Gum analysis

From: Katharine Thayer ^lt;[email protected]>
Date: 10/15/04-07:02:16 AM Z
Message-id: <>

Robert Schaller wrote: (September 6, 2004)
> Dear Katherine,
> I indeed can't see why mylar would introduce any impurities, certainly
> not water-soluble ones, and wouldn't these be the problem? I also want to
> applaud your efforts in this, as it seems long overdue that someone conduct
> this investigation in an orderly, scientific fashion. I wish I was set up
> to do it. I am curious about how the investigation will proceed. What
> tests exactly is the chemist planning to do? I get the impression that just
> hardened gum (or PVA) will be tested first; what next? As I said, I am very
> curious.

Hi Robert,
I'm sorry about the delay in answering; I set your question aside to
answer when I had more time, and forgot about it until September 28,
when I wrote much of the below, and then forgot it again til today.

There are many unresolved questions about the mechanism of gum and
dichromated colloids in general, and I wish I could say that we were
embarking on a comprehensive research project with the intent to answer
them all, but such is not the case. Given that we're volunteering our
time and have other things to do, we had to pick one question that
interested us both and that didn't seem to be adequately answered in the
literature. We are interested in the mechanism of the crosslinking, and
especially in the role of chromium in the crosslinking.

There is a great deal of received wisdom on this issue, but very little
actual data to back it up. It may be that all we accomplish is to
confirm the received wisdom, but that would be something worth doing at
any rate, though less exciting than discovering something no one knew
before. But as far as received wisdom goes, the received wisdom is only
consistent in the general point, that the chromium is involved in the
crosslinking; however, in the specifics of how the chromium is involved,
there is no agreement at all. Several models have been proposed, all of
them different. Duncalf & Dunn concluded that in dichromated PVA,
hydroxyl groups are oxidized to carbonyl groups and that the carbonyl
groups are not oxidized further to carboxyl groups; furthermore the
crosslinking, which they only speculate about but which they provide
indirect evidence to suggest that chromium is involved, is independent
of the oxidation of the colloid; in other words they concluded on the
basis of their research that neither carbonyl nor carboxyl groups are
involved in the crosslinking.

On the other hand Sasaki et al propose a model involving hydroxyl,
carbonyl, and carboxyl groups coordinating with CrIII in crosslinked
PVA. But the Sasaki model is suspect because it was designed to fit a
stoichiometric calculation that was simply calculated wrong, and
furthermore it was based on a molecular weight for PVA that was off by a
magnitude of 10 from what it should have been given their own
assumptions about the degree of polymerization and the molecular weight
of the reacting unit, in other words the Saski model is flawed because
it was designed to closely match numbers that contained two

 Mannivannan et al give what seems to me a definitive description of the
reduction of the chromium, but offer no insight into the mechanism of
the crosslinking. Grimm et al offer a model for the crosslinking in a
phosphor suspension, involving the splitting of the PVA into cleavage
products with terminal carboxyl groups which make octahedral chelate
complexes with CrIII. However, they offer no empirical support for the
theoretical model, and furthermore it seems rather clear that a phosphor
suspension is different in important respects from the films that are
used in photographic processes using dichromated colloids. The phosphor
suspension seems to behave more like a solution, as the mechanisms they
describe are similar to those that occur in irradiated dichromated
colloid solutions, not to the ones that occur in irradiated dichromated
colloid films. (The word "film" here refers to the thin dry layers used
in alt-photo processes.) One clue to how different they are is that in
the phosphor suspension, as the crosslinking proceeds the suspension
becomes less viscous, probably due to the cleavage of the PVA chains. In
dichromated films, however, it is well known (Duncalf & Dunn for
example) that as crosslinking proceeds, the film becomes more and more
viscous until it is completely insoluble.

It just occurs to me that what I'm doing here is writing the literature
review that goes in the introduction to a research article, and that you
all don't need to know the rest of it at the moment; the point was just
to give you some of the flavor of why we think this research might be

The first question we are addressing is a very basic one: IS there
chromium in crosslinked gum, and if so, how much? Duncalf & Dunn are the
only ones to date (1964) that I can find who have asked even the first
part of this question. They say that they found chromium in dissolved
and reprecipitated crosslinked gum "from which residual chromium had
been extracted." However, they do not describe either their method for
extracting the residual chromium or their method for determining the
presence of chromium. Our first analysis then is simply an attempt to
replicate Duncalf and Dunn on this one basic question.

I prepared gum for analysis by coating mylar with dichromated gum and
exposing for maximal hardening without dichromate stain. I put it in
water to dissolve the unreduced dichromate and dried it, then cleared
it in 5% sodium bisulfite to remove any residual chromium, washed it and
dried it again, then scraped off the dried crosslinked gum. If anyone is
interested, this gum is clear and colorless in the film, and the
scrapings are in color and appearance something like very thinly shaved
paraffin wax. It isn't waxy, but it shaves off in a similar way and has
that same whitish translucent look to it.

I sent this gum to the chemist just after Labor Day, and he has been
practicing with a small sample of it since then, to make
sure he's confronted and worked out solutions for any surprises that
might arise, before committing the whole sample of gum I sent him. To
say that the man is meticulous would be a gross understatement, but I'm
glad he is; I will be that much more confident of the results when they

First he dried the sample completely in a 110 degree furnace (to
eliminate any excess water in the gum) and weighed the dried gum several
times to see whether the gum keeps the same weight or whether
the weight changes while sitting, due to gaining or losing moisture from
the air. He has worked that all out and is confident he has an accurate
and reliable weight within .1 mg. The next step is to combust the
crosslinked gum in a furnace that will burn off a lot of irrelevant
stuff (carbon, nitrogen, oxygen, sulfur from the clearing agent, for
example). When the whole sample of gum is combusted, the ash that
results will be analyzed by an analytical chemist to determine exactly
how much chromium is present. So in the end we will know exactly how
much chromium per mg of crosslinked gum. This won't be terribly useful
in developing a model for gum crosslinking, because the structure of gum
isn't completely understood and its molecular weight is indeterminate
and variable. So this first analysis will only be useful for determining
if chromium is present in an amount that would suggest that chromium is
involved in the crosslinking. If the answer is yes, as one would expect
given the conventional wisdom, then we will have to move to PVA and do
the whole thing again. You could say that this first analysis is
actually unnecessary, and you would probably be right, but (1) I was
curious about gum, not PVA, (2) I hadn't got around to finding a good
source of a small amount of pure PVA of a determined molecular weight
and (3) the guy said, why don't we start with gum and see what happens,
and so that's why we started this way.

 Since PVA is simpler and its molecular weight can be determined
accurately, a determination of the amount of chromium per mg of
crosslinked PVA will help point to what model for the crosslinking works
best, and we'll go from there. Any speculation about the model for
crosslinking of gum will have to be extrapolated from the model for

I wrote all the above on Sept 28; since then the chemist has thoroughly
combusted the practice sample, ensuring first that all the material
stays in the crucible during combusion. He weighed the resulting ash,
and has done some initial analyses and determined that he wants to
adjust the furnace temperature higher since he got some irrelevant
inorganics in the ash that would be burned off with a higher temperature
and make for a cleaner analysis. (I hope I'm not putting words in his
mouth; that's not exactly what he said but it's the impression I got
from what he said.)

At any rate, all of this is preliminary, there aren't any actual results
to report yet, but I thought it might be interesting for non-scientists
to see some of the process that scientists go through to set up an
Hope that answers your question,
Received on Fri Oct 15 13:58:06 2004

This archive was generated by hypermail 2.1.8 : 11/03/04-10:51:23 AM Z CST