Greetings:
We are in the process of planning for a new NMR facility that will
contain both high resolution and solids instruments at low to high
fields. A number of questions have come up and I would like to pose
them (I apologize for multiple listings). I have summarized some
information I have found on the Bruker list and on AMMRL.
Bruker list:
After looking through the archives, I have found a reference in Dec
98 regarding magnet lacations and vibrations and in Dec 99 regarding the
relative heights of He/N2 dewars and a 800MHz magnet for a new site. If
I have missed a note regarding site planning, please let me know.
AMMRL Archives:
I read Nathan Dubree's summary of replies to his request for advice
on planning a new NMR facility (Nov. 1999) among other posts in the
AMMRL archives. Thank you for the information! A (very) short summary
of current information precedes the questions in the long version
below. The short version contains the 10 questions only. Hopefully,
these questions/requests will build upon what's in the archives.
If you have had a memorable experience (good or bad) in planning a
new NMR facility, I would be grateful to hear about it.
Thanks in advance!
Marc ter Horst
--
Department of Chemistry Office: (919) 843-5802
CB #3290 Venable Hall NMR Lab: (919) 962-1149
UNC at Chapel Hill Department fax: (919) 962-2388
Chapel Hill, NC 27599-3290 e-mail: terhorst@unc.edu
--------------------------------------------------------------------
Here is the short version:
Is 9.4T (400MHz) a good field strength 'cut-off' for vibration
isolation either through anit-vibration legs and/or a 'vibration-free'
building foundation? That is, should all systems above 9.4T have
vibration isolation? What about 9.4T and above?
What are convenient units/measure of building vibrations and
what
values are acceptable for routine and more demanding (multi-D) NMR
experiments?
Are the benefits of magnet/console bays worth the added costs
for
construction?
Are recessed magnet bays beneficial for high resolution AND
solids
work?
What simple formula/recipes have people used to determine
adequate
environmental controls (temperature, humidity and air flow)? Maybe in
terms of how these variables may affect experimental results?
If you use oxygen sensors in your facility, do you feel it is
worth the
extra cost/maintenance?
Has anybody worked through the numbers in consideration of
reliquification of cryogens given today's costs for He and N2 delivery?
Are permanent cryogen transfer lines a practical consideration
for a
room with many magnets? What drawbacks exist (cost, maintenance of
transfer lines, etc.)?
What drawbacks have people encountered when placing magnets
within
their 5G lines with opposing magnetic dipoles? How often is one magnet
deemed not reliable while work is done on its neighbor, say a cryogen
fill at one magnet affecting the field at the other? Would one expect
to see artifacts in long runs and multi-D work while the neighboring
magnet is in normal or routine use?
What experiences have any of you had with iron shielding? How
often do
you see artifacts due to temperature changes or vibrations in the
shielding? Are long runs and multi-D experiments more susceptible to
these artifacts?
The long version:
I have not included issues regarding sufficient access for
magnets/dewars since this is available from the vendors. I haven't
looked into compressed air and N2 gas alternatives although people have
recommended quiet scroll-type compressors and to make sure that
condensation pipes drain properly. Dedicated circuits have been
recommended for spectrometers and indeed we are considering a large UPS
system to handle the entire facility. Has anybody considered solar
energy (maybe someone in California...)? One should also consider what
colors to use for the floors to better detect metal chips, paper clips
and other unwanted debris (and whether you want carpeting or a sticky
pad be the door). One may also consider what color the walls should be
since a dark wall can help you see N2 and He plumes. Perhaps these
concerns can be addressed in a future note.
1) Vibration
Anti-vibration legs (e.g., TMC) appear to be a must for fields above
9.4T
(400MHz), allowing these systems to be located on 3rd floors and
higher. (Note: the anti-vibration legs/tables seem to isolate the
magnet from frequencies down to less than 10 Hertz.) However, effects
from vibrations can be found in the baselines of spectra acquired at
400MHz where vibration isolation has not been used or the anti-vibration
table is 'turned off'. Indeed this seems to be acceptable for routine
1H and 13C spectra but may introduce artifacts in 2D spectra such as the
ever-problematic t1 noise.
a) Is 9.4T a good field strength 'cut-off' for vibration isolation
either through anit-vibration legs and/or a 'vibration-free' building
foundation?
Speaking of building foundations, there must be a convenient way to
speak about the magnitude and frequency range of vibrations in a
building. I have heard people talk about 'micro inches' or dB's as a
function of frequency.
b) What are convenient units/measures of building vibrations and
what values are acceptable for routine and more demanding (multi-D) NMR
experiments?
2) Magnet bays
I have seen and heard about a number of facilities which house the
magnets and consoles in glass enclosed rooms or bays thus providing
benefits for environmental controls (temperature, humidity, and cryogen
boil off) and sound isolation. What a difference it makes to not have
all those fans buzzing in your ears! It also helps ensure that only
people who need to be close to the magnet actually get there (a
permanent version of the plastic chain/fences we have seen). One can
also lower lighting costs if the lights within the bays are on a
separate circuit and turned on only when needed (track lighting uses a
lot of electricity and the costly lights require frequent replacements).
a) Are these benefits worth the added costs for construction?
Lowering of the floor within the magnet bay makes it convenient for
users to insert samples and for staff to do cryogen fills (assuming a
platform is built to support the dewars). These recessed magnet bays,
in
principle, allow for adequate space between magnets and fluorescent
light bulbs (recommended to be >10 ft away and to use newer, all
solid-state ballasts, see AMMRL: Dec 1998). They also appear to 'open'
up the room but may cause access problems, for example, with probe
tuning and probe changing. It may not be worthwhile to have a solids
instrument in such a recessed bay.
b) Are recessed magnet bays beneficial for high resolution AND
solids work?
3) Temperature, humidity, airflow
Charles Fry has recently (AMMRL: May 2001) posted an excellent summary
of
temperature control issues. The conclusion being: push for better than
+/- 1 degree C. Humidity is usually controlled by 'house' air
conditioning or a dedicated air handler (such as the Liebert systems).
Airflow or circulation is also not often specifically considered but
might become an issue should a magnet quench.
a) What simple formula/recipes have people used to determine
adequate environmental controls (temperature, humidity and air flow)?
Maybe in terms of how these variables may affect experimental results?
Unfortunately magnet quenches happen and they do pose a risk for
suffocation. It seems that O2 sensors might be a consideration for the
safety of ourselves and our users. My guess is that the sensors require
some maintenance and maybe they are not that critical since a quench is
usually obvious anyway and people would not stick around to check an O2
sensor.
b) If you use oxygen sensors in your facility, do you feel it is
worth the extra cost/maintenance?
4) Cryogens
There has been talk about recovering He and N2 boil off and
reliquification for re-use. It sounds like this is done more often in
Europe. From a previous posting (AMMRL: Jul 1997), it appears that it
is not practical to re-use He boil off even for a facility producing
700liters He gas per week.
a) Has anybody worked through the numbers given today's costs for
He and N2 delivery?
Currently we have a room that can hold more than 5 magnets and indeed
with actively shielded magnets this will become more common. It seems
that it might be good to have liquid N2 and maybe He delivered through
insulated tubing. This way, cryogen fills are done by simply connecting
the fill port on the magnet to a N2/He source line from the ceiling.
One then avoids moving dewars around the magnets. I thought the big NMR
lab in Japan does this.
b) Is this a practical consideration for a room with many magnets?
What drawbacks exist (cost, maintenance of transfer lines, etc.)?
5) Magnet positions
You probably guessed this would be an issue with any new facility: how
many magnets can be fit in one room? It is often suggested that magnet
5 Gauss lines should not overlap. Indeed the previous summary suggests
that this is even more important for actively shielded magnets. Of
course, one also needs to make sure that the vertical 5 Gauss (many
suggest 2 Gauss) lines are free from metal (most important is moving
metal), non-NMR related computer monitors and people not involved with
NMR. One way to circumvent the 5 G restriction is to orient the
magnetic fields of neighboring magnets in opposing directions. Then
placing the magnets within their respective horizontal 5G lines reduces
the space required since the fields don't add and also can reduce the
extent of stray fields in other directions.
a) What drawbacks have people encountered when placing magnets
within
their 5G lines with opposing magnetic dipoles? How often is one magnet
deemed not reliable while work is done on its neighbor (say a cryogen
fill at one magnet affecting the field at the other)? Would one expect
to see artifacts in long runs and multi-D work while the neighboring
magnet is in normal or routine use?
Iron shielding has also been used to contain stray magnetic fields (see
AMMRL: Feb 1997 and May 1999). These include iron plates in walls and
ceilings and Faraday cages. I have heard of a Magnex imaging magnet
within an iron cage (I think it was a horizontal, 500MHz imaging magnet
in Florida). A magnet engineer told me that temperature variations or
vibrations in the iron shielding can cause field changes in the NMR
magnet and that this arrangement can only work well with very careful
planning. (It appears that one needs to have a vibration free building
and well controlled ambient temperatures.) Indeed, an AMMRL member
mentioned that there are only a "few cases where passive room shielding
has been completely successful". Passive shielding often reduces stray
fields (although the thickness/size/type/location of the metal may be
challenging to determine), however, what effects (if any) can one expect
on NMR spectra?
b) What experiences have any of you had with iron shielding? How
often
do you see artifacts due to temperature changes or vibrations in the
shielding which can couple into the NMR magnet? Is there any direct
evidence of passive shielding affecting long runs and multi-D
experiments?
This archive was generated by hypermail 2b29 : Mon Jan 21 2002 - 18:08:05 PST