From: JOHN PHILO <JOHN_PHILO@amgengate.amgen.com>
  To  : rasmb <rasmb@bbri.eri.harvard.edu>
  Date: 28 Apr 1994 11:07:11 U

XL-A linearity at high abso

  REGARDING           XL-A linearity at high absorbance
Joachim,
     In my experience, the linearity of the XL-A at high absorbance can be
very good, BUT it depends on both optical alignment and lamp intensity.  The
fact that you see a higher MW at longer wavelengths probably indicates that
your lamp needs cleaning.  The intensity of the UV light in the XL-A falls
off with time due to buildup of deposits on the surface of the lamp.  With
fairly heavy use this fall-off can be significant in only a few months.  When
your XL-A was installed your service person should have recorded an intensity
versus wavelength scan with a window-less cell.  You probably should repeat
this procedure, and I suspect you will find that the output below 260 nm has
fallen by a factor of 10 or more.  If so, have Beckman clean the lamp, and
ask them to teach you how to do it yourself.
      When my XL-A is working well, I have good linearity at 280 nm up to
A=2.0, and even when the lamp is dirty it is usually good at 280 to about
A=1.7-1.8.   Linearity at lower wavelengths requires both good lamp intensity
and low stray light.  On my machine, I was not getting good linearity at 230
or 214 nm until Beckman cleaned the lamp AND adjusted the lamp alignment to
reduce the stray light.  After this adjustment the data were linear to above
A=2 at 230 nm and A=1.8 at 214 nm.  Again, as flash numbers accumulate after
lamp cleaning, both the linearity and the signal/noise will degrade.   I
don't try to work below 214 because there is so little light there.
     Also, the numbers I have quoted only apply if there is no absorbance by
the buffer (common to both sample and reference channels).  For example, if
you have an absorbance of 0.5 from the buffer (or possibly from the windows
if you are using old model E windows) it will reduce the linear range by
about 0.5.  Several times people have given me samples and buffers and
neglected to mention that they contained azide or protease inhibitors such as
PEFABLOC which absorb strongly in the far uv.  Since the amounts of these
absorbers in the sample and buffer were well matched, the only real sign that
there was a high background absorbance was that the signal/noise was poor.