Technical Information

Advanced Optics Version April 2022

TIE-36 Fluorescence of optical glass

Introduction
1. O
 rigin of fluorescence in glass . . . . . . . . . . . . . . . . . . . . 1
Fluorescence is a phenomenon in which a substance absorbs
light of a certain wavelength and almost instantaneously radi- 2. M
 ethods of Characterization . . . . . . . . . . . . . . . . . . . . . . 2
ates light at longer wavelength (lower energy). These pro-
3. Accuracy of method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
cesses are known as excitation and emission. Many sub-
stances, organic as well as inorganic, can exhibit fluorescence. 4. Evaluation of the integral fluorescence . . . . . . . . . . . . 4
In life science applications like microscopes, fluorescence
5. Fluorescence results for optical glass . . . . . . . . . . . . . . 6
plays a major role in the optical design. The “primary fluores-
cence”, or “auto-fluorescence” of an optical material in a lens 6. Fluorescence quantum efficiency measurement
system leads to a significant problem in the application. In flu- set-up [1] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
orescence microscopy it is desirable to have a very low back-
7. F luorescence quantum efficiency measurement
ground fluorescence. In such applications, fluorescing optical
results [1] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
materials may decrease the contrast of the image. Especially
in applications, where the fluorescence of fluorophores or 8. Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
dyes is detected, the fluorescence of optical components is
generally not desirable, as it falsifies the detected signal (fluo-
rescence of optical components as major source of unwanted
background light in an image).

1. Origin of fluorescence in glass emission. This energy is absorbed as phonons (solid state
vibrations), as heat or occurring as long living phosphores-
Excitation of an optical material at a given wavelength will
cence emission as mentioned earlier.
primarily result in partially absorption and transmittance of
the light. Most part of the absorbed light is transformed
Fluorescence in optical glass is for instance generated by
into vibrational energy, generating heat, but a part of it can
point defects like color centers, which are known to be origi-
be re-emitted as light of different wavelength (luminescence
nated by rare earth elements, and certain impurities [2][3]
processes). Fluorescence is a luminescence process in which
[4]. The observed UV excited fluorescence correlates strongly
part of the absorbed light is re-emitted after ns to µs at
with the content of UV absorbing ions in the material like
longer wavelength (lower energy) and is known as Stokes
rare earth elements (REE), Iron, Zinc or Vanadium [5] and also
shift. Many substances, organic as well as inorganic, can
Pb2+. Most of these elements absorb and emit nearly inde-
exhibit fluorescence. Fluorescence can be distinguished
pendent (especially the REE) from the glass matrix.
from phosphorescence by the lifetime or decay of the
excited states, which is e.g., typically in the ms to s range
Optical transitions from the ns2-levels (n = 1,2,3,..), related to
for phosphorescence [1].
structural units of the glass matrix, like e.g. SbO3, AsO3 or
ZnO, behave in a different way. Here the glass matrix and the
The efficiency of energy transfer of the absorption light, known
processing (oxidation and reduction) have a strong influence
as quantum efficiency, strongly depends on the excitation
on the absorption and emission characteristics.
wavelength or wavelength spectrum. It has a maximum close
to the absorption bands and high photon energy respectively.
➜
The fluorescence is in contrast to the non-radiative absorption
where the incident light is transferred to a higher energy level
and the down conversion to lower energy levels without light

1
 Technical Information
Advanced Optics

TIE-36 Fluorescence of optical glass

The fluorescence behavior of a glass can be influenced by Standard fluorescence measurement procedure at SCHOTT is
either process technology or use of higher purified raw mate- the relative integral fluorescence measurement that generates
rials. In general the fluorescence of optical glass is very low a %- value fluorescence in comparison to the integral fluores-
in the spectral range of the application because of the very cence of the reference optical glass SF1 containing a very
high transmittance and therefore low absorption of optical high amount of Pb2+ as fluorophore.
glasses in the visible spectrum. This is in contrast to filter
glasses that have a desired blocking characteristic in certain
ranges of the visible spectrum, here the absorption can lead
to fluorescence [6].

2. Methods of Characterization
The measurement procedure shown in figure 1 works as The Fluorolog-3 system is part of SCHOTT’s chemical and
follows: On the excitation side a 450 Watt high pressure Xenon physical testing laboratories that are accredited within
lamp delivers a broad band radiation spectrum from 200 to D-PL-14645-01-00 according DIN ISO IEC/ 17025:2019.
1500 nm with a maximum intensity between 250 and 650 nm. The typical emission spectral range is 300 nm to 850 nm.
This is the wavelength range that is mainly used for excitation.
The radiation is spectrally selected and guided to the sample
450 Watt high Double
box by a Czerny Turner double monochromator which allows
pressure monochromator sample
a high noise reduction in order to enable a signal to noise Xenon lamp for excitation
ratio better than 4000. The spectral linewidth of the excitation
is 5 – 7 nm FWHM (full width half maximum) having an inten-
sity of approximately 1 µW/cm2 On request LED or Laser lights 28°
with higher intensities are possible as excitation source with
linewidths of 20 – 30 and 2 – 10 nm FWHM respectively. The
modular setup of this equipment allows an easy change of
Double
the excitation source. monochromator
for emission
The fluorescence light emits in all directions, but is either
Fluorescence spectrum
measured in scattering arrangement in reflection at an angle
of 28° to the incident light (14° to the normal of the plane
Fig. 1: Setup of the fluorescence spectrometer Jobin Yvon Fluorolog 3.
of the sample) to avoid the incident (excitation) light in the The arrangement allows flexibility in geometry and excitation source.
fluorescence measurement signal. The spectral fluorescence Due to the double monochromator concept no filters are required.
intensity is evaluated using a second Czerny Turner double
monochromator and a photo multiplier as UV/VIS detector.
Alternatively, pulsed excitation and a corresponding detec-
tion are also possible. Besides the intensity of the excitation
source, the optimization of the detection efficiency depends
on the proper choice of geometry and cut-off filter settings,
which are required to have no intrinsic fluorescence at the
applied wavelength. ➜ ➜ Back to index

2
 Technical Information
Advanced Optics

TIE-36 Fluorescence of optical glass

3. Accuracy of method
For quantitative measurements the grating efficiency has to
be evaluated using different types of calibrated standards
from Starna (Starna Company (Essex, UK) and BAM (Bunde-
sanstalt für Materialforschung und -prüfung [Berlin, Ger-
many]). The grating consists of 1200 lines/mm and has maxi-
mum reflexion efficiency at 420 nm due to blazing of the
grooves (fig. 2). For further measurements the spectrometer
software allows us to correct the original spectrum with a flat
line spectrum simultaneously. One has to consider that the
numerical correction is most accurate in the range of the
maximum sensitivity and decrease with the loss in sensitivity.
The sensitivity range defines the detectable wavelength
range. This means that emission between 300 and ~850 nm
can be investigated.
With this setup emission spectra, as well as excitation spectra,
are obtainable. The emission spectra are recorded by keeping
the excitation wavelength constant, whereas the excitation
spectra are recorded by keeping the emission wavelength
constant.
For determination of the wavelength accuracy the emission,
excitation and absorption spectra of doped samples have
been investigated. In figure 3 these spectra are visualized us-
ing about 485 ppm TmF3 as dopand in CaF2 crystal. Doped
CaF2 is used because it exhibits a higher light yield compared
to optical glass and can therefore be used even for less sensi-
tive fluorescence spectrometers.

1.0
using BAM standards 35000000 0.92
using Starna standards
30000000 0.90
0.8
CaF2 + 485 ppm TmF3
25000000 0.88
Fluorescence [a.u.]

excitation for 345 nm

Efficiency

emission at 250 nm
0.6
20000000 excitation for 450 nm 0.86
emission at 350 nm

15000000 absorption (PE L900) 0.84

0.4

10000000 0.82
0.2
5000000 0.80

0.0 0 0.78
300 400 500 600 700 200 250 300 350 400 450 500 550

wavelength [nm] wavelength [nm]

Fig. 2: Efficiency for 1200 lines/nm grating and the PM detector. The maximum Fig. 3: A
 bsorption, excitation and emission spectra of TmF3 doped CaF2.
efficiency is observed for appr. 420 nm.

➜ ➜ Back to index

3
 Technical Information
Advanced Optics

TIE-36 Fluorescence of optical glass

Results have been compared using theoretical and analytical A relative accuracy or reproducibility better than two percent
work and data from literature. The absorption, excitation and for the FLUOROLOG 3 spectrometer is observed. The absolute
emission spectra for this sample have been obtained. It is ob- uncertainty – considering the uncertainty due to the calibra-
served that the peak of absorption and excitation spectra is tion of 5 % – is therefore given to 7 % over the full emission
exactly on the same position and therefore the consistency range between 300 and 850nm.
of absorption and excitation spectrum becomes visible. This
means that the accuracy is close to 0.1nm (resolution limit of Fluorolog first attemp
Fluorolog improved
the spectrometer) and well below the common step width of 1.0
B [NIST-Chinine sulphate]
1nm. The emission peaks at 350 and 450 nm are situated on C [upper limit of NIST-Chinine sulphate]
D [lower limit of NIST-Chinine sulphate]
the same position as expected from the literature and known 0.8

Fluorescence [a.u.]
from optical glass when excited with 250 and 350 nm respec-
tively.
0.6

The accuracy of the spectrometer has been evaluated using

certified standards BAM-F01 to BAM-F05 [8][9] In figure 4 0.4

the emission spectra of Quinine Sulphate certified by NIST is

shown to underline the correctness of the emission profile. 0.2

0.0
350 400 450 500 550 600 650 700
wavelength [nm]

Fig. 4: E
 mission spectrum of NIST certificates fluorophore Quinine Sulphate.

4. Evaluation of the integral fluorescence

Typical fluorescence measurement results on N FK58 XLD are wavelength. Analog to the publication from Reichel et. al on
shown in Figure 5 for 365 nm, 488 nm, and 532 nm excita- fluorescence measurements on optical filter [6], the fluores-
tion. N FK58 XLD is an eXtremely Low Dispersion glass that cence of optical glasses can be characterized by a single para-
is used in optical design for apochromatic color correction. meter that is called the Relative Integral Fluorescence (RIF).
For 365 nm excitation the RIF is defined by:
For each single excitation wavelength two fluorescence spectra
are shown: The fluorescence signal of the N FK58 XLD sample (1) 
∫ 700
370 sample (λ) dλ
is compared to the spectrum of SF1, which inhabits a high
fluorescence level and acts as a reference sample in all further
RIF365 :=

evaluation. The dashed lines represent always the excitation ∫ 700
370 SF1 (λ) dλ
wavelength (not to scale) and the dashed-dotted curves rep-
resent the spectrum of the SF1 reference at that excitation

➜ ➜ Back to index

4
 Technical Information
Advanced Optics

TIE-36 Fluorescence of optical glass

3500
Where Isample(λ) is the measured fluorescence spectrum (inten-
sity) of the sample (N FK58 XLD in Figure 5) and ISF1 (λ) is excitation 488 nm
3000
the reference fluorescence spectrum of SF1 as a function N-FK58 XLD
9 % RIF_488

of the wavelength. The subscript 365 in RIF365 refers to the 2500 SF1 reference

Fluorescence [a.u.]
excitation wavelength at 365 nm. According Eq. (1) the RIF365
of SF1 is 100 %. The RIF365 of N-FK58 XLD was measured to be 2000

RIF365 = 0.5 % of the integral fluorescence spectrum of optical

1500
glass SF1.

1000
Similar to Eq. (1), the relative integral fluorescence for
488 nm excitation (RIF488 – Eq. (2)) or 532 nm excitation 500
(RIF532 – Eq. (3)) is defined by:
0
450 500 550 600 650 700 750 800
(2) 
∫ 800 sample (λ) dλ
535
wavelength [nm]
RIF488 :=
600

∫ 800
535 SF1 (λ) dλ excitation 532 nm

500 N-FK58 XLD

23 % RIF_532

(3) 
∫ 850 sample (λ) dλ
SF1 reference
Fluorescence [a.u.]

580 400
 
RIF532 :=

∫ 850
580 SF1 (λ) dλ
300

200
The RIF at 405 nm and 642 nm excitation is defined accord-
ingly. Note the different integration ranges (wavelength 100
ranges) of the 3 definitions, Eq. (1), Eq. (2), and Eq. (3).
1000
40000
0
500 550 600 650 700 750 800 850
35000

30000 wavelength [nm]

800
25000

Fig. 5: Fluorescence spectrum of N-FK58 XLD with 4.97 mm sample thickness

Fluorescence [a.u.]

20000

15000
and SF1 (5.0 mm thickness) which serves as fluorescence standard. Top: meas-
600 10000
urement result for 365 nm excitation. Middle: measurement result for 488 nm
5000
excitation. Bottom: measurement result for 532 nm excitation. Note the differ-
0
ent scale of the y-axis.
350 400 450 500 550 600 650 700
400
excitation 365 nm For N-FK58 XLD the relative integral fluorescence for an exci-
N-FK58 XLD tation at 488 nm was measured to be RIF488 = 9 % whereas for
0.5 % RIF_365
200 an excitation at 532 nm it was measured to be RIF532 = 23 %
SF1 reference
Again RIF488 of SF1 is equal to 100 % as well as the RIF532 of SF1
0
is equal to 100 %.
350 400 450 500 550 600 650 700
wavelength [nm] ➜ ➜ Back to index

5
 .
Technical Information .
.
Advanced Optics .
. !"#$%!!!!!&'&&!
. !(#)%! 11/12

TIE-36 Fluorescence of optical glass

5. Fluorescence results for optical glass As has been shown above N-FK58 XLD and all the other fluo-
rophosphates glasses are among the glasses with the lowest
Figure 6 display emission spectra at an excitation wavelength
RIF values around 1 %.
of 365 nm measured for a variety of optical glasses with identi- .
. relative relative
integralintegral
fluorescence (RIF) at 365 nm excitation
cal measurement setup parameters (slit widths, integration . fluorescence (RIF) at 365 nm excitation
. 100%
time, detection regime). The integral fluorescence decreases
. 90%

fluorescence
!"#$%!!!!!&'&&!

integral fluorescence
from SF57HTUltra to N-KZFS8 according to the sequence of the . 80% !(#)%! 12/12

graphs legend. 70%

60%
Fluorescence at 365 nm excitation 50%

relative integral
80000 40%
SF57HTUltra 30%

relative
70000 N-KZFS11 20%
N-SF6HT
N-LAF21
10%
60000 N-LASF44 0%
Fluorescence [a.u.]

N-LASF31
LASF35

N-KZFS8
N-LASF9
N-LASF45
N-KZFS11
N-K5

P-SK60
K10

N-KZFS5
N-KZFS4
N-LASF44
N-LASF41

N-KZFS2
N-PK51
N-LASF31A

N-PK52A
N-FK58 XLD

N-FK51A
SF1
SF5
SF2
F2
LLF1
LF5

N-SF5
N-SF1
N-SSK8
N-BAF4
KZFSN5
N-LAK21
N-SK5
N-LAK22
N-LAK10
N-LAF21
N-LAF33

N-LAK9
N-BAK1
N-LAK14
N-SK2

N-PSK3
N-LAK8
N-FK5
N-BAK4
SCHOTT N-BK7®
N-BALF4
SF1 reference
SF6

N-SF6
N-LAK22
N-SF57HT
50000 N-FK5
N-FK51A
40000 N-KZFS8
N-KZFS5

30000 Fig. 7: Relative integral fluorescence of optical glasses at 365 nm excitation.

FigureThe RIF of the
7: Relative SF1 reference
integral is setoftooptical
fluorescence 100 %. glasses at 365 nm excitation. The RIF of the
20000 SF1 reference is set to 100 %.
relative integral
relative fluorescence
integral (RIF)(RIF)
fluorescence at 405 nm nm
at 405 excitation
excitation
10000
100%
90%
relative integral fluorescence

0
relative integral fluorescence

400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 80%
wavelength [nm] 70%
60%
Fig. 6: Emission spectra of low fluorescent optical glasses using 365 nm 50%
excitation wavelength and a cut off filter (KV 418)
40%
30%
20%
Data on the relative integral fluorescence (RIF) of optical glass
10% of optical glass
at 365 nm excitation is shown in Figure 7. The data shows TIE-36: Fluorescence
0%
typical values for the glass type.
N-LAK22
N-SSK8
N-LASF31A

N-FK51A
F5

N-LASF41
N-LAK8

N-BAK1
N-KZFS11

N-KZFS8
N-KZFS5
N-LAK9

N-SK5

N-PK51
N-BK7

N-K5

N-FK5
SF1 reference

N-BAK4

N-BAF4
N-KZFS4
N-LASF44

N-PK52A
The SF1 reference sample is included in the diagram with a
value of 100 %. The second SF1 data column reflects the pro-
Fig. 8:
Figure Relativeintegral
8: Relative integral fluorescence
fluorescence of optical
of optical glasses
glasses at 405 at
nm405 nm excitation.
excitation. The RIF of the
gress in melt technology over the years. The fluorescence of SF1
Thereference is set
RIF of the SF1toreference
100 %. is set to 100 %.
SF1 glass from recent production runs shows much lower
RIF values than the much older SF1 reference glass.

In general, the N-type glasses show lower RIF values than the TIE-36: Fluorescence of optical glass

non N-type glasses. The environmental friendly N glasses

have been designed to be free of lead, arsenic and antimony.
Especially lead shows strong fluorescence at 365 nm excitation.

➜ ➜ Back to index

6
 Technical Information
Advanced Optics

TIE-36 Fluorescence of optical glass

.
.
. Figure 8 shows the relative integral fluorescence at excitation It has to be pointed out there is no correlation between the
.
. at 405 nm wavelength. The SF1 reference again marks the
!"#$%!!!!!&'&&!
365 nm and the excitation larger than 400 nm. The reason is
. RIF 100 % limit. A smaller total number of glasses have been
!(#)%! 13/12 that the UV excitation/absorption depends mainly on impuri-
Figure 8 shows the relative integral fluorescence at excitation at 405 nm wavelength. The SF1ties whereas for larger excitation wavelengths respectively
investigated at this wavelength. The glass with the lowest
reference again marks
RIF value is N-FK5the RIF 100 %the
within limit. A smaller
range total number
of tested of glasses have beenlower photon energies, where merely absorption bands are
glasses.
investigated at this wavelength. The glass with the lowest RIF value is N-FK5 within the range
of tested glasses. observed, the glass composition is the driver for the fluores-
Figure 9 shows
Figure 9 the relative
shows theintegral fluorescence
relative integral at excitation at 488 nmatwavelength.
fluorescence excitation The SF1cence features.
reference again marks the RIF 100 % limit. The SF1 from current production runs shows much
loweratRFI488 nm
values wavelength.
(less than 60 %) thanThe SF1 reference
the older reference again
sample. marks
N-SSK8 the
shows a higher RIF
than the SF1 reference. The differences between the fluorescence of N-type glasses to non N-
type glasses is not so clear. The non N-type glasses are mixed in the RIF ranking. The glassFigure 10 shows the relative integral fluorescence at excitation
RIF 100 % limit. The SF1 from current production runs shows
with much
the lowest RIF value
lower is N-FK58
RFI values (lessandthan
the non60N-type
%) than glassthe
witholder
the lowest value is LLF1.at 532 nm wavelength. The SF1 reference marks the RIF 100 %.
reference
One can conclude that excitation at 488 nm is less effective than at 365 nm and the origin . of
sample.
fluorescence N-SSK8
is different showstoa365
compared higher RIF than the SF1 reference. .
nm excitation. This time many glasses show a higher RIF compared to the
It hasThe
to bedifferences
pointed out there between the fluorescence
is no correlation between the 365of nmN-type glasseslarger
and the excitation to. thanRIF 100 % of the SF1 reference. N BAK1 shows RIF value way
400 nm. The reason is that the UV excitation/absorption depends mainly on impurities whereas .
nonexcitation
for larger N-typewavelengths
glasses isrespectively
not so clear. lower The
photonnon N-type
energies, glasses
where are.
merely absorption above 600 %. N FK58 is still among the glasses with the !"#$%lowest
!!!!!&'&&!
bands are observed, the glass composition is the driver for the fluorescence features. .
mixed in the RIF ranking. The glass with the lowest RIF value is fluorescence. !(#)%! 14/12

FigureN-FK58
10 shows and the non
the relative N-type
integral glass with
fluorescence the lowest
at excitation at 532 nmvalue is LLF1.
wavelength. The SF1
reference marks the RIF 100 %. This time many glasses show a higher RIF compared to the
One can conclude that excitation at 488 nm is less effective relative integral fluorescence
relative integral (RIF)(RIF)
fluorescence at 532 nmnm
at 532 excitation
excitation
RIF 100 % of the SF1 reference. N-BAK1 shows RIF value way above 600 %. N-FK58 is still
700%
among than at 365with
the glasses nmtheand the
lowest origin of fluorescence is different com-
fluorescence.
fluorescence
integral fluorescence

600%
pared to 365 nm excitation.
500%
relativerelative
integral fluorescence
integral (RIF)(RIF)
fluorescence at 488 nmnm
at 488 excitation
excitation 400%
relative integral

120% 300%
relative integral fluorescence
relative integral fluorescence

100% 200%
relative

80% 100%
0%
60%
N-LASF31A

N-PK52A
N-BAK1
K10
LLF1
N-SK2
N-LAK22
N-PSK3
N-K5
P-SK60
N-SK5
LF5
SCHOTT N-BK7®
N-LAK14
N-LAK8
N-LAK21
N-LAK9
N-LASF44
N-KZFS2
N-LASF31
N-LAF33
N-KZFS11
N-LAK10
N-SSK8
N-KZFS4
LASF35
N-KZFS8
F2
N-LASF41
N-BAK4
N-LAF21
N-BALF4
N-FK5
N-LASF45
KZFSN5
N-LASF9
SF1
SF2
N-PK51
N-KZFS5
SF5
N-BAF4
N-SF1

N-SF5
SF1 reference

N-FK51A

N-FK58 XLD
SF6

N-SF6
40%

20%

0%
FigureFig.
10: Relative integral
10: Relative fluorescence
integral of optical
fluorescence glasses
of optical at 532atnm
glasses excitation.
532 The RIF of
nm excitation.
SF2

N-KZFS5
N-KZFS2

LF5
N-FK5
KZFSN5
LLF1
N-PK51

N-BALF4
N-LASF31
P-SK60

N-KZFS8
SF1
N-K5

LASF35
N-KZFS11
K10
N-LAF33

SF5
SCHOTT N-BK7®
N-LASF9
N-LASF45
N-KZFS4
N-LASF44
N-LAF21

N-LASF41

N-PK52A

N-FK51A
N-FK58 XLD
N-LASF31A
SF6

N-LAK8
N-LAK14
N-SF1

F2
N-SF5

N-BAF4
N-BAK4
N-SSK8

N-BAK1
N-SK5

N-SK2
N-PSK3
N-LAK22

N-LAK21
N-LAK9
N-LAK10
SF1 reference

N-SF6

the SF1 reference

The is SF1
RIF of the set to 100 %. is set to 100 %.
reference

Figure 9: Relative integral fluorescence of optical glasses at 488 nm excitation. The RIF of the
SF1 Fig. 9: Relative
reference integral
is set to 100 %.fluorescence of optical glasses at 488 nm excitation.
The RIF of the SF1 reference is set to 100 %.

TIE-36: Fluorescence of optical glass

➜ ➜ Back to index

TIE-36: Fluorescence of optical glass

7
 Technical Information
Advanced Optics

TIE-36 Fluorescence of optical glass

Fluorescence at 532 nm excitation Figure 12 shows the relative integral fluorescence at excita-
4000
tion at 642 nm wavelength. The SF1 reference marks the RIF
N-LAK22
3500 N-KZFS11 100 %. Again many glasses show a higher RIF compared to
N-SF57HT
N-LASF44 the RIF 100 % of the SF1 reference. N-SF1 and N-PK52A show
3000 N-LAF21
a factor of 100 higher fluorescence compared to SF1. SF1 is
Fluorescence [a.u.]

N-KZFS8
N-SF6HT
2500
SF2 among the glasses with the lowest fluorescence at 642 nm
N-KZFS5
2000 N-FK5 excitation.
N-FK51A
SF57HTUltra
1500
The diagrams Figure 7, Figure 8, Figure 9, Figure 10 and Fig-
1000 ure 12 are irritating in a way that they only show the relative
500
tendency to produce fluorescence light between the different
glass types at the given excitation wavelength.
0
555 575 595 615 635 655 675 695 715 735 755
wavelength [nm] Figure 13 shows the SF1 reference sample spectra for 365 nm,
405 nm, 488 nm, 532 nm and 642 nm excitation. With in-
Fig. 11: Emission spectra of optical glasses using 532 nm excitation wavelength
and without cut off filter. creasing excitation wavelength the maximum and area of the
spectrum decreases. Setting the area below the SF1 reference
In Figure 11 typical emission spectra using 532 nm excitation spectrum at 365 nm to 100% the RIF’s for SF1 reference from
wavelength are shown for several optical glasses. Using the 405 nm to 642 nm can be calculated as shown in the legend
. identical experimental setup the shape of the emission profiles of the diagram. The RFI of SF1 at 642 nm excitation is only
.
. are similar, but the measured intensities in are different and 0.04 % of that at 365 nm excitation.
. overall much lower. The RIF decreases from N-LAK22 to
. !"#$%!!!!!&'&&!
. SF57HTUltra (according to the sequence of the graphs legend)
!(#)%! 16/12
SF 1 reference spectra
35000

SF1 ref. 365 nm excitation,
relative integral
relative fluorescence
integral (RIF) (RIF)
fluorescence at 642at nm
642excitation
nm excitation integral 100 %
30000 
1600,0% SF1 ref. 405 nm excitation,
integral 15 %

SF1 ref. 488 nm excitation,
25000
Fluorescence [a.u.]

1400,0% integral 8,1 %


SF1 ref. 532 nm excitation,
integral 0.8 %
1200,0% 20000 
SF1 ref. 642 nm excitation,
integral fluorescence
fluorescence

integral 0.04 %

1000,0% 15000

800,0%
relativeintegral

10000

600,0%
5000
relative

400,0%
0
200,0%
300 400 500 600 700 800
wavelength [nm]
0,0%
Fig. 13: SF1 reference sample spectrum for 365 nm, 405 nm, 488 nm, 532 nm
KZFSN5
N-LAK14

N-LAK10
N-KZFS8
N-K5

N-LAK22
N-KZFS2
N-KZFS5
N-KZFS11
N-SK5
N-BAK1
N-KZFS4
LF5

SF2
SF1
SF1 reference
N-PK52A

N-LASF31A
N-PSK53A
N-FK51A

N-BK7

LLF1

N-FK5
N-SF1

N-SF6
N-LASF45
N-PK51

N-LAF21

F2
N-SSK8

and 642 nm excitation.

Fig. 12: Relative integral fluorescence of optical glasses at 642 nm excitation.

Figure
The 12:
RIFRelative integral
of the SF1 fluorescence
reference
the SF1 reference is set to 100 %.
is set of
to optical
100 %.glasses at 642 nm excitation. The RIF of ➜ ➜ Back to index

Figure 12 shows the relative integral fluorescence at excitation at 642 nm wavelength. The SF1
reference marks the RIF 100 %. Again many glasses show a higher RIF compared to the RIF
100 % of the SF1 reference. N-SF1 and N-PK52A show a factor of 100 higher fluorescence
compared to SF1. SF1 is among the glasses with the lowest fluorescence at 642 nm excitation.
The diagrams Figure 7, Figure 8, Figure 9, Figure 10 and Figure 12 are irritating in a way that
they only show the relative tendency to produce fluorescence light between the different glass
types at the given excitation wavelength.
Figure 13 shows the SF1 reference sample spectra for 365 nm, 405 nm, 488 nm, 532 nm and
642 nm excitation. With increasing excitation wavelength the maximum and area of the
spectrum decreases. Setting the area below the SF1 reference spectrum at 365 nm to 100%
the RIF’s for SF1 reference from 405 nm to 642 nm can be calculated as shown in the legend 8
of the diagram. The RFI of SF1 at 642 nm excitation is only 0.04% of that at 365 nm excitation.
 Technical Information
Advanced Optics

TIE-36 Fluorescence of optical glass

The relative fluorescence of SF1 between the different excita- The diagrams do not give any information regarding the
tion wavelengths can be used to compare RIF’s of a glass at quantum efficiency of the process at different wavelength. In
different wavelengths. The results shown in Figure 7, Figure 8, general the quantum efficiency for fluorescence decreases with
Figure 9, Figure 10 and Figure 12 can be scaled (multiplied) increasing wavelength and optical glasses in general show
with 100 % (365 nm), ~15% (405 nm), ~8% (488 nm), very low quantum efficiency values. This will be shown in the
~0.8 % (532 nm) and ~0.04 % (642 nm) to take into account next chapters.
the much lower absolute fluorescence in arbitrary units [a.u.],
this is what we call the “weighted RIF”. Figure 14 shows the
weighted relative integral fluorescence RIF for several glass
types in comparison.

weighted relative integral fluorescence (based on SF1)

12%
relative integral fluorescence weighted

10%

8%

6%

4%

2%

0%
A

K8

A
8

4
K1

51
2
K5

A
FS

FS

FS

BK
SK

K2

FK
31

52

51
BA
SS

PK
N-

KZ

KZ

KZ
N-

N-
LA

N-
SF

PK

FK
N-

N-
N-
N-

N-

N-
LA

N-

N-
N-
N-

365 nm 405 nm 488 nm 532 nm 642 nm

Fig. 14: Weighted relative integral fluorescence based on SF1 to compare the
glasses at different excitation wavelengths.

➜ ➜ Back to index

9
 Technical Information
Advanced Optics

TIE-36 Fluorescence of optical glass

6. F luorescence quantum efficiency The C9920 uses a Xenon light source and a monochromator
measurement set-up [1] to precisely adjust the excitation wavelength with a resolution
better than 5 nm. The excitation light is guided into an inte-
The measurement of RIF has the advantage of relatively high
grating sphere. The sample is mounted inside the integrating
precision, but it is a relative measurement. Thus the absolute
sphere under a small angle well below 10° to suppress front-
value of “how many of the incident photons are emitted as
and backside reflection of the probe. The multi-channel de-
fluorescence photons” is not given by this parameter. In order
tector is mounted at a 90° angle relative to both the excitation
to get an estimation of the absolute fluorescence level, the
light and the probe and measures the complete fluorescent
(external) fluorescence quantum efficiency (FQE), is measured
light avoiding any impact of the fluorescence emission angle
for N FK58 XLD and given by:
characteristics. A data analyzer on a computer calculates the FQE.

(4)
∫ fluor (λ) dλ The C9920 can estimate the self-absorption using with the
FQE = , same device. The self-absorption corrected effective FQE (Ab-
 sorption * FQE) is the best value to describe the absolute
∫ in (λ) dλ amount of fluorescence light generated from a given illumina-
tion intensity. In a comparative experiment together with BAM
where fluor is the fluorescent light spectrum (emitted by the
(Federal Institute for Materials Research and Testing, Berlin,
probe) and in is the incident light spectrum (integrated over
Germany) using a sample with perspective to act as a stand-
the whole wavelength spectrum).
ard, we achieved an accuracy of ± 2 % and a re producibility
of better than ± 0.5 % for the un-corrected FQE. The required
FQE is measured using the instrument C9920 from the com-
sample size is 5 – 18 mm in diameter and a thickness between
pany Hamamatsu. The basic set-up is shown in Figure 15.
0.5 mm and 5 mm.

Integrating sphere

Xenon light
source Monochromator

Multi-channel
detector and
Probe (sample) data analyzer

Fig. 15: Basic set-up for measuring the (external) absolute fluorescence
quantum efficiency. By using a calibrated Xenon light source and a monochro-
mator the excitation wavelength can be chosen. An integrating sphere meas-
ures the whole fluorescent light by a multichannel detector. All components
have been calibrated by Hamamatsu.

➜ ➜ Back to index

10
 Technical Information
Advanced Optics

TIE-36 Fluorescence of optical glass

7. F luorescence quantum efficiency
measurement results [1]
The FQE was measured for a variety of optical glasses including
the SF1 reference at an excitation of 365 nm and 488 nm.
The results can be found in Table 1.

365 nm excitation 488 nm excitation

FQE Absorb. eff. FQE FQE Absorb. eff. FQE
[%] [%] [%] [%] [%] [%]
SF1 3.5 41.7 1.46 4.7 1.4 0.07
SF6 0.7 76.0 0.53 n.m. 16.4 n.m.
N-KZFS8 1.1 42.6 0.47 0.15 8.5 0.01
SF57 0.5 89.3 0.45 0.8 12.7 0.10
N-SF6 0.4 93.0 0.37 n.m. 19.9 n.m.
N-LASF45 0.5 76.0 0.34 0.2 22.7 0
N-LAK9 0.8 36.0 0.29 n.m. 17.3 n.m.
N-BAK4 0.9 28.0 0.25 n.m. 10.2 n.m.
SCHOTT N-BK7® 0.4 9.5 0.04 0.1 4 0.00
N-FK58 XLD 1.3 3.1 0.04 5 0.8 0.04
OG590 [5] 1.7 71 1.2 1.8 83 0.05

Tab. 1: Effective FQE of several optical glasses and OG590 filter glass in
comparison. “n.m.” means “not measurable”

Among the optical glasses SF1 shows the highest FQE of
3.5 %. The effective FQE is about 1.5 % with an absorption
measured of 41.7 % with the C9920. The lowest effective
FQE at 365 nm excitation was found in N-FK58 and SCHOTT
N-BK7® (in both cases the effective FQE is less than 0.05 %).
The effective FQE of SF57 and N-KZFS8 is a factor of 10 higher.
The values at 488 nm excitation are in general smaller for all
optical glasses due to the lower absorption, except for N-FK58
that exhibits an effective FQE of 0.04 % at 488 nm excitation,
identical to the effective FQE at excitation at 365 nm. For
many optical glasses the fluorescence signal was too low in
the C9920 at 488 nm excitation. These results are indicated
with “n.m: not measurable” at 488 nm excitation. To judge
the reliability of the measurements it is important to look at
the measured spectral intensity distribution.
Figure 16 shows the spectral response of the multi-channel
detector during measurement for excitation at 365 nm of SF1,
N KZFS8, SCHOTT N BK7® and N FK58 XLD. The excitation
peak at 365 nm is clearly visible. The intensity at the excita-
tion peak in comparison to the spectrum in air reflects the ab-
sorption of the glass. At longer wavelengths most of the
measurements are on the same level as the spectrum in air.
No significant fluorescence can be detected. The only excep-
tion is SF1, which is significantly above the other signals. A
similar behaviour can be observed in Figure 17 for excitation
at 488 nm. Even the fluorescence of SF1 is not dominant any-
more. The signal is on noise / air level.
➜ ➜ Back to index

11
 Technical Information
Advanced Optics
TIE-36 Fluorescence of optical glass
Therefore, the absolute results in Table 1 should be inter-
preted with care. At least it is obvious that the internal fluo-
rescence level of most of the optical glasses is below 1 % at
365 nm and decreases almost by an order of magnitude at
488 nm excitation. The C9920 is not sensitive enough to
100000
100000
1000
intensity [a.u.]
intensity [a.u.]
100
10
1 1
0.1
300 350 400 450 500 550 600 650 700 750 800 850 900 950
wavelength [nm]
Fig. 16: FQE measurement spectrum of different optical glasses at excitation
of 365 nm.
8. Literature
[1] R. Jedamzik F. Elsmann, A. Engel. U. Petzold, J. Pleitz,
Introducing the quantum efficiency of fluorescence of
SCHOTT optical glass, SPIE Proceeding 10375, 2017
[2] W. Geffcken, The molar absorption of different ions
in glases, Glastechn. Berichte, 35, p. 27 f, 1962
[3] D. Ehrt, P. Ebeling, U. Natura, U. Kohlberg, K. Nau-
mann, and S. Ritter, Redox equilibria and ultraviolett
radiation induced defects in glasses, Intern. congr.
on Glass Vol. 1, p. 84, 2001
[4] J.S: Stroud, Optical absorption and color caused by
selected cations in high-density, lead silicate glass,
J. Am. Cer. Soc. 54, p. 8 ff, 1971
[5] W. Basiev, P-G. Zverev, V.V. Fedorov, and S.B. Mirov,
Appl. Opt. 36, p. 2512 ff, 1997
For more information please contact:
schott.com
SCHOTT AG, Hattenbergstrasse 10, 55122 Mainz, Germany
Phone +49 (0)6131/66-1812, info.optics@schott.com
generate reliable fluorescence quantum efficiency values for
optical glasses. The sensitivity must be an order of magnitude
higher at least. The C9920 is optimized for substances with
fluorescence levels significantly above 1 %.
100000
100000
1000
100
10
0.1
300 350 400 450 500 550 600 650 700 750 800 850 900 950
wavelength [nm]
Fig. 17: FQE measurement of optical glasses at excitation of 488 nm.
[6] R
 eichel, S., Biertuempfel, R., Engel, A., “Characterization
and measurement results of fluorescence in absorption
optical filter glass.” Proc. SPIE, 9626 (2015)
[7] P. deRose, “Bioanalytical and biomedical applications
of fluorescence techniques: instrum. characterisation
and validation, traceability and need for Reference
Materials”, Fluorescence Workshop@ BAM, Berlin 2003
[8] BAM-F01 to BAM-F05, “Calibration Kit SPECTRAL
FLUORESCENCE STANDARDS,” Federal Institute for
Materials Research and Testing (BAM) (2006)
[9] BAM-F012, “Glass-based Multi-emitter Fluorescence
Standard,” Federal Institute for Materials Research
Version May 2022
and Testing (BAM) (2014)
➜ Back to index