From:	Tarek Mamoun El-Gogary <asmasomy-: at :-mans.edu.eg>
Date:	Thu, 31 Oct 2002 08:56:53 +0200
Subject:	summary: solvent cutoff wavelength

Dear CCLer's:
Here is the replies to my question:

Dear CCL members,
> I am sorry to post this general question.
> I am using a double beam spectrophotometer to measure the electronic
> absorption spectra of some organic compounds, so I am using the solvent as
a
> blank and run autozeroing before scanning the sample. My question is:
should
> I stop scanning at the cutoff wavelength (240 nm) of the solvent or may I
> measure the samble beyond the cutoff wavelength of the solvent? As the
> solvent is in the blank cell so its interference will be eliminated, is
that
> correct?
> regards
> Tarek
----------------------------------------------------------------
----------------------------------------------------------------

REPLIES:
Dear Tarek,

Well, if you ran autozeroing with the same solvent
in both cells and use that information in running the sample against
the solvent, you are perfectly OK. One of your test should
proabbly be in the autozeroing step, where you should get
either a rising line or a descending one as you gather
the spectrum from longer to shorter wavelength. The reason
is of course that the all solvents have very large peaks
in the sigma-sigma* region which is under 200nm
and even a tiny difference between the cells, of 0.001cm
will give rise to large absorption differences - you have a
hign molarity of your solvent (~55 for water so A= eps*l*c
can get fairly large).

I hope that this helps,

Amiram

=================================================
Dear Tarek,

There is usually no need to stop scanning at the cutoff point.

However, any results below that point should be treated with caution.
It is possible that you could get spurious peaks in that region; this
depends on what particular type of spectrophotometer you are using. The
older machines with mechanical chopper mechanisms were very prone to
this type of artifact.  Modern ratio recording machines are a lot
less prone to this effect

Remember also that the cutoff is not necessarily a sharp point. My
advice would be to check the actual absorption of the solvent
by running a spectrum  of the solvent (+ cell) against air.

Remember also that you will be able to work at lower wavelengths
if you use a thin cell (1 mm) rather than the usual size (10 mm).

Also read what the manufacturer's manual says about the sensitivity
at low transmissions.

Cheers

Peter Bladon

===========================================================
Hi,

It is also useful to measure the solvent and the sample separately and
extract the solvent spectrum from that of the sample. It can be done by
most of theavailable spectrophotometers with a built-in function.
Best,
Csaba

============================================================
Dear Tarek,

        If the solvent does not transmit light below 240 nm, then any
data collected in that solvent would be meaningless.  If some small
amount of  light is transmitted (the most likely case) then you will
obtain some data below 240 nm, but the noise will be greatly increased.
 You could check the definition of 'cutoff wavelength' to see whether it
means 0.1%, 1% or 10% of full intensity, etc.  Generally, data collected
below the solvent cutoff would be considered unreliable.

        Regards,

                 Steve

=========================================================
Dear Tarek:

My experience with this kind of problem suggests that you can't go much
below the cutoff wavelength for the solvent, even if you are using it as
blank in a double beam instrument.  The problem is that as you get beyond
the cutoff wavelength, you still aren't going to get enough light through
the blank to allow for a really proper correction.  Most instruments will
try to take things as far as you can, but sooner or later you get so little
signal through the blank that the instrument has a hard time trying to
apply any blank correction because the signal is so weak and noisy.  So,
you might be able to push it a bit, but really not too far.

Hope this was of some help.

Brian Williams,Chemistry
Buckenll University

============================================
 I take it you are using UV-Vis.  Generally, one doesn't scan beyond the
cutoffs of the equipment, whether this is the cufoff of the optics, the
glassware (quartz in your case, I suspect, but Pyrex will do for > 290 or
so),  or the solvent.  Even well matched cells will have some differential
absorption for the solvent in this situation.  A good peak deconvolution
program might recover some sample information in this region -- this may be
a function of software provided with modern spectrometers, but not with
older dual-beam instruments.  You should consult the manufacturer's
literature, or some of the standard texts on care and feeding of UV
measurements.

Cheers,
Steve

================================================
Tarek

The problem with measuring the absorbance in a region where the solvent
absorbs is that very little light makes it to the instrument's detector,
even for the blank.  Thus you would be trying to detect the difference
between the very low light intensity for your reference and an even lower
intensity for the sample plus solvent.  In practice, most instruments
allow enough stray light to reach the detector to make the measurement
you're describing suspect at best.  Therefore, the solvent blank does not
assure a meaningful result, and subtracting the spectrum of the solvent
> from the spectrum of the solution is not likely to help.  Adherence to
Beer's law (the linearity of absorbance with concentration) would perhaps
suggest that your results are meaningful.

ciao,
Bruce



Bruce Palfey
Department of Biological Chemistry
University of Michigan
Ann Arbor, MI 48109-0606


*END*
================================



Raw Message Text