Green Lasers and Fluorescein (or: Get rid of that 530/30!)

Howard Shapiro hms at
Mon Jan 21 19:43:25 EST 2008

Mario wrote:
> This message is solely intended for those of you who are outfitting a
> new instrument, or retrofitting an existing one, and are thinking about
> green or yellow-green lasers.
> I recently visited some laboratories that are acquiring new LSR II
> instruments, and was told that BD is recommending that they equip the
> instruments with "yellow-green" lasers (~560 nm) rather than green
> lasers (532 nm).  The stated reason for this is that the green laser has
> been observed to result in deteriorating fluorescein (FITC) measurements.
> The big reason to use a green laser is because of the significantly
> better detection of PE and PE tandems.    This comes about for two
> reasons -- one is the increased excitation efficiency at 532 nm
> (compared to 488 nm), and the second is the availability of high-power   
> lasers (200 watts).  In multicolor experiments, we find as much as
> 10-fold increased sensitivity on these channels using a high-power green
> laser compared to low-power blue laser.  (The benefit over a high-power
> blue laser is still significant, albeit less so).
> It is possible that the green-yellow laser may provide this advantage as
> well; I'm not sure because we haven't done the comparison.   
> Certainly, if you get a 560 nm laser at less than 200 watts, I would
> expect it to perform less well than the 200 watt 532 nm laser.
> So why the recommendation for the switch to a the green-yellow laser?   
> Because of the poor FITC results.  But changing lasers is not the
> solution -- changing filters is!  Most FITC measurements are made with   
> a 530/30 nm filter -- a filter that is nearly centered on the green
> laser.    Thus, the high powered green laser provides all sorts of stray   
> light that contaminates this measurement (off the blue laser), and leads
> to high background.  There are two solutions to this:  one is a "notch"
> filter that selectively blocks the 532 light; the other is to use a
> shorter bandpass filter.
> First, why are manufacturers supplying a 530/30 filter for FITC?  FITC's
> emission peak is ~512 nm -- the 530/30 is only collecting less than half
> of the tail of the FITC emission!  We should use filter that	 
> is much closer to the peak.  The only reason 530/30 filters are still
> supplied is purely historical inertia. Years ago, when filter technology
> was less advanced, a filter bandpass needed to be far from a laser line
> to block the laser light -- i.e., anything closer to the 488 than a
> 530/30 let through some of the blue light.  But filter technology is
> vastly better these days; we can come much closer to the   
> laser line and still block it out.
> Currently, we use a 515/20 filter for FITC.  Not only does it
> efficiently block the 488 nm line, it also efficiently blocks the 532 nm
> line.  AND... it collects more FITC fluorescence than the outdated
> 530/30 filter.
> As for the notch filter solution:  don't use it!  Why take less than
> half of the available fluorescence, and eliminate a large fraction of
> it, simply to block the green laser light!  It's a silly solution that   
> costs much more than the relatively inexpensive 515/20 nm filter (which
> can be ordered from any filter company).
> There are reasons to get a yellow-green laser (for example, to detect
> certain fluorochromes or fluorescent proteins).  But don't get it
> because the green laser makes FITC look worse -- that's not a valid reason.
> Finally .... if you get a high power green laser (or even if you get a   
> yellow-green laser), then you should use a low-power blue laser (NOT the
> high power blue laser).  The high-power blue laser helps for detecting
> PE and PE tandems, but it does not help with FITC, and it hurts
> significantly with PerCP.  If you use the green laser for PE etc., then
> get the cheaper low power blue laser: you will get as good (or better)
> results.
> PS, the information about the green laser, PE tandems, and the use of
> alternative FITC filters is fully described in Perfetto et al.,
> Cytometry A, 71, 73-9 (2007).

First of all, I think Mario dropped a "milli-" in a few places and meant
200 mW lasers, not 200 W lasers.

I otherwise endorse what he said, particularly about being able to get
by with relatively low power at 488 when you use a green laser to excite
PE and tandems. It's generally a good idea to "titrate" laser power as
you would titrate antibody concentration; the relationship between
excitation power and fluorescence signal is only linear up to a point,
and, with some labels (notably PerCP), increasing laser power may
actually decrease signal (with PerCP, this is due to formation of
triplet states).

I should also stress that in general, but particularly when you are
trying to get filters with center bandwidths close to laser lines that
might otherwise interfere with your detection (e.g., the 515/20 filter
for fluorescein (488-490 excitation) when you are also using 532 nm
excitation), you need to get and look at an O.D. (optical density) curve
for the filter, not just the usual per cent transmission curve, to make
sure the filter will block the laser wavelength you don't want. The
filter makers will provide O.D. curves if you ask for them. Some filters
that block potential interference near the passband have higher
transmission farther away; in general, you want O.D. of at least 6
outside the passband (some curves don't go higher than 5, but that's
usually OK if there are no drops below that line outside the passband).


More information about the Cytometry mailing list