>> From the Library of
Congress in Washington DC.
>> Lynn B. Brostoff:
I'm very happy
to introduce Dr. Rachel
Obbard today as our speaker.
Rachel is an Assistant Research
Professor at the Thayer School
of Engineering at
Dartmouth College.
She holds degrees in Material
Science and Engineering
from the Colorado School
of Mines, the University
of New Hampshire, and
Dartmouth College.
Rachel is most interested
in porous materials
and her fascinating work
on sea ice takes her
to both the Arctic
and Antarctica.
We did not have our
Skype meetings
when she was in Antarctica.
She has been interested
in the study of materials
and cultural heritage for many
years and she first reached
out in 2013 to the Metropolitan
Museum of Art with an interest
in doing a project of this sort.
And then they reached out to
the Library of Congress in 2015,
and it took a few
years to get funded.
But the work she will
present today is the result
of a collaboration that resulted
between PRTD, Dr. Sylvia Centeno
at the Met, and Rachel, of
course, as well as scientists,
students, and interns at
our institutions in Cornell
and Argonne National Labs.
And I just like to add
that as you likely know,
some of the early work on
this subject of tidelines,
which is very important to paper
conservators, was conducted
by our very own Elmer Eusman.
And we are really pleased to
build on that work as well
as the work of Ann
Laurence [phonetic] Dupont,
which many of you know, and
her colleagues in France.
>> Rachel Obbard: Thank you
Lynn for that introduction.
Hello everyone.
Thank you so much for having
me here and for attending.
I'm sure this is a bit
out of your busy day,
and I'm excited to
present this work.
This is something we've
been working on for a couple
of years now and we've got
an enormous amount of data
and some really pretty pictures
so I hope you'll enjoy this.
So the order of my
talk will be first
to give you a very small
amount of background.
There is certainly huge
body of work on this topic
but I'll give you the background
that I think is necessary
for what I'm going
to talk about.
Talk about our objectives for
this talk and for the study,
our analytical approach or
techniques, the results we got,
and a little bit of a
summary and some next steps.
So tidelines, as you very
well know, are residual damage
on paper at a former
wet/dry boundary.
We know that visible
tidelines can be created even
with ultrapure water
on purified cellulose.
We also know that repeated
wettings with increasing amounts
of fluid will produce
successive tidelines.
This study builds on the
previous work as Lynn mentioned,
particularly that of
Elmer Eusman, which showed
that these tidelines
can be produced even
with purified substrate
and water,
and this is a very
important result.
The example shown here is
perhaps not a good example
of that because here you can
also see bleeding of the inks.
But this was one of the
deaccessioned objects
from the Metropolitan Museum of
Art that we began by looking at.
In this study, we examine
some historical samples
but have also tried to avoid
such possibly confounding
factors such as this
by concentrating our work
on tidelines produced
in the laboratory
on 100% cotton rag
or cotton linen papers
using melaque water.
So paper fascinates
me personally
as a material scientist because
it's what we call a hierarchical
composite or a hierarchical
structure.
It starts with cellulose, of
course, long chain polymer,
which is organized into
elementary fibrils,
microfibrils, and then
microfibrils or bands.
And these form what we call
fibers and what we'll see
in images in the
following slides.
This structure is very
complex and it's fundamental
to paper's mechanical
and transport properties.
So as you know, paper
is hygroscopic.
It absorbs water.
And wetting and drying both
produce macroscopic changes
as well as microscopic ones.
We'd like to do some future
work on the structural changes
at the fiber level, but for now
the important thing to take away
from this slide is to remember
that wetting promotes transport.
As Dr. Eusman pointed out,
it was previously assumed
the brown lines formed
at the wet drive boundary
were caused by the migration
of soluble material there.
But research has shown that it's
actually much more complicated.
Cellulose degradation
at the wet/dry boundary is
still not completely understood.
Recent research has shown
the importance though
of hydro peroxide production
in oxidative degradation,
but hydrolytic degradation
is still important.
Chemical changes at the wet/dry
boundary lead to degradation
of cellulose a seen by the
formation of carboxylic acids,
formic acid, and acetic acid, as
well as is seen through evidence
of depolymerization
and of reduction
in mean length of
the polymer chain.
And this reduction in chain
length is primarily the result
of chain scission
which essentially halves the
mean length of polymer chains.
So our objective is to
further the understanding
of tideline formation in
the historical examples
that we selected, in the sized
model papers that we produced,
and in some undersized model
papers that we produced.
And you'll see that
the two types
of model papers gave
very different results.
We are trying to, or have been
trying to, record the migration
and deposition of
inorganic material associated
with the tidelines
through complementary
in situ techniques.
And to further investigate
the effect of artificial aging
on the tidelines on both the
sized and unsurpassed papers.
As I mentioned, we began
by selecting a few historical
objects that we could sample.
But because we wanted
to constrain some
of the variables inherent
in that sort of sample,
we also identified some
modern 100% cotton rag papers
and actually cotton linen papers
and produced tidelines on these
under very controlled
conditions.
So what I'll present
today is just a subset
of all the paper samples
that we looked at so
as not to confuse matters.
We used a wide variety
of techniques,
primarily qualitative
x-ray fluorescence
at several different
resolutions,
scanning electron microscopy
with energy dispersive
spectroscopy
for element identification, and
x-ray micro-computed tomography.
So first I'll talk about
micro XRF techniques.
So XRF, as you probably know,
x-ray fluorescence mapping
or imaging, was developed
at the University of Antwerp
and at the Delft
University of Technology
which I had the pleasure
of visiting last summer.
It's an awesome place.
My colleagues at the
Metropolitan Museum of Art
and the Library of
Congress both conducted XRF
with their respective machines
with the specifications
shown here.
The Brucker [phonetic] M6 system
at the Metropolitan Museum
has a 30 wide rhodium target
microfocus x-ray tube
which is operated
at 50 kVA and 600 microns.
It has a silicon drift detector
and poly capillary optics
and measuring had moved
across the object surface.
The Artax [phonetic] system has
similar components and spot size
and resolution but
is less sophisticated
in terms of mapping.
So it was operated in the study
in the line scan mode as well
as for individual spot analysis.
We show here again that
the print that we called
for shorthand "Flowers," both
sides of it, and two areas
that were analyzed which are
labeled, despite it being turned
on its side, upper
right and bottom right.
These are the elemental maps
from the upper right-hand
tideline area shown along
with a photograph
of that region.
Note how the sulfur
and potassium are the most
concentrated in the edges
of the tideline while the
chlorine is more diffuse.
Here's the sulfur and the
potassium is quite concentrated
right in the edge there.
You see here the
chlorine is quite diffuse
in the tideline region.
Now here are the elemental
maps from the other area,
the bottom right area, and
here again the tide line seems
to have swept up and
concentrated the chlorine sulfur
and potassium although the
concentration of the sulfur
at the edges is the
most pronounced here.
These are both a
little bit more diffuse.
We also looked at
several printed books
from the well-known William
H. Barrow book collection
that you have here
in the preservation research
and testing division.
Shown here is a page from
Barrow book number 1077,
the title which is "the
Modern Practice of Physic"
which was printed
in London in 1768.
So this is the very latest
in medical literature,
I want you to know, and
this is definitely the go-to
for your doctor,
should he or she ask.
Shown here is the page from the
book in UV light on the right.
And you notice that in the UV
light you can see the tideline
here is actually rather thick.
When you look at it
here, your eye is drawn
to the darkest part, the very
leading edge, but you can see
that the tideline itself, it's
not a thin or narrow line.
And we're to be looking
at the accumulation
of elements throughout this
tideline zone, if you will.
So this is a typical appearance
in normal and UV light.
This is actually a gelatin sized
paper with fairly creamy color
and the sizing felt
pretty significant.
The next few slides
show our analysis
of the paper from this book.
So here we see both the
elemental scanning maps here
and the line scan.
So we're using both XRF's here
at the Library of Congress
and at the Metropolitan
Museum of Art.
And the two types of analyses
show the same information
but in a different way.
And while we can get
great visualizations
of the distribution of
elements in the map,
we get somewhat more
detail in the line scans.
So the maps on the left
beautifully highlight the
concentration of sulfur and
potassium at the tideline.
The iron is concentrated
there but not as dramatically
and the calcium and chlorine
really don't show up.
The line scan here is very
busy and that's in part
because of the normalization of
each of the individual curves.
But it includes all of
these elements as well
as aluminum phosphorus.
The other ones, manganese,
iron, copper, zinc, and lead.
So without trying to tease the
individual lines here apart,
it's apparent that all of the
elements actually show a pattern
of increased concentration here
at the visible tideline
indicated
with the blue dots here.
So all of the elements are
increased in concentration
in the region at just
below the tidelines.
And there's even,
if you look closely,
sort of a hint of
a double tideline.
You see this peak here
and then there appears
to be another peak
closer to the tideline.
In a minute, we'll look at
and XRF of this same paper
with the lab produced tidelines.
So remember, this is an old
tideline, we don't know how old.
In a minute, I'll show you a lab
produced tideline on that paper.
So here's how we
produce the model systems
and the laboratory tidelines.
So we selected a variety of
modern papers, as I said,
which were either 100% cotton
or a cotton linen cellulose,
and they were either
sized or undersized.
And we're trying to represent
the historical papers and also
to experiment on some different
preparation conditions.
We also made fresh
tidelines, as I mentioned,
on the bearable papers
from the same book.
It's the pages that we showed
you the historical tidelines on.
So the tideline formation was
done by vertically dipping
in melaque altra-pure
deionized water the papers here.
So they're held on a rack and
they're set down into this pan
of water for 15 hours in
a controlled environment.
And this work was all done here.
In our testing, we noticed
that no new gelatin sized
papers formed visible tidelines,
although UV did show a faint
fluorescence at the water line
in all of these samples.
So in other words, the
sizing was very effective
at stopping any water transport
in the new sized papers.
We will even pre-aged some
of these papers before
tideline formation
and we still didn't see any
transport in the sized ones.
The samples were also
artificially aged
at the Library of Congress.
You see here Leah, the intern,
standing in front
of the aging oven.
And then samples were mailed
to the Metropolitan Museum
of Art and to Dartmouth.
So here is a paper from
that same Barrow book,
now with the laboratory
made tideline.
The aged paper produces
a much sharper
and more pronounced tideline.
Take a look at the maps over
here, particularly the sulfur,
the potassium, and the overlay
with the sulfur and chlorine.
Presumably, the sizing in this
paper is degraded over time
which allowed that
transport, but the aging
of the cellulose itself could
have had an effect as well.
So this shows similar movement
in sulfur and potassium to that
which we saw in the
old tideline, but iron
and calcium are now
also apparent.
The line scan shows
some additional detail,
and again these different
curves are normalized.
The timeline is very
complex and, in particular,
notice the offset of
the elements here.
So the x-axis is vertical
distance across the tideline.
And here's the top
of the dark tideline
and then the tideline
fades, and the bottom
of the dark tideline is here
and then this area is
below the tideline.
So this paper where you see
the separation of the peaks
like this, the paper
is essentially acting
as a thin layer chromatography
substrate.
And in this case, the magnesium
and the zinc appear at the front
of the tideline and the
calcium and sulfur the back
with the potassium
in the middle.
So the takeaway here is that
the massive concentration
of what were trace
elements in the paper,
including the transition
metals, and both sets
of results are qualitative.
I should remind you
not quantitative,
but they do show very
good signal-to-noise ratio
and also good correspondence
to one another.
So now, I'm going to
show you some XRF mapping
of model samples.
And what we did was, Sylvia
at the Metropolitan
Museum took the strips
and set them all next
to each other here.
So this photograph shows
you all of the samples.
I realize that the writing here
is to too small for you to read,
but I'm going to refer to
some of the different strips
so that you'll know
what they are.
And then here on the
left are the line scans
of the different elements for
those papers, so you can sort
of get a correspondence between
the strips and the line scans.
So all of the unsized
papers form tidelines
with strong concentrations of
sulfur, chlorine, and potassium.
And we were initially surprised
to see such a high concentration
of these elements in these
tidelines because, again,
this was melaque water and we
believed very clean cellulose,
and we thought it might've been
because the polypropylene trays
that we were using hadn't been
cleaned thoroughly enough.
So we went back and
super cleaned them
and also super cleaned
and used some glass trays,
and the test do showed that
these elements at the tideline.
So they're not coming from
the trays or the water.
So one of the things to note
in this set is that none
of the gelatin sized papers form
tidelines although the UV did
show some fluorescence
there at the water line.
And I don't have
the UV image here,
but the one circled here
are the sized papers.
This is the UICB 2000.
And the other two are Alana and
another sample of the UICB 2000.
So you can see that we're
not getting any action here
in the line scans.
If I look now at
these three papers,
the second through
the fourth samples,
the second through the fifth,
these are Whatman [phonetic]
number one at different stages
of aging post tideline.
And you notice here that
the tideline appears
to become more diffuse
with aging.
And so, this is from zero
days to 3.5 days of age
in an 80 degree C oven.
Okay, these ones circled are
the undersized UICB samples,
again with aging of over
a period of several days.
In contrast to the
Whatman [phonetic] ones,
the UICB papers, you
don't see any widening
of the tideline with aging.
So there doesn't seem to be
any diffusion going on there.
So that was what
we call micro XRF.
In March 2016, three
student undergraduates
and I took some samples
to Argonne National Lab,
the advanced photon
source there,
where a very high power source
produces very high resolution
analytical techniques
of many types.
And we were able to do
essentially nano XRF there.
This is, to my knowledge
at least,
the first time Synchrotron
XRF has been used
in a systematic study of
tidelines such as this.
So at Argonne, we
used beam line 26 ID,
which is I think
somewhere up here.
Each of those little
triangles sticking off
from the essential circle at
Argonne is a separate beam line.
And at the time that we
applied for beam time
to beam line 26 ID, it
offered both x-ray fluorescence
and x-ray tomography.
And we were hoping to use
both techniques simultaneously
in order to get both structural
and chemical information
on the samples.
Unfortunately, by the
time we got there,
the tomographic capability had
been moved to another beam line,
so we did the only x-ray
fluorescence there.
Before we went there, I'll also
talk about micro CT which we did
on machine at Cornell.
And our goal was to try
to do both structural
and chemical analyses
at the same spot
on a sample using the
two different instruments
in two different locations.
So what we did was we
designed a small sample holder
that would fit both systems.
And my undergraduate
student actually came
up with a solid works model
of a tiny sample holder.
It's 4 millimeters across
the top here and it has
in it a tiny 2 millimeter
square window.
And we glued the sample to that
with a tiny bit of rubber cement
so it was something
that was fairly viscous
and wouldn't be absorbed
by the paper.
And here is an image
of the sample mounted
in the beam line at Argonne.
That's one of my students.
So the idea was to be able
to identify the tideline
on this sample.
You can't see on
this particular one.
This may have been an
early [inaudible] test
of the sample holder.
But typically you can see
the tideline on these,
and then the idea was to
be able to know what areas
of the sample you were
sampling when it was
in different [inaudible].
Okay, so result.
So these are the element maps
from the Synchrotron XRF.
This is, again, Whatman
[phonetic] one.
Note here that it
picks up more elements
than the micro XRF's do,
and also that it shows their
distribution on the fiber level
which I think is just so cool.
So particularly if you look
here and here, you can tell
from the distribution
of the elements
that the fibers are there.
And I'm sorry I don't
have an optical image
to show you to compare to those.
Note that these elements are
in order of atomic number,
and unfortunately
sodium cannot be seen
because we were using
a 9 KV source
and sodium was below the
florescent below the cut off.
It's interesting to note
looking at these are some
of the elements, such as
the sulfur and chlorine,
are very much associated
with the fibers,
whereas others are not.
For instance, if you
look at the calcium,
this is clearly just
some kind of a particle.
And the fact that it's also
associated with aluminum
and silicon phosphorus suggest
to me that it may be some kind
of a mineral particle,
basically a piece of dirt.
And incidentally, the
line scanning micro XRF
and in the SEM also showed
the presence of particles,
particularly calcium
particles in these tidelines.
Some other observations, we
wondered of course what form
of sulfur and chlorine
are co-located here,
just because we have the two
elements doesn't tell us what
the compound actually is,
which is a problem of all
of the techniques I'm talking
about today unfortunately.
It's probably not calcium
sulfate or calcium chloride
but perhaps it is potassium
chloride or potassium sulfate.
We really don't know
and we don't know how those
compounds might affect the
paper degradation.
It's also really interesting
to see that the metals,
such as the silicon here
and the copper down here
and the manganese are
themselves associated
with the fiber structure.
And this suggests that
they are associated
with the paper itself, and
beyond that we don't know.
There's clearly a lot of
chemistry going on here
that we haven't figured out yet.
Okay, now I'm going
to talk a little
about the scanning
electron microscopy result.
So we have an SEM at
Dartmouth which we use
and we also use the one here
at the Library of Congress.
So at Dartmouth, what
we did was, of course,
with the skinning
electron microscope,
once you put the sample in the
SEM, you can't see the surface
of it so you can't tell
what the tideline is.
So we marked the tideline
using a tungsten needle.
We just made little
punctures along the tideline.
So on this sample here, here's
the tideline, and you notice
that you can also see
it due to the presence
of these bright particles.
And on the sample, this
is about the tideline
and that's below the tideline.
We did not code the
samples we were using
in an environmental SEM.
We actually started coding them.
In the beginning we coded them.
We got, of course, an enormous
gold peak that we said, "Oh,
heck, we don't need this."
So we started using it
in environmental mode.
We collected energy
dispersive spectra
from our associated EDS system
and we found sodium calcium
and sulfur prevalent in the
areas with the particulates.
So this is a spectra
of somewhere here.
We took three spectra
in each location,
so I don't know whether it
was here or here or here.
I could go back to the
date and tell you that,
but they were all very similar.
So on the tideline we saw
sodium, sulfur, calcium,
a smaller chlorine peak, and
of course, I shouldn't fail
to mention the enormous
carbon and oxygen peaks
that would naturally be
associated with paper.
Then these are really cool.
These false color SEM
EDS images were produced
by Tana [phonetic] here at
the Library of Congress.
So this false color
map helps you see
where the different
elements are associated
on the structure of the paper.
So here you see sulfur
is red, sodium is green,
calcium is blue, and of
course when you put the red
and blue together you get
shades of purple here.
So you can see here that the
sulfur is strongly associated
with both the sodium
and the calcium,
so you've got the red bits
here and the red adding
to the purple there, even
the particulate matter
that you can see here in
the backscattered image.
And here's a close-up of
a particulate on a fiber,
and here the element
maps suggest
that there may actually be
two types of sulfates present,
sodium sulfate and the
calcium sulfate maybe.
And you notice that this
technique is a nice complement
to the Synchrotron XRF because
there we couldn't detect the
sodium, whereas here we can.
And both, of course, give
us fiber resolution images.
Tana [phonetic] also examined
the Barrow book samples
in the SEM, and here
was an example
from the same Barrow book
that I mentioned earlier.
And here we see a strong
association of calcium
and sulfur, possibly
calcium sulfate,
but of course we
can't know for sure.
As you know, UV light is a great
tool for illuminating tidelines,
and we thought these samples
were very interesting.
The top is the Whatman
number one
with the artificial
tideline prepared in the lab
after 21 days of aging.
And you notice here that
you've got a lot going
on underneath it here.
You've got essentially
a double tideline,
but then I don't
see it well there
but there is also all these
streaks running down from it.
So you could say double,
maybe multiple tideline.
It's a very complex thing.
And we can see similar
phenomenon here
in the UICB paper.
It looks different, and this is
the unsized UICB Ptolemy paper.
But again, you can tell
that it is quite complex.
There's gradation going on here,
possibly again I don't know
if you can see it here.
When I look at it on my laptop,
you can actually
see the presence
of double tideline here, much
closer to the primary one
than in the Whatman above.
So these lines can spectra
show just the key elements
in the Whatman number
one tideline before
and after 21 days of aging.
So here's to the day zero,
so you've got the chlorine,
potassium, sulfur, and calcium
peaks all fairly co-located
around the tideline.
And then afterward,
now remember each
of these lines is normalized,
so essentially what's happened
here is that the chlorine
and the potassium, this
is really just noise.
Those have really been lost
from this tideline area.
All you see now are the strong
sulfur and calcium peaks.
We wonder, you know, what is
happening to the chlorine?
Where is it going?
Here's, again now,
Dartmouth SEM images
of the Whatman number one
tideline after 21 days of aging.
Again, you see the
tungsten needle marks.
Here's the tideline here.
We cut little notches
in the edge of the paper
to help us orient them
on the sample holders.
So here, 21 days of aging.
If you remember back
to the earlier spectra
of Whatman number one, where
we had these decent size peaks
of sulfur and sodium
and calcium.
Now they're much, much smaller
with respect to the carbon
and [inaudible] peaks.
And here's a comparison on the
UICB, the unsized UICB paper
after only 1.5 days of
aging above in the tideline
and below the tideline.
So again here's the tideline
marked with the holes.
And here, you can actually see
the double tideline in the SEM
and you can see that
in the tideline --
sorry the font size
is so small --
but essentially you
can see the sodium
and calcium peaks here
and the magnesium.
The peaks in order
are carbon, oxygen,
magnesium, sulfur, and calcium.
And above the tideline
and below the tideline,
all you can basically
see is the calcium peak.
So all of those elements
in the paper are getting
brought to the tideline.
Here's more really nice false
color maps of that paper,
the UICB unsized
paper at age 21 days.
You can see here the zones
with the different
elements are concentrated.
And this analysis also detected
sulfur and magnesium and calcium
but didn't detect any
chlorine or potassium.
The bottom right image here,
which looks like a rainbow,
combines the assigned colors
and shows how the different
elements are deposited
as particles or bands and
they co-locate, for example,
in the case of the
magnesium and the sulfur.
You see the red and the green
co-located here in the middle.
And it's hard to say here
whether carbonates are present
but sulfates certainly are.
So a big part of my work
on other materials is
a nondestructive method
called micro-CT.
And if you don't
know micro-CT is,
I tell people it's essentially
like a CAT scan except
that people always are always
amused when I tell them this.
The CAT scan in the hospital
it looks like a big doughnut
and it's got a source
and a detector in it,
and they rotate around
the patient.
And that's because if
you're sick particularly,
you don't want to
be spun in place.
Well, l fortunately our
nonhuman samples don't have any
such concerns.
So micro-CT's work by having
a fixed source and detector
and spinning the
sample in between them.
So the sample is rotated
in a fixed number of steps,
and in each step an x-ray
attenuation image is captured.
And the projected
images actually impinge
on the scintillator
which converts the
x-rays to visible light.
And then these images
are reconstructed
to produce 3-D images of
the internal structure
of the object, again,
nondestructively.
So it's a fantastic technique
and the post processing can
produce not only beautiful
visualizations of the 3-D
structure, which we'll see
in a second, but an awful lot
of quantitative data about it
on a very detailed level.
For this project,
so at Dartmouth I have
a sky scan micro-CT
that has a theoretical
5 micron resolution.
I can push it down
to 3 sometimes
but for this project we wanted
a much higher resolution,
so we used an [inaudible]
of 550 micro-CT at Cornell
at their School of Biotechnology
and it has a .77
micron resolution.
So these, again, we
used the sample holders
that we designed for this.
So when you look at these images
here you can see the presence
of the sample holder.
So in this image, here you're
looking face on at the sample.
You can see the frame
here holding the sample.
And these other images
are orthogonal slices
through the sample.
So here you're looking
essentially sideways at it.
You can see the edges of the
frame at the top and the bottom.
And here you're looking
down on it.
You can see the edges of
the frame there and there.
And so the area that we
were actually focusing
on was in the center here.
It was right on the tideline.
And again with these samples,
we marked the tideline
with the tungsten needle
so that our colleague
at Cornell would know
exactly where to focus.
And you will see in some
of the images that follow,
you can actually see
the little needle hole.
So the orthogonal
reconstructions on that page,
these orthogonal projections
are actually created
by the [inaudible]
software itself.
This beautiful projection was
actually made using Ocyrix
[phonetic] software, and while
it makes great visuals it's not
too useful for quantitative
analysis.
So what my student and I have
been doing is looking at a lot
of different packages for
quantitative analysis.
In retrospect, we could
probably travel to Cornell
and spend some time there using
their software but we're trying
to find something that we
could use at home at Dartmouth.
And we've tried a bunch
of different things,
including [inaudible],
which is an FVI software,
an image J flavor called Fiji.
So let's see.
The Fiji image is this one here.
And then most recently we
have started using Matlab
to create construct 3-D
volume from the TIF images,
actually the images that
the [inaudible] produces are
in a format called DICOM.
So first you have to
convert that to TIF,
and then you can use Matlab
to reconstruct the 3-D images
from that and to
segment the data.
And the segmentation is
a really important part,
so I'll talk about that next.
So being a structural person,
being a material scientist
type structural person,
I'm really interested in
what's going on structurally,
I mean chemically, yes, as well.
But I'm really a structural
kind scientist at heart.
So I'm really interested in
how are those fibers changing
at the microscopic
level as a result
of the damage they're
undergoing?
We know there's
depolymerization.
But are the fibers or the
microfiber rolls being changed?
And they probably are.
We'd like to be able
to see that.
So in order to see that, we have
to actually isolate individual
fibers in these reconstructions
so that we can look at
them individually rather
than as a clump if you will.
So this is known as a
segmentation problem.
And to give you an example of
how much of a problem it is,
I show you a contrast here
between sea ice on the left
and paper on the right.
And I know that's a weird
combination to make.
It just happens to
be the other material
that I do a lot of micro-CT of.
So what you see on the left
here, you see ice, of course,
is a three phase
system, so you've got ice
which is fairly fresh.
You've got super
salty brine channels,
and then you got
some air bubbles.
And what you can do with the
micro-CT reconstruction is
to separate these
phases out and look
at them independently
of one another.
So here we've made
the ice invisible
and the air invisible
which is minuscule.
Most of the sample
is ice or brine.
And all that's left here
are the brine pockets.
These are vertical brine
channels that formed
when the sea ice formed.
And we've actually sorted them
based on their throat size.
When you have distinct
objects like this,
so whenever you analyze
micro-CT data,
you're working with objects.
You have to say, "What's the
object I'm interested in?"
In paper, the object
would be a fiber.
You know here the brine channels
as objects are quite distinct
from one another,
really easy to analyze
in isolation, if you will.
Not so with paper.
So the segmentation problem
refers to the fact that we have
to figure out how to digitally
peel those fibers apart
from one another so we
can analyze the changes
in each separately and yet
within its native environment,
and we're still working on that.
So in summary overall,
in our model papers,
visible timelines do not form
on the gelatin sized papers.
Also in summary, we found
that the complementary methods
that we used were a very
effective way of looking
at the tideline system.
Micro-XRF line scanning
and mapping were two
complementary methods.
The Synchrotron or Bernano
[phonetic] XRF and the SEM
with the elemental mapping
were nice compliments
with one another.
And they all indicated
significant concentration
of metals in the tideline.
And in particular, the elements
that were the most concentrated
were sulfur, potassium,
chlorine, along with sodium,
magnesium, iron, and calcium.
And we also found that
the localized deposition
and diffusion of elements
within tidelines after formation
and during aging is going on.
And we don't really understand
yet what their significance is
in relation to the
known degradation,
the depolymerization
at the tidelines,
but that's an area
for future research.
So for future work, we plan
to look for more evidence
of the metal compounds or
complexes at the tidelines
and try to look at their
association with the effects
of degradation, possibly using
size exclusion chromatography.
We hope to further investigate
this idea of the multiple zones
within the tidelines that
form with aging and we hope
to solve the segmentation
problem in the micro-CT images
so that we can collect
statistically significant data
on the changes going on
structurally at the fiber level.
So I would like to thank
my collaborators hugely
and their wonderful
colleagues, especially here
at the Library of Congress.
I'd also like to thank
the National Center
for Preservation Technology and
Training for funding my work
and the work of my students on
this project, and the Library
of Congress and Metropolitan
Museum of Art
for supporting my colleagues.
And I like to think the beam
line scientists at Argonne
and Dartmouth Newcomb Institute
for Computational Sciences
for supporting our work on
the segmentation problem.
Thank you.
[ Applause ]
Questions?
[ Inaudible ]
>> Our feeling is that the metal
is coming from the paper itself
because the water
was melaque water.
Yeah?
>> From your fluorescent images,
you have an idea what the
fluorescent material is?
[ Inaudible ]
.
>> Do you mean in the UV images?
>> Lynn B. Brostoff: It
wasn't the focus of our work,
but the work that Elmer
[Inaudible] Dupont has been
doing has focused a lot on
production of hydro peroxides,
and the organic compounds
that are forming peroxide
and that are soluble
or not soluble.
And Elmer and I would say
that there is an assumption
that the fluorescence
is associated
with those organic components.
>> Rachel Obbard: Yeah, and
so what's interesting is
that in some cases we
don't see much fluorescence
but we do see --
or I should say,
we see independently
sometimes fluorescence
and sometimes concentration
of elements at the tideline.
And those two things do not
necessarily go together.
As Lynn pointed out, the
fluorescence we think is coming
from the organic components.
>> When you've noted
diffusion of elements
after aging [inaudible], is
at the moisture in the air
that is enabling them to move
through the paper
during aging process>
>> That's a good question.
Have you thought about?
>> Lynn B. Brostoff:
Yeah, presumably.
But there's a lot of
moisture in paper.
We know that a lot of transport
goes on in normal atmospheres.
You know, the paper is
not immune even in the lab
to having things move around.
>> It's very interesting
to contrast these
results with a dry oven.
>> Rachel Obbard:
Yeah, it would be.
I'm making a note of that.
>> Lynn B. Brostoff: And
in Elmer's early work,
he did that actually,
both dry and light aging,
and with slightly
different results the way the
changes were.
But he was looking
mostly with fluorescence.
I don't want to like,
[inaudible] sitting right there
so if you want to comment.
>> I would just say that--
[ Inaudible ]
>> Lynn B. Brostoff: In my
experience the sulfa moves
like gang busters more
than anything else.
So we really see that
in the tidelines.
We've seen it with iron going.
Of course, there
are artificial aging
and they're meant to accelerate.
I think human oven aging does
make sense because we don't want
to [inaudible] aging often
changes the mechanisms.
It's artificial aging.
You hope it's giving you an
indication of what might happen
in the future but you see
that's why we look so much
at the historical samples,
because we really thought a lot
of the same kinds of
suggestions and phenomena.
>> Rachel Obbard: It's
worth doing a dry oven run
if we do another batch.
Other questions?
>> Are there any plans
to possibly washout any
of these tidelines
and see what remains
or do a secondary tideline
to see how that compares?
Taking one of these that have
already been established and...
>> Rachel Obbard: And produce
another tideline on it?
>>To see like where the movement
of the elements have gone?
>> Rachel Obbard: Yeah.
Another thing that would be
really interesting to do.
No, we don't have specific plans
to do so but we are planning
to work on this going forward.
And I'll add that to my list.
>> Lynn B. Brostoff: I think
as had been pointed out before,
that one of the reasons we
want to look at timelines is
to understand better what might
occur with local treatment.
So [crosstalk] about that.
So yeah, I think if you are
doing an overall washing,
you know you're doing
something different.
But the work that's been done
shows that even with washing,
there's some lingering
damage that can occur.
So that's what the concern is.
It doesn't seem reversible,
all of it.
>> I was just curious in
the steps too when you try
to eliminate these tidelines,
would you consider including
a conservator with your team.
Because the methods that
we use might not be known.
>> Lynn B. Brostoff:
We did actually have
someone at the Met.
What was her name?
>> Rachel Obbard:
Put me on the spot.
I'm bad with names.
>> Lynn B. Brostoff:
You know her.
>> Rachel Obbard: Lorena?
>> Lynn B. Brostoff: Yes.
She was working with us.
But of course, of right now --
>> Rachel Obbard: I
was depending on her
to convey the expertise
of the conservators,
but I realize she can't
speak for all of you.
>> Lynn B. Brostoff: The grant
funding is over for that grant.
>> Rachel Obbard: But we'll
be applying for some more.
So that's a great idea.
Yeah, we should have a
conservator on our team.
Any other questions?
>> All right, then thank
you very much for coming.
[ Applause ]
>>This has been a presentation
of the Library of Congress.
Visit us at LOC.gov.