This subject obviously concerns many of us.
Let me add my perspective from a relatively large European academic lab.
In our open access NMR operation (8
Spectrometers, 180 users, mostly organic
synthetic Ph.D students and post-docs), we still
use the manual approach and not full automation
(this is mostly for historical and financial
reasons, not because I am opposed to automation
in principle).
The goal of our user training (ca. 2h for each
new user) is very limited, namely:
1) to make sure they do nothing harmful to the
instrument (like contaminating the probehead with
dirty samples, measuring at too high temperatures
etc.), and
2) to make sure they do not waste instrument time
because they do not know how to lock/shim or set
up appropriate experiment parameters.
For the purpose of this training and in order to
improve the flow of information between the NMR
staff and the research groups we use a "superuser
system" where every research group has one or two
users who receive a more thorough training
(including trouble shooting and helping the
others when there is a problem) and it is
normally these superusers who train new group
members. The training of the superusers and
training of users for more complicated
experiments is done by the NMR staff.
I have to add, however, that our undergraduate
students (BSc and MSc) get quite extensive
training in NMR. All have to follow two half-year
courses in empirical spectrsocopy (IR, UV, MS,
NMR) and one half-year practical in analytical
chemistry where one week is dedicated to hands
on-training in 1D NMR (1H, 13C). They have one
course on the theoretical basis (QM) of magnetic
resonance and, in the third year, one optional
course in advanced structure elucidation by NMR.
Of course, most of our Ph.D. students and
post-docs did not do their undergraduate time at
ETH and therefore their NMR knowledge varies from
very good to non-existing when they come here.
Some actively ask for advice by the NMR staff and
develop very good spectroscopy skills, others
still produce very ugly spectra at the end of
their Ph.D. time and couldn't care less.
This brings me to my main concern:
It is the responsability of the group leaders
(faculty) to ask critical questions and to reject
lousy NMRs or -more generally speaking-
experiments done unprofessionally . The big
question is, whether faculty are still willing
and able to take on this responsability under
todays pressures . I observe more and more often
that the supervisor no longer looks at the
primary data but only at results in
"finished/edited" form (as tables or write-ups
for publications).
Recent examples: doing thermodynamics by
measuring equilibrium constants via NMR at
variable temperature without calibrating the
temperature in the NMR; integrating routine 13C
spectra; doing chemical shift vs temperature
curves in D2O without internal reference and not
realising that the lock signal is very
temperature dependent etc. etc.
I am afraid that the percentage of unreliable or
simply false results appearing in the literature
and the time lost by the research groups because
of wrong conclusions drawn from their own spectra
is quite dramatically increasing because of this
development.
As NMR specialists we can improve this situation
by keeping in close contact with the group
leaders and their students and by giving - even
unsolicited - advice. In my experience, the group
leaders and their coworkers are quite thankful
for this and only the very unwise ones complain
about lost time.
It is not so important whether the actual
measurement process is largly automated and
"black box". What is important is that the user
nows how to choose the right experiment and how
to judge the results critically.
Bernhard Jaun
--
Prof. Dr. Bernhard Jaun
Laboratory for Organic Chemistry
ETH Zurich
CH-8093 Zürich
Switzerland
Phone: ++41 1 632 31 44
Fax: ++41 1 632 14 75
E-MAIL: jaun_at_org.chem.ethz.ch
Received on Thu Apr 13 2006 - 11:32:21 MST