Swelling of Dichromated Gelatin Film

[Joe Farina]

Last page of paper:

[Dinesh]

Be aware that these papers and results refer to HOE type dcg holograms, ie monochromatic reconstruction. In this regime, it's not particularly difficult to get the entire visible range. For example, we got the plots below for a client wanting a multi-spectral imager. All these reconstruction wavelengths are on the same plate.

However, display is a totally different beast. The very first plot in the paper shows the efficiency "according to Kogelnik", but the authors don't say which of three possible expressions they used. The one usually quoted is the tanh squared function. This is the efficiency for unslanted, pure phase Bragg planes. Firstly, a display hologram cannot have unslanted planes, secondly, there can never be a pure phase hologram. I also find it odd that he claims that a thicker emulsion will give more efficiency. Yes, Kogelnik does infer this, but I think the kogelnik theory does not take into account the material absorption of the light as a function of actual efficiency. In other words, the efficiency is calculated from a coupled wave model, but the actual efficiency/sensitivity of a real hologram is dependent on the ability of the medium to absorb the radiation. Ammonium dichromate has an absorption peak around 400nm, thus, it absorbs 400 - 500 and actinic reaction takes place. However, that same property that allows actinic reaction also causes an attenuation of the wave as it passes through the emulsion. Thus, the intensity is dropping. I remember arguing this at POC, where the Kogelnik model required 40 nm for the best efficiency. We tried it anyway (I didn't have a lot of choice despite my misgivings! It's one reason I started Triple take Holographics; "senior management" simply look to the "experts" and the "famous holographers", without any actual rational analysis of the problem! But I digress with a personal gripe). When I took the plot, it looked like a profile of the Himalayas. The project at the time required a 25 nm bandwidth with a peak at OD 4 - ie straight up and straight down.

MSI plots.jpg (40.8 KiB) Viewed 3563 times

[Joe Farina]

Thank you Dinesh. Those sure are some sharp, narrow, and efficient peaks. I was wondering, can you share your method of narrowband red shifting? If it's proprietary, I understand.

[Steven]

Many thanks for the info Tony/Dinesh.

I cast four fresh plates yesterday evening using mould coating.
I will sensitize two, each with a one minute dip in 3% AmDi tomorrow night and do a couple of exposures early Tuesday morning.
I intend to leave them for two days after exposure before processing.

I have attached a photo of my best 5"x4" coin Denisyuk hologram that I made a couple of months ago.
It finally came out a gold/copper colour after reprocessing several times to get it right.
It took me six hours of reprocessing to get there!
Initial six or so passes through the processing gave gold at edges, but green streaks in centre.
I concluded that the centre was harder and therefore needed a higher swelling bath temperature to shift colour.
I'm trying to achieve the same colour with a four inch deep scene I'm shooting, but I'm finding that difficult without it going broadband.

Details:
Mould coated 2mm picture framing glass using a 30-200 mix of 240 bloom gelatin, shims = 56um Scotch Magic tape.
Dried in still air for 48 hours at ambient (15-20C)
Dip sensitized in 3% AmDi for 30 seconds at ambient.
Dried in still air at ambient for 24 hours.
Exposed at 532nm with a power density of approximately 100mJ cm2, assuming 70% reflection from the coins.
24 hours dark reaction.
Kodak Rapid fix until clear + 60 seconds (about two minutes total) temperature of fixer bath 22C.
Wash 10 minutes in cool tap water.
Swelling bath at 46C
Allowed plate to cool to temperature of fist IPA bath, 27C.
60% IPA at 27C for 8 minutes.
90% IPA at over 30C for 2.5 minutes - I forgot to document actual temperature.
!00 % IPA hot >50C for 5 minutes. Again, I forgot to document actual temperature.
Dried with my magic hair dryer.

Steven.

[Joe Farina]

This one may interest you, Steven.

[Joe Farina]

last page:

[Steven]

This one may interest you, Steven.

Many thanks Joe, it certainly does.

It looks like the old school of thought that only giving a five minute dark reaction time after exposure before development is not doing our diffraction efficiency any favours. Funny, but I came to similar conclusions many months ago, not by experiment, but just by using some logic.
What I wasn't aware of, is how long this dark reaction DE enhancement went on for.
Maybe something should be added to the DCG wiki?

I'm not using a high concentration of AmDi for sensitizing, my standard dip is only 3%, but I have gone down to 1.5% at times.
My reason for using low concentrations is to obtain a more equal exposure through the depth of the emulsion, as well as the fact that I'm shooting SBR.

Steven.

[Joe Farina]

It looks like the old school of thought that only giving a five minute dark reaction time after exposure before development is not doing our diffraction efficiency any favours. Funny, but I came to similar conclusions many months ago, not by experiment, but just by using some logic.
What I wasn't aware of, is how long this dark reaction DE enhancement went on for.
Maybe something should be added to the DCG wiki?

I've always thought that the dark reaction has not been used enough in DCG. I guess impatience has something to do with it, that is, people probably have the desire to see the results of the exposure right away. However, the dark reaction has a lot going for it. It just keeps cranking away over time, there are no worries about vibration, temperature changes, fringe drift, mode hops, etc. The fringes are locked in place, the exposed areas keep outpacing the unexposed areas during the dark reaction. In my case, with Jeff's TMG system, I've been told that the dark reaction isn't the same (it's supposed to be very much less) so I haven't looked into the issue yet. But I've always been a trifle skeptical, and suspect there may be a useful dark reaction for dye-sensitized TMG/DCG. If you're getting some good results, it would definitely be an excellent contribution to the wiki. Also, it might inspire me to do some tests with dye/DCG, to see if there's a noticeable effect with a post-exposure delay.

[Dinesh]

Thank you Dinesh. Those sure are some sharp, narrow, and efficient peaks. I was wondering, can you share your method of narrowband red shifting? If it's proprietary, I understand.

Thanks, but I'm afraid it is proprietary. When I retire in about 5 years or so, all will be revealed. But, until then, there are probably lurkers who are working on HOE type applications and I'd hate to deprive them of the joy of thinking!

However, the point is that the techniques would not work on display holograms. Techniques involving swelling to shift the peak lambda work because of the fact that the planes are mostly parallel to each other and at some angle to the face of the emulsion. In this circumstance, as the emulsion swells/contracts, the planes get carried along, and possibly twisted, and so simply increase or decrease the spatial frequency. In a transmission geometry, the planes are almost perpendicular to the face of the emulsion, and so swelling/contraction of the emulsion has little effect. Bear in mind that even under these ideal circumstances, the "planes" are actually not planes, but sinusoidal variations. Thus, any non-linear swelling due to local variations in density will introduce additional Fourier components.

In a display hologram, the "planes" are not planar, but convoluted, "non-Euclidean" surfaces. As such, as the emulsion swells/shrinks, the planes will not expand uniformly and parallel to each other. The forces on these surfaces will depend on their curvature and on the local density and the local density profile. This warping will probably introduce noise as low frequency Fourier components will appear proportional to the radius of curvature, the surrounding index and the index modulation. As the swelling/shrinking continues past some point, the radius of curvature will change. I suspect a full theory of the effects of swelling/shrinking inside a display hologram set of surfaces may well involve tensor components.

So, the work on colour shifting by swelling/shrinking the emulsion should be taken with an eye towards the fact that in a display hologram, there are additional effects besides simply an alteration of the spatial frequency - there are noise effects that may be introduced as well as non-linear effects that may come in with large percentage expansions.

[Dinesh]

In my case, with Jeff's TMG system, I've been told that the dark reaction isn't the same (it's supposed to be very much less) so I haven't looked into the issue yet. But I've always been a trifle skeptical, and suspect there may be a useful dark reaction for dye-sensitized TMG/DCG.

The TMG may make a difference, but there should be some dark reaction. In any complex system, if the system has two or more stable states at different energies, but the system has to "go over" a potential hill to get from one stable state to the other stable state, then energy needs to be inserted into the system, in the form of light, for example. However, the transition between the two states can occur at random, without any deliberate energy input - think tunneling for example - or by energy getting in from the surround - charged particles in the environment. So, the cross linking of the gelatin needs energy, but can, and does, occur at random, due to various random events. However the addition of TMG, changes the energy picture - the height of the potential hill - so that the frequency of random events is also changed.

[edit: this whole thing hinges on the entropy inside the dcg/tmg. I've always thought that the entropic relations inside an emulsion would make a fascinating research project, but as a friend of mine says, that's project number 56. perhaps in 5 years when I retire!]