that when the thickness decreases during the dehydration, it has more resistance to shrinking, and doesn't shrink as much. There was some disagreement in that thread, and the results didn't seem conclusive. But based on some of my results, I think the idea has merit.

Before I began to reorganize my lab about a year ago, one of the last holograms I did turned out very well. It's shown in the thread about speckle noise reduction. My first bath in water is the "fixing" or development bath, and this contains a water solution of aluminum sulfate and sodium metabisulfite. As a test, I added formaldehyde (formalin) so that the final volume of formalin was about 3%. This hologram (2-color) turned out quite narrowband, as can be seen in the green swirl in the marble. I forget what other color was in that particular hologram (which was a narrow strip cut from a larger exposed hologram) but it was enough to show how narrowband it was, and quite bright and low-noise.

Since then, I've found that the pH of this solution was about 2.3. Also, checking a swelling curve for type B gelatin, it seems that the swelling is extremely high, perhaps maximum, at this low pH level. After exposure, I let the plate set in this room-temperature solution for 5 minutes. The composition of the bath was 50ml (4 grams per liter water) aluminum sulfate and 50ml (20 grams per liter water) sodium metabisulfite, plus 3ml formalin. While in this bath, the layer must have been in about the maximum swelling condition. Then warm water bath at 30C for 1 minute, then 70, 91, and 100 IPA. The resulting hologram is shown in the photo.

So, this tells me that hardening with formalin in the swollen state may have potential. I am just getting geared up for some more tests.

[Joe Farina]

For reference, this is the thread started by John Pecora, unfortunately the images are no longer available. It was a DCG hologram which was shifted narrowband towards the red (from blue to green as I recall), very nicely.

http://holoforum.org/oldforum/viewtopic.php?f=2&t=4815

There are quite a few discussions in various threads regarding the red-shifting of DCG in the old forum archive.

[Joe Farina]

This is the hologram which turned out nicely using formalin. It has the marble in the middle with the green swirl, and the red tile in the lower left corner.
http://holoforum.org/forum/download/file.php?id=707

[Joe Farina]

Since Steven brought up the subject of bandwidth broadening associated with red-shifting of DCG, I've attached an 11-page paper called "Wavelength shifting and bandwidth broadening in DCG" by Corlatan/Schafer/Anders. The maximum number of attachments is five per post, so I broke it up into three posts.

[Joe Farina]

Pages 6-10 of paper: