Luminescent
solar
concentrators.
2:
Experimental
and
theoretical
analysis
of
their
possible
efficiencies
J.
S. Batchelder,
A. H.
Zewail,
and
T. Cole
Experimental
techniques
are
developed
to determine
the applicability
of a particular
luminescing
center
for
use in
a luminescent
solar
concentrator
(LSC).
The
relevant
steady-state
characteristics
of eighteen
common
organic
laser
dyes
are given.
The
relative
spectral
homogeneity
of such
dyes
are shown
to
depend
upon
the surrounding
material
using
narrowband
laser
excitation.
We developed
three
independent
tech-
niques
for measuring
self-absorption
rates;
these are
time-resolved
emission,
steady-state
polarization
an-
isotropy,
and spectral
convolution.
Preliminary
dye
degradation
and prototype
efficiency
measurements
are
included.
Finally,
we give
simple
relationships
relating
the efficiency
and gain
of an
LSC
to key
spectro-
scopic
parameters
of its
constituents.
1. Introduction
The
luminescent
solar
concentrator
(LSC)
offers
the
promise
of
reducing
the cost
of
photovoltaic
energy
conversion
by the
use of
high
gain
concentrators
which
do not
require
tracking.
The
conceptual
operation
of
the
LSC
is based
on
light
pipe trapping
of
luminescence
induced
by the
absorption
of
solar radiation.
A trans-
parent
material,
such
as Plexiglas,
is
impregnated
with
guest
luminescent
absorbers
such
as
organic
dye
mole-
cules.
Solar
photons
entering
the
upper
face
of the
plate are
absorbed,
and
photons
are
then
emitted.
Snell's
law
dictates
that a large
fraction
of these
lumi-
nescent
photons
will
be trapped
by
total internal
re-
flection;
for example,
-74%
of
an isotropic
emission
will
be
trapped
in a
PMMA
plate
having
an
index
of re-
fraction
of
1.49. Successive
reflections
transport
the
luminescent
photons
to the
edge of
the plate
where
they
can enter
an edge-mounted
array
of solar
cells.
We previously
discussed
in LSC-11
the primary
issues
governing
the applicability
of LSC
devices;
these issues
are
solar
absorption
bandwidth,
self-absorption
(or
the
reabsorption
of
emission
from a
particular
dye or inor-
ganic ion
by another
similar
emitting
center),
and
photodegradation.
Our
primary
concern
in this
paper
is to
demonstrate
experimental
techniques
and
appro-
priate theoretical
interpretations
for
determining
the
efficacy
of a given
dye in
an LSC
device. We
will also
All authors
are with
California
Institute
of Technology,
A. A. Noyes
Laboratory
of Chemical
Physics,
Pasadena,
California
91125.
Received
8 June
1981.
0003-6935/81/213733-22$00.50/0.
(© 1981
Optical
Society
of America.
investigate
the effects
of the relative
spectral
homoge-
neity,
as determined
by
the type
of substrate
material
used.
This paper
is organized
as follows:
Section
II de-
scribes
spectroscopic
techniques
and results
of mea-
surements
made
on organic
laser
dyes.
Such mea-
surements
include absorption,
emission,
and excitation
spectra,
relative
spectral homogeneity
in a
variety of
host
materials,
spatial
filtering
effects,
and the
polar-
ization
dependence
of the
output
emission.
We
als3
utilize
transient
spectroscopic
methods
to find the
ob-
served
lifetime
of dye
excitations
as a function
of
con-
centration
in the presence
of
self-absorption.
Prelim-
inary dye
degradation
data and
prototype
efficiencies
are also
described.
Section
III continues
the
treatment
of our
previous
paper (LSC-1),
extending
the
analysis
of the
self-absorption
phenomenon
to model
the
ob-
served
spectral
shifts, the
rate of
depolarization
of
the
output
emission,
and
the lifetime
of the
emission
with
concentration.
In
Sec. IV we
show how
these
results
allow
us to
determine
the efficiency
of
a single dye
LSC
plate
as a
function
of the
size or geometric
gain of
the
plate. We
conclude
in Sec.
V with
our perceptions
of
the impact
of the
above developments
on the
potential
utility
of the
LSC.
II.
Experimental
Techniques
and Results
A.
Materials
We
restricted
our study
of luminescent
centers
to the
organic laser
dyes
because
of their high
quantum
effi-
ciency
of luminescence,
their solubility
in methyl
methacrylate
and organic
solvents,
and
their ready
availability.
The source
of
the dyes
was Exciton
Chemical
Co.
2
Dyes
were used
as received
without
1 November
1981
/ Vol. 20,
No. 21 / APPLIED
OPTICS
3733
further
purification.
Liquid
samples
were
made
at
known
concentrations
by
dissolving
quantities
of
dye
weighed
on
a
Cahn-25
electrobalance
in
reagent
grade
methanol.
These
samples
were
stored
in
soda-lime
glass
bottles
and
kept
in
darkness.
The
concentration
of
the
samples
for
absorption
measurements
was
chosen
so
that
the
peak
optical
density
of
a
1-cm
path
length
of
the
solution
between
10,000
and
30,000
cm-'
was
between
0.5
and
1.5.
Spectra
were
also
taken
of
some
of
these
dyes
in
a
variety
of
other
hosts.
The
principal
solid
matrix
ma-
terial
was
polymethyl
methacrylate
(PMMA),
which
usually
contained
5%
hydroxy
ethyl
methacrylate
by
weight
to
increase
the
solubility
of
the
more
polar
dyes.
Large
plates
(1
m
square
X
0.3
cm
thick)
were
com-
mercially
made
to
our
specifications
by
Acrilex
Inc.
3
Smaller
test
samples
were
fabricated
in
our
laboratory
in
the
following
manner:
Aldrich
monomer,
containing
hydroquinone
monomethyl
ether
as
an
inhibitor,
was
purified
by
fractional
distillation
in
a
nitrogen
atmo-
sphere
using
a
vacuum-jacketed
vigreaux
column.
This
distilled
monomer
was
combined
with
technical
grade
hydroxy
ethyl
methacrylate,
and
the
desired
dyes
were
dissolved
therein.
Small
amounts
of
methanol
and/or
acetic
acid
were
added
to
the
solution
to
increase
the
dye
solubility
if
rhodamine
dyes
were
to
be
used.
Some
samples
were
prepared
by
adding
the
concentrated
dye-monomer
solution
to
prepolymerized
PMMA
and
continuing
the
polymerization.
However,
best
results
were
usually
obtained
by
polymerizing
just
the
mono-
mer-dye
solution.
Two
percent
by
weight
of
azobis-
isobutyronitrile
was
added
as
an
initiator,
and
the
mixture
was
poured
into
a
mold
formed
by
two
glass
plates.
The
plates
were
separated
by
a
polyethylene
tubing
gasket
and
by
aluminum
spacers
around
the
periphery
to
maintain
a
constant
plate
thickness.
A
very
thin
coat
of
silicon
vacuum
grease
on
the
glass
plates
acted
as
mold
release
agent.
These
molds
were
200.000
100,000
0X720
CV670
S640
c
KR620
R610
1R590
R575
R560
C540
C500
C1460
10
. 000
WRVENUMBER(S
(/CM)
30.000
Fig.
1.
Composite
of
extinction
coefficient
spectra
for
a
variety
of
representative
organic
laser
dyes.
Vertical
axis
is
extinction
coeffi-
cient
(liter/mole/cm),
and
horizontal
axis
is
in
wave
numbers.
All
spectra
are
from
low
concentration
methanol
solutions.
then
immersed
in
a water
bath
with
an
oil
surface
film
and
placed
in
a
convection
oven.
Polymerization
was
initiated
at
-85
0
C,
when
a
noticeable
increase
in
vis-
cosity
occurred,
at
which
time
the
temperature
was
lowered
to
550
for
48
h,
followed
by
a
final
curing
at
950
for
4
h.
Typically,
significant
fractions
of
the
dye
did
not
go
into
solution,
so
that
the
dye
concentrations
in
the
final
plates
were
assayed
by
measuring
the
peak
optical
density
and
assuming
that
the
peak
extinction
coefficient
was
that
of
the
methanol
solutions.
We
found
that
dye
concentrations
in
excess
of
10
AM
caused
significant
amounts
of
monomer
to
remain
unpoly-
merized
in
the
cured
plates
in
the
case
of
rhodamine
and
oxazine
dyes,
and
that
this
monomer
could
be
slowly
driven
out
by
vacuum
degassing
at
50°C.
After
curing,
the
plates
were
removed
from
the
molds
and
were
scribed
and
broken
to
size.
The
edges
were
polished
with
a sequence
of
grits;
the
final
buffing
compound
was
a
cerium
oxide
rouge.
We
also
developed
an
alternative
technique
of
doping
the
plastic
which
has
the
great
advantages
of
not
re-
quiring
distillation,
casting,
or
curing.
If
a
commercial
transparent
PMMA
plate
or
rod
is
immersed
in
a
methanol
solvent
containing
the
dye
of
interest,
the
dye
will
diffuse
into
the
plate
along
with
the
methanol.
A
solution
of
9%
dichloromethane
by
volume
in
methanol
was
found
to
be
the
best
compromise
between
speed
of
infusion
and
maintaining
good
surface
finish
on
the
plate.
The
time
required
to
achieve
useful
dye
con-
centrations
in
the
plastic
for
a
20-cm
(8-in.)
rod
was
15
min,
and
for
a
40-cm
(16-in.)
plate
the
time
was
.12
h,
both
at
room
temperature.
Coumarin-540
infused
faster
than
rhodamine-640,
possibly
due
to
the
differ-
ence
in
molecular
weight.
It
appeared
to
the
eye
that
the
dyes
typically
resided
in
a
film
between
0.1
and
1
mm
from
the
surface
of
the
plate,
depending
on
tem-
perature
and
soak
time,
so
that
the
infusion
technique
does
not
yield
a
uniform
dye
concentration
across
the
thickness
of
the
plate.
It
is
unlikely
that
the
dye
con-
centration
in
the
film
would
be
greater
than
that
of
the
soaking
solution,
so
that
the
solution
concentration
yields
an
upper
bound
on
the
local
dye
concentration.
Measuring
the
peak
optical
density
and
assuming
a
uniform
dye
dispersion
in
the
plate
gives
a
lower
limit
on
the
concentration
in
the
region
of
the
plate
where
most
of
the
dye
reside.
B.
Steady-State
Spectroscopy
Absorption
spectra
of
all
samples
were
made
using
either
a
Cary-14
or
a
Cary-17
dual-beam
spectropho-
tometer.
These
spectra
and
their
associated
base
lines
were
digitized
using
a
Houston
Instruments
Hipad
digitizer.
The
digitized
spectra
were
corrected
for
base
line
drift
and
stored
on
diskette
using
a
Terak
packaged
PDP
11-03/2.
Figure
1
shows
a
composite
of
a number
of
these
absorption
spectra
plotted
by
a
High
Plot
dig-
ital
plotter
on
the
same
system.
The
digitization
ac-
curacy
was
five-hundredths
of
an
inch,
which
is
finer
than
the
pen
line
of
the
chart
recorder
output.
The
base
line
noise
in
the
spectra
is
typically
1000
liters/
mole/cm
(SNR
100).
3734
APPLIED
OPTICS
/
Vol.
20,
No.
21
/
1
November
1981
LAMP
g OR W
STEPPER
MOTOR
NTR
11-03/2
SUPPLY
LAMP
CONTROLLER
CHOPPER
I LOCK-IN
AMP D
MONOCHROMATOR
R;
oSTEP PERl
MOOR
eRAHC
SAMPLE
MONO-
CHROMATOR
H .V.
|CHART
l
Fig.
2. Schematic
of
the
apparatus
for
emission
and
excitation
measurements.
Regulated
mercury
or
tungsten
source
was
focused
down,
chopped,
and
monochromated
prior
to
exciting
the
sample.
Resulting
emission
was
monochromated
and
detected
by
a PMT,
and
the
signal
was
amplified
with
phase-sensitive
detection.
Remote
computer
controlled
both
monochromators
and
recorded
the
mea-
sured
spectra.
Emission
and
excitation
spectra
of
the
dyes
were
made
at
micromolar
concentrations
using
a computer-
controlled
apparatus,
as
shown
in
Fig.
2.
The
excitation
source
was
either
a 200-W
Oriel
6323
tungsten
lamp
or
a
200-W
Oriel
6137
high
pressure
mercury
lamp.
The
light
was
collimated
with
quartz
optics,
chopped
by
a
PAR
191
chopper,
and
monochromated
by
a
Jarrell-Ash
(model
82-410,
f/3.5,
0.25-m)
monochromator
with
ei-
ther
a 6000-A
blazed
grating
with
1180
grooves/mm
or
a 3000-A
blazed
grating
with
2365
grooves/mm.
For
the
phase
sensitive
detection
to
produce
accurate
excitation
and
emission
spectra
when
chopping
the
excitation
source,
the
lifetimes
of
the
dyes
must
be short
compared
with
the
chopping
period
(2.5
msec
in these
experi-
ments).
This
condition
was
fulfilled
in
all
these
mea-
surements.
The
excitation
linewidth
was
fixed
at
90
A
across
the
tuning
range.
The
emission
90°
from
the
excitation
was
analyzed
by
a similar
Jarrell-Ash
monochromator,
in
this
case
with
an adjustable
resolution
of
2-90
A.
The
output
light
was
detected
by a
928
Hamamatsu
PMT
biased
at
900
V;
the
PMT
output
was
terminated
by
100k
Q
parallel
with
a PAR-HR8
lock-in
amplifier.
The
analog
output
of
the
lock-in
was
digitized
recorded
by
the
PDP
11-03/2.
Care
was
taken
to keep
the PMT
current
at
least
20%
below
its rated
output
current
of
100,glA.
The
RC
time
constant
of the
lock-in
was
kept
at
least
as
short
as
the
time
between
digital
sampling
of
the
emission.
These
samples
were
taken
typically
every
10
A.
Both
monochromators
were
driven
by
a
Slo-Syn
SP151
driver
so
that
either
one
or both
monochromators
could
scan
under
the
control
of
the
PDP
11-03/2.
This
allowed
acquiring
spectra
in
cm-1
even
though
the
monochromators
had
wavelength
drives.
Computer
control
also
provided
automatic
backlash
correction.
Frequency
calibration
of
the system
was
done
with
the
5451-A
line
from
a
mercury
germicidal
lamp,
and
the
system
response
was
measured
using
an Eppley
Labo-
ratory
calibrated
EPI-1669
quartz
halogen
lamp
in
a
custom-made
housing.
The
output
of
the
HR-8
was
digitized
in
a 12-bit
analog-to-digital
converter
built
around
the
Intersil
ICL-7107.
This
generated
a parallel
digital
input
to
the
remote
PDP
11-03/2.
The
system
response
function
was
stored
on diskette;
this
was
de-
termined
to be
the
response
of the
system
to the
cali-
brated
tungsten
lamp
divided
by the
known
spectrum
of the
tungsten
lamp.
Emission
spectra
were
corrected
by
dividing
out
this
system
response
and
were
normal-
ized
so that
the
luminescence
integrated
over
all
wave
numbers
equals
1.
The
peak
position
of the
emission
spectra
for the
dyes
tested
is given
in Table
I. Cor-
recting
the
excitation
spectra
for the
response
of the
system
requires
a calibrated
detector
as
well
as
a cali-
brated
source.
Using
both
it
is possible
to
determine
the response
of
the monochromator
that
is scanning
the
excitation.
We
made
the
approximation
that
our
PMT
response.was
flat over
the
region
of interest.
In
this
approximation,
the
variation
of
the intensity
of
the
scanning
excitation
is
the
measured
response
to the
calibrated
tungsten
lamp.
Excitation
spectra
were
corrected
by
dividing
out
this
measured
response.
Both
the emission
and
excitation
spectra
of these
dyes
were
taken
only
at micromolar
dye
concentration
to
minimize
the
red-shifting
effect
of self-absorption.
The
optical
density
of the
sample
in the
region
where
the
absorption
and
emission
overlap
must
be sufficiently
low
that
the
blue
tail
of
the
emission
spectrum
is
not
artificially
filtered
out.
The
excitation
samples
must
be
dilute
for
a second
reason.
Away
from
the
peak
of
the
absorption,
where
the
absorption
per
unit
length
is
small,
the
exciting
beam
intensity
does
not
vary
greatly
along
the
beam
length
in
the sample.
However,
as
the
excitation
is
scanned
across
the peak
of
the
absorption,
most
of
the exciting
light
is
absorbed
in a thin
surface
layer
in
the
sample
if
the
concentration
is too
high.
This
effect
changes
the
spatial
distribution
of
the
emission
and
alters
the
observed
emission
spectrum.
Measurements
were
also
made
on
the
spectral
char-
acteristics
of
multiple
dye
solutions.
Dyes
could
not be
combined
indiscriminately,
possibly
due
to
agglomer-
ation
phenomenon
between
some
dyes
which
quenched
the
emission
when
they
were
in solution
together.
For
example,
our
oxazine-720
or oxazine-750
solutions
showed
quenching
of their
emission
when
they
were
in
a solution
with
rhodamine-610
or rhodamine-590.
An
example
of
a successful
multiple
dye combination
is
sulforhodamine-640,
rhodamine-590,
and
coumarin-
540.
In
Figures
3
and 4
we show
emission
and
excita-
tion
spectra,
respectively,
for an
oxazine-720
methanol
solution
and
for
an oxazine-720-rhodamine-640-cou-
marin-540
methanol
mixture
at micromolar
and hun-
dred
micromolar
concentrations.
The
emission
spectra
of
the multiple-dye
solution
at low
concentration
are
dominated
by
rhodamine
and
coumarine
emission,
while
at high
concentration
the
oxazine
dominates
the
emission.
If we
detect
the
emission
intensity
of oxazine,
as
shown
in
Fig.
4, as
a function
of excitation
energy,
the
1 November
1981
/ Vol.
20, No.
21
/ APPLIED
OPTICS
3735
Table
1.
Optical
Properties
of Dyes
used
In
the
LSCs
-
Stoke
Shift
Exciton
Dye
max
oe(
max
(fmax)
V(fmax)
max
rax)
-
'(max)
CODE
(Kodak
name)
in
methanol)
cm,
A
cm-'
cm-,
Solvent
±
30%
Coumarin-480
22,000
25,730
3,890
21,200
4,720
4,
560
0
.
58
a
110.
(Coumarin-102)
ethanol
Coumarin-500
19,
900
25,
660
3,900
20,
200
4.
950
5,440
0.
5
3
a
140.
ethanol
Coumarin-535
52,200
23,120
4,320
20,400
4,900
2,720
90.
(Coumarin-
7)
Coumarin-540
52,
200
21,
860
4,
580
19,
700
5, 070
2,130
0
78a
80.
(Coumarin-6)
ethanol
DCM
28,
900
21,
500
4, 650
15,
700
6,360
5, 760
0.
71
b
240.
DMSO
Rhodamine-560
82,000
20,120
4,970
19,000
5,250
1,080
0.
85c
25.
(Rhodamine-
10)
ethanol
Rhodamine-575
93,800
19,330
5,170
18,300
5,460
1,010
35.
Rhodamine-590
107,000
18,940
5,280
18,000
5,550
900
0.98d
25.
(Rhodamine-6G)
methanol
Rhodamine-610
114,
000
18,
380
5,440
17,
500
5, 710
870
o
5e
36.
(Rhodamine-B)
ethanol
Kiton
red-620
111,
000
17,990
5,560
17,300
5,
800
740
0
.
8
3
g
16.
(Sulforhodamine-B)
ethanol
Rhodamine-640
106,000
17,670
5,660
16,800
5,940
830
.
17.
(Rhodamine-101)
ethanol
.Rulforhodamine-640
120loOO
17,360
5,760
16,700
6,000
690
h
17.
(Sulforhodamine-
1 01)
ethanol
Cresile
Violet-670
57,
900
16,
880
5, 920
16,100
6,
220
810
0. 54f
17.
(Oxazine-9)
methanol
Oxazine-720
81,
800
16,170
6,190
15,
600
6,420
600
17.
(Oxazine-
1 70)
Oxazine-750
90,600
15,140
6,600
14,500
6,920
680
25.
DODCI
238,
000
17,160
5,
830
16,
500
6,050
630
11.
DOTCI
236,
000
14, 770
6, 770
14,000
7,140
770
8.
IR-144
153,000
13,560
7,370
12,000
8,340
1,560
16.
RG.
A.
Reynolds,
K.
H.
Drexhage,
Optics
Comm.,
13,
No.
3, 222
(1975).
b
P.
R.
Hammond,
Optics
Comm.
, 29,
No.
3,
331
(1979).
CK.
H.
Drexhage,
"Structure
and
Properties
of
Laser
Dyes,
"ed.
F.
P.
Schafer,
Topics
in Dye
Lasers,
Applied
Physics
I (Springer
New
York,
1977)
p. 144.
dA.
Baczynski,
T.
Marszalek,
H.
Walerys,
B. Zietek,
Acta
Phys.
Polo.,
A44
,
No.
6,
805
(1973).
eT.
Karstens,
K.
Kobs,
J.
Phys.
Chem.,
84,
No.
14,
1871
(1980).
1
D.
Magde,
J. H.
Brannon,
T. L.
Cremers,
J. Olmsted
I,
J.
Phys.
Chem.,
83,
No.
6, 696
(1979).
J.
M. Drake,
R.
T.
Morse,
R.
N.
Steppl,
D.
Young,
Chem.
Phys.
Lett.,
35,
No.
2, 181
(1975).
hC.
F.
Rapp
et
al.,
Final
Report
of Owens
Illinois,
Sand77-7005,
p.
40.
3736
APPLIED
OPTICS
/ Vol.
20, No.
21
/
1 November
1981
ONE DYE T
, LOW CONC.
3 DYES
T
,HIGH CONC.
3 DYES T
I , , , , LOW CONC.
11.000 16,000 21,000
WPVENUMBERS
Fig. 3. Three emission spectra of methanol dye solutions resulting
from 22,220-wave number (4500-A) excitation. Top spectrum is from
a micromolar oxazine-720 solution, and the bottom two are micro-
molar and hundred
micromolar solutions, respectively, of couma-
rin-540, rhodamine-640, and oxazine-720.
ONE DYE T
I I I LOW CONC.
3 DYES T
,{A ,HIGH, COIGH CONC.
3 DYES AT
I , ,J , , ~ LOW CONC.
13,000 18,000 23,000
WRVENUMBERS
Fig. 4. Three excitation spectra of methanol dye solutions due to
15,630-wave number (6400-A) detection. Three spectra correspond
to the same three solutions described in Fig. 3.
Dewar
Fig. 5. Diagram of the apparatus used to measure the luminescence
spectra of liquid, cast plastic, and diffused plastic samples as a func-
tion of excitation energy and temperature. Either the argon-ion laser
or the tunable dye laser was used as the excitation source. Apparent
sample temperature was maintained at either 77 or 300K. Resulting
emission was analyzed by a Spex 14018 double monochtomator.
HN C
2
H
5
,#
I
C3
\ /O K, H C N
2H5
1 I I I
K~ ~~ "_. .; ~C
/
l I I
Fig. 6. Structure of rhodamine-575.
low concentration
multiple dye solution is dominated
by the pure oxazine absorption peak, while
the high
concentration
multiple dye solution gives a nearly
constant response across the
visible spectrum.
C. Spectral Inhomogeneity
A primary objective of our work was determination
of the homogeneity of the absorption and emission
spectra in solution, in cast plastic, and in diffused plastic
and to determine the temperature dependence of the
anti-Stokes shifted emission. We shall call this the
spectral inhomogeneity of the dye in these various cases.
To achieve these objectives the luminescence spectra
were taken using the apparatus shown in Fig. 5. A
Spectra-Physics 160 argon-ion laser was used directly
as the exciting source or served as a pump for a Spec-
tra-Physics 375A dye laser. The dye laser operated
with a glycol solution of rhodamine-590. When we
excited the sample at an energy considerably below its
peak absorption, it was important that very little
spontaneous emission of a higher energy than the
principal lasing wavelength be present in the sample
excitation beam. The dye laser was therefore followed
by several sharp cut filters, a dispersive prism, and a slit
aperture. Neutral density filters were used to adjust
the input power to the sample. The samples were held
in a glass Dewar with windows positioned for both the
excitation beam and for the resulting right-angle
emission. The luminescence spectra were taken both
at room temperature and with the samples immersed
in liquid nitrogen. Since the excitation was not
chopped, it was important to eliminate stray room light
as a source of excitation. The luminescence resulting
from the laser excitation was collected by a Spex 1419A
sample illuminator and then analyzed by a Spex 14018
double monochromator with 2400-groove/mm holo-
graphic gratings. The sample illuminator was a col-
lection optics stage, originally designed for Raman
spectroscopy, which contained the dispersive prism, slit,
sample translation stage, and matched f/No. optics for
the spectrometer. A polarizing filter in the collection
optics was used to suppress scattered laser light, fol-
lowed by a depolarizing prism to compensate for the
polarization response of the gratings. The spectrally
resolved output was measured by a Hamamatsu 955
photomultiplier and a Spex DPC-2 photon counter.
The final chart recorder output was digitized and stored
in the PDP 11-03/2. System response correction was
effected by digitizing the measured
spectrum of a cali-
brated
tungsten lamp, as described in the previous
section.
Rhodamine-575 was chosen for study because it can
be excited on the extreme red edge of its absorption
band by the rhodamine-590 dye laser. The structure
of rhodamine-575 shown in Fig. 6 is also very similar to
other xanthene dyes, so that the results of these mea-
surements should be representative of most of that class
of laser dyes. The liquid sample studied in this manner
was 18-gM rhodamine-575 in methanol. This was
mounted in a 1.3-mm o.d. capillary tube, oriented so as
to minimize the path length of the output emission.
1 November 1981 / Vol. 20, No. 21 / APPLIED OPTICS 3737
I A- I I .
I I -