496
J. Phys. Chem. 1992, 96, 496-502
Dopant Site Sizes and Dopant Rates of Diffusion in Low-Density Polyethylene Films with Covalently Attached Pyrenyl Groups. Fluorescence Quenching by a Homologous Series of N,N-Dialkylanilines in Unstretched and Stretched Films Roseann M. Jenkins, George S. Hammond, and Richard G. Weiss* Department of Chemistry, Georgetown University, Washington, D.C. 20057 (Received: July 15, 1991)
Low-density polyethylene films with covalently attached pyrenyl groups at internal dopant sites (Py-LDPE) were prepared and their fluorescence properties were explored as four homologous N,iVdialkylanilines (DAA, alkyl = methyl, ethyl, propyl, butyl) were allowed to diffuse from methanolic solutions into the films or out of the films into neat methanol. From the data, the fraction of pyrenyl-occupied dopant sites capable of including at least one DAA molecule and the rates at which that inclusion occurs have been calculated. In the temperature range examined, the rate constants decrease as the molecular volume of a DAA increases. However, the larger DAA exhibit the smaller assoCiated activation energies for pyrenyl-fluorescence quenching in the films. Closer scrutiny of the data reveals that the activation energies for the larger DAA are based upon a smaller fraction of pyrenyl-occupied dopant sites than for the smaller DAA. In essence, entry into dopant sites with insufficient free volume is uery difficult. The fraction of dopant sites with adequate free volume is larger in unstretched than in stretched Py-LDPE. The distribution of dopant site volumes in the stretched films is also markedly narrower. The decrease in the quenched fraction of sites in the stretched films cannot be explained on the basis of bulk DAA concentrations in LDPE or their relative quenching efficiencies of pyrenyl fluorescence in other media. The results are discussed in terms of the various parameters expected to be important for diffusion within the films.
Introduction The free volumes available in dopant sites of low-density polyethylene (LPDE) films probably exist as a distribution which depends upon the molecular weight, branching, and thermal history of the specific polymer.' Although LDPE consists of about equal amounts of crystalline domains (where dopant molecules apparently do not enter2) and amorphous regions? the other factors mentioned above make LDPE a generic definition which fits many different specific materials.' In each, the dopant molecules reside preferentially in the amorphous region and at the interfaces between crystalline and amorphous regions: In some cases, dopants may relocate to the latter sites when the polymer is stretched.*s4 In spite of the lack of a specific LDPE, many studies with different polymer formulations have produced consistent spectroscopic dataS on molecules doped into stretched LDPE films6 This implies that (macroscopic) cold stretching of films may force the dopant molecules into specific, oriented microscopic sites.6b We have probed the differences between dopant sites in unstretched and stretched LDPE films using several technique^.^-^ All lead to the same conclusions: (1) the average free volume of the dopant sites decreases when a film is stretched; (2) the (1) (a) Axelson, D. E.; Levy, G.C.; Mandelkem, L. Macromolecules 1979, 12, 41. (b) Glenz, W.; Peterlin, A. J . Macromol. Sci. Phys. 1970, 8 4 , 473. (c) Hadley, D. W. In Structures and Properties of Oriented Polymers; Ward, 1. M., Ed.; Wiley: London, 1975; Chapter 9. (d) Nordmeier, E.; Lanver, U.; Lechner, M. D. Macromolecules 1990, 23, 1072, 1077. (2) Phillips, P. J. Chem. Rev. 1990, 90, 425. (3) (a) Abbate, S.;Zerbi, G.; Wunder, S. L. J . Phys. Chem. 1982, 86, 3140. (b) Chang, S.S.;Bestul, A. B. J. Res. Nafl.Bur. Sfand.,Sect. A, 1973, 77. 395. (c) Peterlin, A. Macromolecules 1980. 13. 777. (4) Jang; Y. T.; Phillips, P. J.; Thulstrup, E.W. Chem. Phys. Lett. 1982, 93, 66. ( 5 ) Vander Hart, D. L. Macromolecules 1979, 12, 1232. (6) (a) Michl, J.; Thulsrup, E. W. Spectroscopy with Polarized Light; VCH: Deerfield Beach, FL, 1986. (b) Meirovitch, E. J . Phys. Chem. 1984, 88,2629. (c) Hentschel, D.; Sillescu, H.; Spiess, H. W. Macromolecules 1981, 14, 1605. (d) Yogev, A.; Riboid, J.; Marero, J.; Mazur, Y. J . Am. Chem. Soc. 1969, 91, 4559. (e) Read, B. E. In Structure and Properties of Oriented Polymers; Ward, 1. M., Ed.; Wiley: London, 1975; Chapter 4. (0 Gottlieb, H. E.; Luz, Z. Macromolecules 1984, 17, 1959. (7) Ramesh, V.; Weiss, R. G. Macromolecules 1986, 19, 1486. (8) (a) Naciri, J.; Weiss, R. G. Macromolecules 1989, 22, 3928. (b) Naciri, J. Ph.D. Thesis, Georgetown University, Washington, DC, 1989. (9) (a) He, Z.; Hammond, G. S.; Weiss, R. G. Macromolecules, submitted for publication. (b) He, Z. Ph.D. Thesis, Georgetown University, Washington, DC, 1991.
diffusional activation energy for the fraction of sites which remains accessible (i.e., whose free volume is large enough to accommodate a dopant molecule or molecules) does not change appreciably when a film is stretched. As a means to understand better the factors governing the accessibility and size distribution of dopant sites in unstretched and stretched LDPE films, we have attached pyrenyl groups covalently at their 1-position to the interior of LDPE films,'O and then used this material to measure the fraction of dopant sites capable of including an additional noncovalently bound molecule and the dynamics associated with that inclusion? The noncovalently attached dopants are a series of homologous NJ-dialkylanilines (DAA) whose van der Waals volumes have
DAA DMA: R = Me DEA: R = Et DPA: R E n-Pr DBA: R = n-Bu
I-pyrenyl group (Py)-
been calculated and which quench the pyrenyl groups in LDPE at a diffusion controlled rate. Experimental Section Materials. Low-density polyethylene (LDPE) films (Sclairfilm 300 LT-1; 76 pm thick; 0.92 g/cm3; and a number average molecular weight ( M , ) of 112 600'') were supplied by DuPont of Canada. Prior to use, film strips were immersed overnight in chloroform to remove antioxidants, rinsed in fresh chloforom and methanol, and dried with a stream of nitrogen. Pyrene (Aldrich, 99%) was recrystallized three times from-95% ethanol, passed through a silica column using 1/4 (v/v) ethyl acetate/hexane as eluant, and sublimed to yield white crystals, mp 150-151 O C (lit.'* ~
~
~~
(10) (a) Lamotte, M.; Pereyre, J.; Joussot-Dubien, J.; Lapouyade, R. J. Phofochem. 1987,38, 177. (b) Lamotte, M.; Joussot-Dubien, J.; Lapouyade,
R.; Pereyre, J. In Photophysics and Photochemistry above 6 eV; Lahmani, F., Ed.; Elsevier: Amsterdam, 1985; p 577. (c) Lamotte, M.; Bagno, 0.; Lapouyade, R.; Joussot-Dubien, J. C. R. Acad. Sci. (Paris) 1984, 299 (II), 1321. (1 1) Private communication from Ann Walksi, DuPont of Canada. (12) Birks, J. B.; Kazzazz, A. A.; King. J. Prog. R. SOC.(London) 1966, A291 556.
0022-3654/92/2096-496%03.00/0 0 1992 American Chemical Society
The Journal of Physical Chemistry, Vol. 96,No. 1, 1992 497
Diffusion in LDPE Films
TABLE I: Putitiopiag Coefficients for N,N-DiaUrylanilnes between Methanol and Py-LDPE Films K O
DAA
temp, O C
DMA
20
30
unstretched 0.18 f 0.01 0.23 f 0.01
40
DEA
20
DPA
40 20 40
DBA
20 30
stretched
(0.19 f O.O1)b (0.21 f 0 . 0 1 ) b (0.21 f 0.01)b
0.15 f 0.01 0.14 f 0.01
0.16 f 0.01 0.18 f 0.01 0.12 f 0.01 0.15 f 0.01 0.10 f 0.01 0.11 f 0.01
(0.15 f O.O1)b (0.13 f O.O1)b (0.13 f O.O1)b
0.12 f 0.01 0.13 f 0.01 0.11 f 0.01 0.12 f 0.01 0.13 f 0.01 0.13 f 0.01
"Each value of K is an average of five experiments with different DAA concentrations: for DMA, 0.3, 0.4, 0.6, 0.8, and 0.95 M; for the other DAA, 0.3, 0.4, 0.5, 0.6, and 0.7 M. From ref 8a, using other Py-LDPE films. mp 149-1 51 "C). N,N-Dimethylaniline (DMA, Aldrich), N,Ndiethylaniline (DEA, Aldrich), N,N-dipropylaniline (DPA, American Tokyo Kasei), and N,N-dibutylaniline (DBA, American Tokyo Kasei) were purified according to the procedure of Meltzer and Tobolsky:13 for DMA, bp 68 OC/lO Torr (lit.14 bp 95 OC/30 Torr); for DEA, bp 91 OC/lO Torr (lit.I4 bp 96 "C/13 Torr); for DPA, bp 65 OC/O.2 Torr (lit.I4 bp 122 OC/14 Torr); for DBA, bp 100 OC/O.2 Torr (lit.I4 bp 130 OC/7 Torr). By gas chromatographic analyses on a 10 m X 0.5 mm RSL 300 wide bore capillary column, all of the DAA were >99% pure. They were stored under nitrogen, refrigerated, and protected from light until being used. 2-(Dimethylamino)ethanol (DAE, Aldrich, 99%) was dried over potassium carbonate and fractionally di~til1ed.l~The fraction with bp 134 OC (lit.16 bp 134 "C) was collected and stored under nitrogen. Chloroform {Fisher ACS grade, 0.7 5% ethanol) and methanol (Baker Photorex reagent) were used as received. heparation of LDPE with Covalently Linked Pyrenyl Group (Py-LDPE): Typically, a 2 X 4 cm piece of LDPE was immersed overnight in a chloroform solution of 0.24 M pyrene. The film was quickly rinsed with methanol to remove any surface pyrene and dried with a stream of nitrogen. The dopant concentration of free pyrene (OD 1 at 300 nm) was 2.7 X lo-* M. The film was suspended in a Pyrex tube which was flushed with nitrogen and irradiated with a 450-W Hanovia medium-pressure Hg arc for 1 h. The irradiated film was rinsed with chloroform aliquots until their UV spectra showed no evidence of pyrene absorbance. A 2.7 X IO4 M concentration of pyrenyl was calculated from an average of six UV absorption measurements at various positions in the dried film and using the film weight and density to estimate the volume. The film was cut longitudinally and one piece was stretched over a smooth mandril to '+00-500% of its original length. The film shrank slightly ( ' P Y PBBIPY
+'
> 'PY> 'Py
(PY'PBB) Py + hv
PY or 'PY
+ A
(4)
(5)
as before at 395 nm as a function of time. Growth curves for stretched Py-LDPE with DPA required a slightly modified procedure. They were measured with films which had been equilibrated initially with methanolic DPA at an elevated temperature, 35 OC, for 30 min prior to being placed in neat methanol at the various temperatures noted. All fluorescence data were recorded in the front-face mode with 1.25-mm slits using a Spex Fluorolog spectrofluorimeter equipped with an Osram 150 W/lXBO high-pressure Xe lamp and an IBM compatible 386 computer. Samples were thermostated (fO.5 "C) by circulating water through a cell block holding the cuvette. Temperature was monitored using a calibrated thermistor that was in contact with the cuvette. Data were collected and analyzed with a Basic program written by Mr. William Craig. Partition Coefficients (K) of N,N-Dielkylanilines between Methanol and Py-LDPE Films (Table I). A weighed Py-LDPE film was placed in 3.5 mL of a thermostated, nitrogen-saturated solution of DAA in methanol. After 1 h, the film was removed and washed rapidly with methanol. The DAA was extracted from the film by placing it in 3.5 mL of methanol for 1 h. Further washing showed no DAA in the methanol by UV/vis spectroscopy. The concentration of DAA in the methanol was determined from the UV absorbance of the solution and the known molar extinction coefficient^:'^ for DMA, e 1942 at 298 nm; for DEA, t 2231 at 304 nm; for DPA, e 2275 at 305 nm; for DBA, t 2481 at 305 nm. Thus K = [DAA]nlm/[DAA]MeoH after equilibration and the concentrations of DAA in methanol before and after equilibration with the film are taken to be approximately the sameS8
Results and Discussion The protocol for attachment of pyrenyl groups at their 1-position to polymethylene chains in the interior of LDPE films has been described previously.* The modified films of Py-LDPE, once formed, are very stable and can be used repeatedly. Since each film preparation is slightly different due to the initial loading of pyrene in the LDPE and the concentration of pyrenyl groups eventually attached to the film, comparisons between unstretched and stretched Py-LDPE were conducted with films that shared the same history. In fact, only one unstretched and one stretched film have been employed in all of the experiments reported in this (17) Eustace, D.; Grunwald, E. J. Am. Chem. SOC.1974, 96, 7171.
490
The Journal of Physical Chemistry, Vol. 96, No. 1 , 1992
Jenkins et al.
TABLE 11: RepresentativeRate Constants at 15 and 35 O C for DAA Diffusion to and from Pyrenyl-OccupiedDopant Sites According to Various Data Treatments
DAA DMA DEA DPA DBA
concn,c M 0.95
0.70 0.70 0.70
unstretched
stretched
unstretched
stretched
unstretched
stretched ~~~~~
35 OC 14.3 28.7d 8.3 6.7 5.2
15 "C 3.6 6.7d 3.0 2.9 2.7
35 OC 10.0 24.7d
15 OC 1.6 2.9d
6.6 3.7
1.8 1.3
35 "C 2.4 4.g 14.4 1.2
0.8 0.5
15 OC 0.6 1.11 3.18 0.4 0.3 0.3
35 OC 1.2 3.6 14.79 0.7 0.4
15°C 0.2
0.d 1.38 0.2 0.1
35 "C 11.9 23.9 14.39 7.1 5.9 4.7
15 OC 3.0 5.6 3.68 2.6 2.6 2.4
35 OC 8.8 21.71 10.09 5.9 3.3
15 "C 1.4 2.9 1.69 1.6 1.2
"An average of 23 runs. bFrom growth curves using eq 9. 'Initial concentration of DAA in methanol used in decay curve experiments and to dope films for growth curve experiments. dFrom decay curves using eq 8. 'From growth curves using eq 9 and K . /From decay curves using eq 8 and K. gUsing eqs 8 and 10. The k,,, values are the same as kin+ k,,, reported from growth curves. paper. The most useful information comes from comparisons between properties of the unstretched and stretched piece. The method for obtaining the data of interest relies upon the fluorescent properties of the pyrenyl excited singlet state and the quenching of that state by DAA molecules.'* A simplified mechanism for the processes involved is shown in Scheme I. Operationally, a fdm is suspended in a methanolic solution of DAA and the fluorescence intensity ( I , hv') is measured as a function of time while the film (and solution) is irradiated with a constant intensity monochromatic light source (hv). When the fluorescence intensity no longer changes, the liquid is exchanged for pure methanol and the intensities are measured again as a function of time. These are called, respectively, decay and growth curves. Since methanol does not swell LDPE, only DAA molecules diffuse into the film or escape from it. Due to the relatively short lifetime of pyrenyl singlet states in Py-LDPE (-200 nsI9) in comparison with the time required for DAA molecules to establish a diffusional equilibrium between methanol and LDPE (several minutes), quenching by DAA should be almost exclusively static: virtually all pyrenyl groups of PyLDPE which do not share their dopant site with at least one DAA molecule a t the moment of excitation will not be quenched; all pyrenyl groups sharing their dopant sites with a t least one DAA molecule at the moment of excitation will be quenched. Experiments conducted with LDPE films containing covalently substituted anthryl groups support this hypothe~is.~ In essence, eq 5 is an unimportant contributor to the quenching scheme. Additionally, it is assumed implicitly in our data treatment that pyrenyl groups which do and do not share their dopant sites with a DAA molecule have equal excitation probabilities. Since a pyrenyl group does not form ground-state complexes with DAA molecules,18their influence on the pyrenyl absorption spectrum will be relatively minor. Thus, the changes due to the presence of DAA in the films can be related directly to the fraction of pyrenyl occupied dopant sites that can also accommodate at least one DAA molecule. Rate Constants for Diffusion of DAA Molecules to and from Pyrenyl-Occupied Sites in LDPE. Based on these considerations, data from the decay and growth curves can be replotted according to eqs 8 and 9, respectively.* Experimentally, Zo(decay) is determined when a supported film is placed in a cuvette filled with methanol; I,(decay) = Zo(growth)is measured after the methanol has been removed, replaced by methanolic DAA without moving thefilm or nwette, and waiting until the intensity stabilizes; finally, I,(growth) is obtained sequentially by removing the methanolic DAA solution, replacing it with neat methanol (again, without film movement), and waiting until the signal stabilizes. Ir is the intensity at time t after either methanolic DAA is added to the (18) (a) Gordon, M.,Ware, W. R., Eds. The Exciplex; Academic: New York, 1975. (b) Birks, J. B. Photophysics of Aromatic Molecules; WileyInterscience: London, 1970; Chapter 9. (c) Taylor, G. N.; Chandross, E. A.; Schiebel, A. H. J. Am. Chem. SOC.1974, 96, 2693. (19) Lifetimes of 500 ns have been reported for pyrene singlets under optimal condition^.'^^ Even they would not alter our conclusion. The ca. 200-11s lifetime for pyrenyl singlets in Py-LDPE films is a realistic estimate. (a) Wintgens, V. In Handbook of Organic Photochemistry; Scaiano, J. C., Ed.; CRC Press: Boca Raton, FL, 1989; Vol. I, Chapter 17.
SCHEME II: Representation of the Steps Associated with Partitioning of DAA Molecules between Methanol and LDPE Films Accordmg to Thermodynamic and Kinetic Models"
"The vertical line is the interface between methanol and an LDPE film. A and B are sites within LDPE along the route to or from a pyrenyl-occupied site, the circled Py. cuvette (decay) or is replaced by neat methanol (growth). Due to the lack of cross-sectional homogeneity of amorphous and crystalline regions, movement of the film during measurements of either Io I , sequence must be avoided.
-
(9)
The sum of the rate constants for DAA motion to a pyrenyloccupied dopant site and escape from it, kin k,,,, can be calculated from semilogarithmic plots of the data.* Representative curves and their semilogarithmic analogues are shown in Figures 1 and 2 and the kin k,,, for each DAA as calculated from eqs 8 and 9 are included in five tables in the supplementary material (seeparagraph at the end of this article and any current masthead page for ordering instructions). Table I1 contains values of kin and k,,, at 35 and 15 O C as calculated by the two approaches described above. The precise meaning of kin+ k,,, in these experiments requires some explanation. Entry of a DAA molecule into the film is insufficient to cause quenching of pyrenyl fluorescence; the DAA must enter a pyrenyl-occupied site within the film to be measureable in Scheme I. Thus, kinrepresents the rate-limiting step in the process which transports DAA across the methanol/LDPE interface (A), through the amorphous region of the film (B), and finally into a pyrenyl-occupied dopant site (Scheme 11). The other rate constant, k,,,, represents the slow step for separation of a pyrenyl-DAA contact pair by at least one polymethylene chaini8 (Le., for escape of a DAA molecule from a pyrenyl-occupied site). Since the activity of a DAA molecule in neat methanol (as in growth curve experiments) and in methanol containing a large concentration of DAA (as in decay curve experiments) need not
+
+
Diffusion in LDPE Films
The Journal of Physical Chemistry, Vol. 96, No. I, 1992 499
If
Figure 1. Representative fluorescence growth curves from unstretched Py-LDPEin methanol at 25 OC after being equilibrated with 0.7 M DBA in methanol (a), 0.7 M DPA in methanol (b), 0.7 M DEA in methanol (c), and 0.95 M DMA in methanol (d). hX= 343 nm; h,,,= 395 nm.
Figure 2. Treatment of data presented in Figure 1 according to eq 9. Correlation coefficients for fits to linear slopes of each data set are 0.991 (a, DBA),0.992 (b, DPA),0.994 (c, DEA),and 0.997 (d, DMA).
be the same, kinfrom the two types of plots may have slightly different values. Additionally, the nature of the key microscopic process in growth curves, k-3 in Scheme 11, combined with the vast volume excess of methanol and K being