Absorption and fluorescence spectra of poly(ethylene …

Loading...

Taking too long?

Reload document

| Open in new tab

Downloaded from http://polymerphysics.net

Absorption and fluorescence spectra of
poly(ethylene terephthalate) dimers

M. F. Sonnenschein and C. M. Roland
Chemistry Division, Code 6120, Naval Research Laboratory, Washington,
DC 20375-5000, USA
(Received 12 July 1989; revised 2 November 1989; accepted 20 November 1989)

The concentration and temperature dependence of the absorption and fluorescence spectra of poly(ethylene
terephthalate) (PET) solutions were measured. At concentrations above that associated with the onset of
chain overlap, ground-state aggregation was observed. This aggregation phenomenon is similarly seen in
the solid state, as well as in dimethyl terephthalate solutions. Elevated temperature effects dissociation of
PET dimers, although their fluorescence intensity increases, presumably due to enhanced energy transfer.

(Keywords: poly(ethylene terephthalate); dimers; fluorescence)

to as

INTRODUCTION
Recently it was reported that concentrated solutions and
terephthalate) (PET) exhibited
films of poly(ethylene
fluorescence emission from two ground-state entities1 -.
It was hypothesized that the two fluorescing structures
were the aromatic unit of the phthalate residue (herein
the ‘monomer’) and a ground-state
referred
aggregation of the same residue (herein referred to as the
‘dimer’). While Soxhlet extraction to remove small-
molecule impurities from the PET did not alter the
fluorescence spectra3, the possibility that the emission
results from impurities bound to the polymer chain
remained. Another series of experiments4 6 under a
different set of conditions, primarily at lower concen-
trations, have also been described.

In order to probe the origins of the PET fluorescence
in more detail, a series of solution-phase experiments
were conducted
to test the ground-state aggregation
hypothesis. The results of fluorescence and absorption
spectroscopic measurements are reported herein.

EXPERIMENTAL
Amorphous PET (additive-free pellets; intrinsic viscosity
[,n] = 0.94 dl g -1) was kindly provided by Mr Craig Trask
of Allied Signal Inc. (Petersburg, VA) and used without
purification or further characterization. Dimethyl tereph-
thalate (DMT) of better than 99% purity from the
Aldrich Chemical Co. was recrystallized from boiling
acetone prior to use. Solutions of DMT were filtered
through Fisher Q2 filter paper prior to use. An absence
of microcrystals in the filtered solutions was confirmed
by inspection with a Zeiss D-7082 transmitting light
microscope using crossed polarizers. Better than 99%
pure 1,1,1,3,3,3-hexafluoro-2-propanol
from
Aldrich was used without purification. Ultra-violet/
visible absorption spectra, background-corrected
at
each temperature, were obtained using a Perkin-Elmer
3840 Lambda Array spectrophotometer (with purported
linearity to OD = 3) mated to a Perkin-Elmer 7500
microcomputer. Fluorescence spectra were taken on a
Spex Fluorolog 212 spectrofluorimeter with a 450 W

(HFIP)

high-pressure xenon lamp. The solvent HFIP exhibited
no absorption above 225 nm, and no fluorescence emis-
sion over any of the excitation wavelengths utilized in
this study. All fluorescence measurements were made
observing front-face emission with slit widths constant
at 0.35 mm. After 30 min lamp warm-up, lamp jitter was
less than 5% and low-frequency intensity fluctuations
were negligible.

Temperature-dependent

absorption measurements
employed a Forma Scientific 2095 bath, circulating
temperature-controlled water through the fluorescence
cell holder. Cell temperature was determined by monitor-
ing a thermometer placed in a cell containing water within
the cell holder. At the desired temperature, a closed cell
containing PET/HFIP solution was placed in the cell
holder and allowed to equilibrate for 20 min before
measurements were made. Longer equilibration times
gave quantitatively equivalent results. All measurements
were made in order of ascending temperature. Samples
were not routinely deaerated, since even the most
concentrated solutions at the lowest temperatures showed
negligible oxygen quenching, as a consequence of the
lifetimes (-~ 10- 9 s) of both
very short fluorescence
monomer and dimer entities3. All concentrations are
reported herein based on the concentration of the PET
repeat unit.

RESULTS AND DISCUSSION
Absorption spectra and fluorescence excitation and
emission spectra of PET films have been reported
previously8′ 9 . We have essentially reproduced this pre-
vious work. The spectra of PET dissolved in HFIP are
qualitatively similar to that of PET films, but several
crucial differences exist. In Figure I
is shown the
absorption spectrum of PET in HFIP under relatively
dilute conditions. Using the assignments of ref. 9, the
A1 -A’ state (with maximum absorption at wavelength
Amax =290nm) and the A-B’2 state (with ‘max at 244 nm)
exhibit red shifts of 10 and 14 nm, respectively, compared
with the film. The extinction coefficients are – 104 and
– 105 cm 1 mol-P1 1, respectively, in HFIP, in contrast to

0032-3861/90/112023- 04
C 1990 Butterworth-Heinemann Ltd.

POLYMER, 1990, Vol 31, November 2023

Absorption and fluorescence spectra of PET dimers: M. F. Sonnenschein and C. M. Roland

the nearly equal extinction coefficients measured for
films9.

Previous workers have reported that observation of
emission at 338 nm yielded a fluorescence excitation
spectrum with A.aX at 321 nm (assigned to the monomer),
while observation of emission at 390 nm obtained a
fluorescence excitation spectrum with Amax at 340 nm
(assigned to the dimer) with measurable excitation out
to 380nm” 3. Figure 2 shows the absorption spectrum
between 330 and 400nm from a concentrated solution

of PET in HFIP. The fluorescence excitation associated
with emission at 390 nm is associated with the absorbing
tail of the Al-A’ absorption. The resulting solution-
phase fluorescence excitation and emission spectra associ-
ated with the absorbing electronic transitions and the
tail absorption are presented in Figure 3.

To determine if the origin of the dimer excitation might
actually be due to hot band absorption, the temperature
dependence of the long-wavelength absorption was
measured. An increase with solution temperature is, of

200

250
WAVELENGTH (nm)

300

350

400

350
370
360
WAVELENGTH (nm)

380

390

400

Figure 1 Room-temperature absorption spectrum of 5 x 10-I M (by
monomer) PET in HFIP. The absorption at wavelengths less than
220 nm is due to HFIP

Figure 2 Absorption spectrum of 0.4 M PET in HFIP at 320C. At
wavelengths less than 320 nm the optical density is greater than 3;
absorbance at 330nm is 0.45

w
0z

0
C,,

WAVELENGTH (nm)

3D

w
z

0

z
II-z

280

330

370

360

WAVELENGTH (nm)

550

Figure 3 Excitation and emission spectra of 0.3M PET in HFIP: (a)monomer excitation, Aem=340nm; (b) monomer emission, Al,=300nm; (c)
dimer excitation,

,.. = 390 nm; (d) dimer emission,

. = 340 nm

2024 POLYMER, 1990, Vol 31, November

Absorption and fluorescence spectra of PET dimers: M. F. Sonnenschein and C. M. Roland

I”‘2

0.04

Z
oCD
(0

C-
8,-0.04

f

z

-0.08 –

-0.2

290

310

330

350

370

390

410

430

WAVELENGTH (nm)

Figure 4 Differential absorption of 0.4 M PET in HFIP. Spectra are
the difference in optical density at 24 and 320C (El) and at 24 and 45’C
(+). Absorption below 320 om was beyond the linear range of the
instrument

course, expected if the intensity represents a hot band
absorption, while if the absorption is from ground-state
dimers, their thermal dissociation at elevated temperature
will result in an intensity reduction’0 . On the other hand,
the absorption of impurities would be essentially un-
affected by temperature*. Figure 4 shows the difference
in absorption between that at 24 and 320C, and 24 and
450C. It is seen that, as the temperature is increased, the
absorption of the tail decreases. A similar experiment,
measuring the temperature dependence of the dimer
fluorescence intensity, gave the opposite result; that is,
dimer fluorescence increases with increasing temperature,
consistent with previous results 6 “‘l. Since the reduction
in dimer absorption with increasing temperature indicates
thermal dissociation of the dimers, the increasing dimer
fluorescence with temperature evidently results from
thermally enhanced energy transfer3 “l2.

An experiment was performed to determine if the
fluorescence intensity of the dimer emission was related
in a predictable way to the total concentration of
polymer. Because of energy transfer from the monomer
excited state to the dimer excited state, it is not possible
to derive relative concentrations from the fluorescence
intensities; nevertheless, these measurements can provide
insight into the nature of the aggregation process.
Spectral intensity resulting from intrachain dimer for-
mation would be consistent with a zero intercept for the
emission intensity versus concentration, while interchain
dimers would give rise to non-zero intercept. Figure 5
shows that the dimer fluorescence intensity does, in fact,
extrapolate to a non-zero (and positive) concentration
intercept. These data suggest that the origin of the dimer
fluorescence is primarily aggregation between different
polymer chains. Also, it is noted that the monomer
fluorescence intensity has a strong negative correlation
with this intensity, as expected if the latter is due to dimers
rather than impurities.

Even in the melt, a polymer chain occupies a small
portion of the volume it pervades’ 3. In a good solvent,
the segment density for a single chain is very low; hence,

interchain dimer formation is expected to dominate. The
PET, with a weight-average molecular weight equal to
75000 (based on its intrinsic viscosity”4), will have a
radius of gyration (RG) equal to 23 nm under theta
conditions1 5. This corresponds
to a chain overlap
concentration of roughly 0.01 M in monomert. Figure 5
shows that appreciable dimer fluorescence intensities do,
in fact, require concentrations beyond 0.01 M. Because
of the strong concentration dependence of R., the
absolute transition from the dilute to semidilute concen-
tration regime cannot be calculated; nevertheless, the
congruence between the calculated and measured concen-
trations for chain overlap is instructive. Further results
suggestive of a marked growth of PET dimer fluorescence
at the dilute to semidilute concentration transition will
be reported later’6 .

The dimer formation seen in PET should also occur
in chemically similar small molecules. In fact, dimer
fluorescence has been reported in fluorescence spectra
taken of crystals of dimethyl terephthalate3 . Although,
previously, no evidence of aggregation was obtained from
solutions of DMT in various solvents 3, more concen-
trated solutions of DMT in either HFIP or chloroform
do exhibit dimer fluorescence (Figure 6), as well as energy
transfer from the monomeric lumophore to the ground-
state aggregate. DMT can
thus serve as a model
compound for the crystal and solution photophysics of
PET.

SUMMARY
Experimental observations have been presented that are
consistent with the hypothesis that the long-wavelength
excitation/emission spectra of PET are associated pri-
marily with interchain ground-state aggregation. Under
the conditions reported in this paper, the ground-state
absorption and its inverse dependence on temperature,
when coupled with the reported absence of a ‘growing-in’
of dimer fluorescence3’1 7, provide strong evidence for the
presence of PET ground-state aggregates in solution. The
increase in the long-wavelength fluorescence intensity
with increasing temperature, notwithstanding thermal

I

I

I

I

I

I

IF

I

t

1

;

I

—

S-

S

,,

,’

,’

I

08.

wU 0.6
LLUO

> 0.4

LU
cc 0.2

0

0.02

0.05
CONCENTRATION (Moles Monomer/Liter)

0.04

0.1

0.2

0.3

Figure 5 Concentration dependence of monomer (
) and dimer
(—- ) integrated emission intensities. Values in each series are
normalized to the solution of highest fluorescence intensity. Monomer
spectra are corrected for the slight change in fluorescence excitation
spectrum as a function of concentration in this range. See footnote to
text for details

* The volume expansion of HFIP over the temperature range is
expected to be less than 3%, assuming corresponding states with water
between the melting and boiling points. The actual changes in the
absorption are too large to be accounted for in this way

t c* is calculated using RG as the radius of a sphere that the polymer
in solution excludes. There is no a priori reason to choose any particular
model for the calculation of c* in these experiments, but the method
chosen here reflects in part their chain dynamic aspect

POLYMER, 1990, Vol 31, November 2025

Absorption and fluorescence spectra of PET dimers: M. F Sonnenschein and C. M. Roland

non-zero intercept of the fluorescence intensity when
plotted as a function of polymer concentration suggests
primarily interchain aggregation. The implications of this
ground-state aggregation on the Theological behaviour
of PET in solution and in the melt remain to be
investigated.

ACKNOWLEDGEMENTS

thanks

M.S.
post-doctoral fellowship.

the National Research Council for a

1
2

3

4

5

Allen, N. S. and McKeller, J. F. Makromol. Chem. 1978,179,523
Hennecke, M. and Fuhrmann, J. Makromol. Chem., Macromol.
Symp. 1986, 5, 181
Hemker, D. J., Frank, C. W. and Thomas, J. W. Polymer 1988,
29, 437
Takai, Y., Misutani, T. and leda, M. Japan. J. Appl. Phys. 1978,
17, 651
Renyuan, Q. in ‘Macromolecules
Rempp), Pergamon Press, Oxford, 1982

(Eds. H. Benoit and P.

6 Mendicuti, F., Patel, B., Vellarkad, V. N. and Mattice, W. L.

10

7
8
9

Polymer 1989, 29, 1669
Hirayama, F. J. Chem. Phys. 1965, 42, 3163
Phillips, D. J. and Schug, J. C. J. Chem. Phys. 1969, 50, 3297
Padhye, M. R. and Tamhane, P. S. Angew. Makromol. Chem.
1978, 69, 33
Irie, M., Kamijo, T., Alkawa, M., Takemura, T., Hayashi, K.
and Baba, H. J. Phys. Chem. 1977, 81, 1571
11
Hennecke, M., Kurz, K. and Fuhrmann, J. Polymer 1989,30, ???
12 MacCallum, J. R. in ‘Photophysics of Polymers’ (Eds. C. E.
Hoyle and J. M. Torkelson), American Chemical Society,
Washington DC, 1987
deGennes, P. G. ‘Scaling Concepts in Polymer Physics’, Cornell
University Press, Ithaca, NY, 1979

13

16

14 Weisskopf, K. J. Polym. Sci., Polym. Chem. Edn. 1988,26,1919
15

Naoki, M., Park, I.-H., Wunder, S. L. and Chu, B. J. Polym.
Sci., Polym. Phys. Edn. 1985, 23, 2567
Sonnenschein, M. F. and Roland, C. M. J. Polym. Sci., Polym.
Phys. Edn. submitted for publication
Birks, J. B. ‘Photophysics of Aromatic Molecules’, Wiley, New
York, 1970

17

240

z

Z

285
WAVELENGTH (nrro

330

(b)

REFERENCES

300

340
WAVELENGTH (nm)

380

Figure 6 Excitation spectra of DMT-saturated HFIP: (a) monomer
spectrum, A2.. =340nm; (b) dimer, Am=4 50nm. Energy transfer from
the monomer to the dimer is evidenced by the presence of the monomer
spectrum (<330nm) in the dimer excitation. Emission spectra are substantially similar to those in Figure 3, but reflect a greater overlap of the absorption spectra than in PET dissociation of the dimers, probably reflects an enhance- ment in the energy transfer from monomers to dimers. The growth of monomer fluorescence intensity is negatively correlated with that of the dimer, while the positive, 2026 POLYMER, 1990, Vol 31, November