6/27/96
They are spectacular! The plates are loaded in the same film holders as
regular film. We ran a test where we spread the the beam out until the
lowest current reading on the screen was indicating. Then stepped the
exposure through two plates using an aperture. The exposure times went
from .12 to 600 sec and all the steps were there without saturating the
plates. We did the same test on Kodak SO-163. >
.../...
> Fuji has a program where you can try out their imaging plates if you
are interested. That is how we got them to try out. >
> - -Scott Walck
Just one question. The plate number is usually printed on the
plates using a photonic system, (seems to be a small screen on which data
are printed, and one or two mirrors and a lens allowing to focus the data on
the plate) which works quite well with usual photographic plates.
In the case on Fuji IP, it seems that it does just not work, because IP
are sensitive to electrons and NOT to photons (I may be equivocated but
I do not think so). Do you know if any improvement is in progress? this
could be a new device developed by EM companies or by Fuji itself,
allowing an auxilliary electron beam to print the relevant data on the
plates, or any other system that I have not though about.
It appears that with many users using the same microscope, it is likely
that plates be mixed up, so that just printing numbers on the back of the
plates might be unsufficient.
Another question. What will Fuji price policy be in the future. At the
moment their plates seem to be rather expensive, so that if we want to
equip 5 microscopes with this system we might think twice...
Yet another one. I would be happy to be able to get the plate out of the
microscope just after taking the picture.
I elaborate: with photographic plates I understand that we have to work in
batches because plates are light sensitive. As a consequence one has to
wait till the plate is processed in order to know if it is good or not.
For this reason camera makers do have a great argument telling that the
image can be computer stored instantaneously.
If you can get the IP out of the microscope and have it processed within
minutes this would be a great improvement. It "just" takes a modification
of the photo box, and I imagine that a good engineer could do that. By
solving these two problems I believe that Fuji would clear any arguments
against their system.
Yves MANIETTE
Universitat de Barcelona
Serveis Cientifico Tecnics
Unitat ESCA TEM
Carrer Luis Sole i Sabaris
E-08028 BARCELONA
Tel +34 3 402 16 95
Fax +34 3 402 13 98
yves@giga.sct.ub.es
The plates are the same size as film. They have a uniform coating of a doped phosphor with other appropriate coatings for antireflections and such. The doping of the phosphor sets up "trapped" energy levels. These energy levels are stable when exposed to electrons (and x-rays) and the number of electrons excited to these levels are proportional to the incident electrons. At room temperature and in darkness (such as the camera chamber), the exposed plates are stable. When we shipped them to ASU, we put packed them with dry ice. After the plates are exposed, they are processed in a "reader". The reader scans a specific wavelength of light over the plates. This excites the electrons in the "trapped" state to a less stable higher energy level which decays with a specific higher energy of light which is read. The intensity of this light is proportional to the electron dosage. The plates are erased fully be exposing them to light. They can be loaded into the camera boxes !
under full lights. They can be ta
ken out of the box after exposure with very subdued light. (about the same as a door cracked
slightly.) Important environmental concern: no wet processing, i.e. no wet chemicals or silver
disposal problems, and the plates are reused.
Working with the plates:
The exposure range was 4 orders of magnitude (not 5 as I said before). The exposure range of film is the same as about the 3rd order for the plates. What this means in practice is that you can use your normal film settings for images. What before would have been very low exposure areas of the film where no detail was seen now has info in it. Jim Darley recently stated, "an image is better when formed by more electrons". From a statistical point of view, i.e. the signal to noise ratio, this is true. The imaging plates allow about an order of magnitude higher exposures than film with the subsequent. The contrast resolution of the plates is 14 bit. When doing diffraction work, it looks like it is relatively easy to saturate the plate. For a convergent beam image, I used too high an exposure settings and the brght field disk was saturated. There was plenty of information in the dimmer areas of the pattern. It was suggested that the exposures for diffraction patterns be!
taken at less times than normally
with film.
Advantages and disadvantages:
We did the exposure test and showed that over the four orders of magnitude the plates were
linear. If you compare that to film, the exposure range for film is about 1 order and it is more
parabolic in shape, being flatter at the low end. Klaus-Ruediger Peters stated,
"Film provides up to 14 to 15 bit of contrast resolution using the AGFA Arcus II scanner ($
2,000). This is an exceptionally high contrast resolution and provides astonishing precision and
quality of the digitized information."
This may be true, but you are still limited in the exposure range and unless a correction is made
for mapping the input to output upon digitization, the result for the film is still not linear. You
are also adding a conversion process with light through the exposed film and a detector to the
digitization process which will further complicate electron dose input to the digitized output
intensity. For the plates, the output is directly proportional to the number of electrons arriving at
the plate.
An application for the plates that is ideally suited for them is with image simulations.
Experimentally, the plates give the electron dose per point, i.e. the contrast is exactly known for
a particular imaging condition. This can then be compared with the predicted value for the
simulation program.
A disadvantage to the plates is that the information box (film number, mag, text, etc.) when
exposed in the microscope is not recorded. This is because that is done with light. If the
microscope manufacturers embrace this technology, this could easily be overcome by recording
the information on a disk and incorporating that to the digital image when they are read. A more
exciting possibility would be to incorporate the reader into the camera chamber itself. For
example, when a new plate is ready to be exposed, an exposed plate could be read. An elegant
design might be to have a turntable that brings two plates or coated areas alternately under the
beam and reader. This arrangement could also lead to larger image formats.
I think that there is a future for imaging plate technology in TEM. If you are in a lab that must
get away from film processing, they can do what film does now and then some. Unfortunately,
the cost of the systems are a little high, and you can't soley justify the switchover based on film,
processing, and printing costs even in a fairly high output laboratory. The cost of computer
RAM, large size permanent storage, and publication quality digital output is sufficiently low
now to almost eliminate those concerns in the switchover consideration.
If you have read this to this point, thank you for listening to me ramble. I'm done now and won't
say another word.
Scott D. Walck
walcksd@ml.wpafb.af.mil
> plates using a photonic system, (seems to be a small screen on which data
> are printed, and one or two mirrors and a lens allowing to focus the data on
> the plate) which works quite well with usual photographic plates.
>
> In the case on Fuji IP, it seems that it does just not work, because IP
> are sensitive to electrons and NOT to photons (I may be equivocated but
> I do not think so).
It is my understanding that IP's *are* sensitive to photons (at least
in some frequency ranges). Fuji would know whether the usual illumination
system would work. We have a xenon flash tube and glass prisms/lenses, so
maybe there are enough high-frequency photons.
> Do you know if any improvement is in progress? this
> could be a new device developed by EM companies or by Fuji itself,
> allowing an auxilliary electron beam to print the relevant data on the
> plates, or any other system that I have not though about.
It would seem that a CRT-type beam could be used to write this way.
> Another question. What will Fuji price policy be in the future. At the
> moment their plates seem to be rather expensive, so that if we want to
> equip 5 microscopes with this system we might think twice...
>
Since the plates are reusable, it is only a one-time expense.
> Yet another one. I would be happy to be able to get the plate out of the
> microscope just after taking the picture.
>
> I elaborate: with photographic plates I understand that we have to work in
> batches because plates are light sensitive.
IP's are also light-sensitive--you should not expose them to fluor-
escent light when taking them to the reader.
> As a consequence one has to
> wait till the plate is processed in order to know if it is good or not.
> For this reason camera makers do have a great argument telling that the
> image can be computer stored instantaneously.
>
> If you can get the IP out of the microscope and have it processed within
> minutes this would be a great improvement. It "just" takes a modification
> of the photo box, and I imagine that a good engineer could do that. By
> solving these two problems I believe that Fuji would clear any arguments
> against their system.
>
This is as easy as removing a single piece of film--no problem on
our scope. The readers I've seen take on the order of a few minutes to
process an IP.
Yours,
Bill Tivol
tivol@wadsworth.org
system there are some points remaining unclear for me. Maybe I am asking
too much...
On Fri, 28 Jun 1996, William Tivol wrote:
> > Another question. What will Fuji price policy be in the future. At the
> > moment their plates seem to be rather expensive, so that if we want to
> > equip 5 microscopes with this system we might think twice...
> >
> Since the plates are reusable, it is only a one-time expense.
this is not perfectly true. Fuji claims their plates are reusable 500
times, so you may have to buy new ones every few years. It comes to be
a bit less expensive than photo plates, but you need to add the reader's
price. Therefore it is more expensive actually.
> > Yet another one. I would be happy to be able to get the plate out of the
> > microscope just after taking the picture.
> > I elaborate: with photographic plates I understand that we have to work in
> > batches because plates are light sensitive.
>
> IP's are also light-sensitive--you should not expose them to fluor-
> escent light when taking them to the reader.
One of the argument of Fuji is precisely that dark room is not necessary
any more. So here is something that I do not understand clearly. On the
other hand I repeat that I have never seen any number printed on any IP,
therefore it must be quite less sensitive than photo plates.
> > As a consequence one has to
> > wait till the plate is processed in order to know if it is good or not.
> > For this reason camera makers do have a great argument telling that the
> > image can be computer stored instantaneously.
> >
> > If you can get the IP out of the microscope and have it processed within
> > minutes this would be a great improvement. It "just" takes a modification
> > of the photo box, and I imagine that a good engineer could do that. By
> > solving these two problems I believe that Fuji would clear any arguments
> > against their system.
> >
> This is as easy as removing a single piece of film--no problem on
> our scope. The readers I've seen take on the order of a few minutes to
> process an IP.
Right. But if you make high resolution images, you may want to know
quickly if the photo was successful or not (focus, drift). Therefore being
able to see the result just after taking the picture can be a real
improvement versus normal photo plates. On a Philips microscope, removing
one plate is the same as removing 36: you HAVE to switch off the HT. this
means that when working in high resolution it is not so easy. The system
I imagine would allow you to get the plate out the microscope within
seconds, without generating any perturbation in the vacuum system.
Yves MANIETTE
Universitat de Barcelona
Serveis Cientifico Tecnics
Unitat ESCA TEM
Carrer Luis Sole i Sabaris
E-08028 BARCELONA
Tel +34 3 402 16 95
Fax +34 3 402 13 98
yves@giga.sct.ub.es
Note: We used these numbers to calculate our exposure times for the current that we were using.
We came close to using the full range on sensivity 50.
- -Scott
- ----------------------------------------
>From Paul Perkes regarding the imaging plates:
More notes:
The files are 3760 x 3000 x 16-bit (14-bits used) unsigned integer raw
files (23MB). These can easily be imported into Spyglass,
DigitalMicrograph, Semper or any other sophisticated image processing
package that can handle 16-bit data (if you have enough memory!).
The details for converting the image plate data to electron dose are
outlined below:
The basic equation is
rho = (200/SE) * 10**((QL-A)/B) * 10E-12
rho = C/cm2
SE = 200 (for 1.0E-14 - 1.0E-10 C/cm2 sensitivity)
100 (for 2.0E-14 - 2.0E-10)
50 (for 4.0E-14 - 4.0E-10)
20 (for 1.0E-13 - 1.0E-9)
10 (for 2.0E-13 - 2.0E-9)
5 (for 4.0E-13 - 4.0E-9)
QL = original IP (log) data
A = (2E+14/2) = 8192
B = (2E+14/4) = 4096
N = rho / ((1.6E-19)*400*400) = number of electrons per 25 micron pixel.
(400x400 25 micron pixels per square centimeter)
In your case SE = 50.
Regards,
Paul R. Perkes (602) 965-5218
Senior Applications Systems Analyst (602) 965-9004 FAX
Center for Solid State Science (602) 440-0803 Pager
Arizona State University paul.perkes@asu.edu
Box 871704
Tempe, AZ 85287-1704