E. M.
1932
(Otrs-l D) that are used in eq A.8. Since the subroutine computes the column vector (Otrs-l D)“instantaneously” while performing the diagonalization, the square matrix Ot’ is not stored as it would be in other diagonalization procedure^.^^^^^ References and Notes (1) M. T. Watts, M. L. Lu, R. C. Chen, and M. P. Eastman, J. phys. Chsm., 77, 2959 (1973). (2) R. Chang and J. Piehler, Bull. Chem. Soc. Jpn., 46, 1296 (1973). (3) M. A. Komarynsky and A. C. Wahl, J. fhys. Chem., 79,695 (1975). (4) C. P. Cheng and S. I. Weissman, J. Phys. Chem., 80, 872 (1976). (5) M. P. Eastman, D. A. Ramirez, C. D. Jaeger, and M. T. Watts, J . Phys. Chem., 80, 182 (1976). (6) M. P. Eastman, C. D.Jaeger, D. A. Ramirez, S. L. Kelly, and P. W. Percival, J. Magn. Reson., 22, 65 (1976). (7) B. Kaempf, S. Raynal, A. Collet, F. Schue, S.Boileau, and J. Lehn, Angew. Chem., Int. Ed. Engl., 13, 611 (1974). (8) M. A. Komarynsky and S. I. Weissman, J . Am. Chem. Soc., 97, 1569 (1975). (9) 0. W. Webster, W. Mahler, and R. E. Benson, J. Am. Chem. Soc., 84, 3678 (1962). (10) K. Hofelman, J. Jagur-Grodzinski, and M. Szwarc, J . Am. Chem. Soc., 91, 4645 (1969). (11) B. F. Wong and N. Hirota, J . Am. Chem. Soc., 94, 4419 (1972). (12) N. Kushibiki and H. Yoshida, J . Am. Chem. Soc., 98, 268 (1976).
(13) (14) (15) (16) (17) (18) (19) (20) (21) (22)
L. H. Piette and W. A. Anderson, J. Chem. Phys., 30, 899 (1959). R. A. Marcus, J . Chem. fhys., 43, 3477 (1965). S. Chandrasekhar, Rev. Mod. Phys., 15, 1 (1943). N. Hirota, J . Phys. Chem., 71, 127 (1967). N. Hirota, R. Carraway, and W. Shook, J . Am. Chem. Soc., 90, 3611 (1968). G. L. Malinoski, W. H. Bruning, and R. Grlffin, J. Am. Chem. SOC., 92, 2865 (1970). (a) T. J. Katz, J. Am. Chem. Soc., 82, 3785 (1960); (b) T. J. Katz and H. L. Strauss, J . Chem. fhys., 32, 1873 (1960). H. L. Strauss, T. J. Katz, and G. K. Fraenkel, J. Am. Chem. Soc., 85, 2360 (1963). F. J. Smentowski and G. R. Stevenson, J. Am. Chem. Soc., 89, 5120 (1967). P. H. Riege;, I. Bernal, and G. K. Fraenkel, J . Am. Chem. SOC., 83. ._. 3918 .. - 11961). -R. A. Marcus, J . Chem. fhys., 43, 679 (1965). D. Live and S. I. Chan, J . Am. Chem. SOC.,98, 3769 (1976). J. A. Pople, W. G. Schneider, and H. J. Bernstein, “High-Resolutlon Nuclear Magnetic Resonance”, McGraw-Hill, New York, N.Y., 1959, Chapter 10. R. 0 . Gordon and R. P. McGlnnls, J. Chem. fhys., 49, 2455 (1968). R. G. Gordon and T. Messenger, “ESR Relaxation In Llqulds”, L. T. Muus and P. W. Atkins, Ed., Plenum Press, New York, N.Y., 1972, Chapter 13. G.V. Bruno, Ph.D. Thesis, Cornel1 University, 1973. N. Haran, 2. Luz, and M. Shporer, J . Am. Chem. Soc., 96, 4788 (1974). I
(23) (24) (25) (26) (27) (28) (29)
Schulman and R. T. Parker
I
Room Temperature Phosphorescence of Organic Compounds. The Effects of Moisture, Oxygen, and the Nature of the Support-Phosphor Interaction’ E. M. Schulman” and R. T. Parker Department of Chemistty, University of South Carolina, Columbla, South Carollna 29208 (Received May 23, 1977)
The effects of moisture, 02, and the nature of the support-phosphor interaction on room temperature phosphorescence (RTP) of sodium 4-biphenylcarboxylateand sodium 1-naphthoate (polar organic compounds) adsorbed on paper, sucrose, starch, glass fiber paper, silanized paper, and others were studied. It is found that an abundance of surface hydroxyl groups are necessary for a suitable RTP support, and the presence of moisture in the atmosphere serves to both allow collisional deactivation of the excited phosphor and permit increased quenching by aiding in the transport of O2 into the sample matrix. The results support the hypothesis that hydrogen bonding of the phosphor to the support is the primary mechanism of preventing collisional deactivation of the excited phosphor by restricting translational, vibrational,and rotational movements,and a sample matrix is set up which resists the penetration and subsequent quenching effects of 02.
Introduction Isolated cases of phosphorescence from organic molecules in fluid solution have been observed with particularly rigid molecules,2 rigorously deoxygenated solutions of dyestuff^,^ and by use of spin forbidden transition enhancers such as dimethylmer~ury.~However, strong phosphorescence of organic compounds is a phenomenon5 which has until recently6-8been normally observed only in the gas phase, in rigid media, or at cryogenic temperatures. Such techniques were required to prevent the nonradiative decay of the triplet state by collisional energy transfer (internal conversion processes) and to inhibit quenching by ground state triplet oxygen. Room temperature phosphorescence (RTP) in the air of a wide variety of polar organic compounds when adsorbed on silica, alumina, and cellulose was first observed by Roth6 and, later, independently measured by Schulman
* Address correspondence to this author at the Department of Chemistry, State University College at Buffalo, 1300 Elmwood Ave., Buffalo, N.Y. 14222. The Journal of Physical Chemistry, Voi. 81, No. 20, 1977
and Walling.7ps The analytical utility of room temperature phosphorimetry, which does not require the time consuming and cumbersome cryogenic techniques normally required for phosphorimetry,was pointed out by Schulman and Walling,7i8and the qualitative and quantitative applicability of the method was demonstrated by Winef~rdner.~-llIncreasing the quantum efficiency of room temperature phosphorescence (RTP) at the expense of lifetime and fluorescence by use of the “external heavy atom effect” has been exploited by Seybold,12and applied to improve analytical results for compounds which phosphoresce weakly at room temperature by Winef0rdner.l’ A rapid and selective method for the determination of p-aminobenzoic acid at room temperature when adsorbed on a sodium acetate support has been reported by Hurtubise.13 Although much work has been done to develop RTP as an analytical technique, no systematic investigation of the physical basis of this novel and exciting phenomenon has been reported to date. We suggested earlier that surface adsorption of the phosphor molecules to the support
1933
Room Temperature Phosphorescence
somehow inhibits collisional deactivation of the triplet state and restricts oxygen quenching when the sample is rigorously dried.7i8 Jones has investigated organic phosphors imbedded in poly(methy1methacrylate) and other polymer films which serve to restrict collisional deactivation of the phosphor at room temperature and has found that oxygen is sufficiently small to diffuse into the polymer film with time and quench triplet states.14J5Lifetime studies have shown that both moisture and oxygen are effective quenchers of the room temperature phosphorescence of tryptophan residues in wool keratin and in poly(viny1 alcohol) matrices. Quenching of the phosphorescence due to moisture from the poly(viny1 alcohol) film was attributed to the disruption of the hydrogen bonding network and subsequent loss of rigidity of the film, thus allowing collisional deactivation to occur.16 Lewis' classic (1941) study of fluorescein solutions in glycerol and crystal oil (both of which become viscous and rigid in the same temperature range) showed higher luminescence intensity (both E-type delayed fluorescence and phosphorescence were being observed) from glycerol than from hydrocarbon solutions of the same viscosity.17 Lewis ascribed this increased luminescence yield to the higher molecular rigidity associated with the hydrogen bonding network of the triol. Lewis also noted variation of phosphorescence intensity from boric acid glasses depending on incorporated water. We here report the effects of moisture and oxygen on organic phosphors adsorbed on paper at room temperature and the ability of various materials to serve as a suitable support for RTP. Experimental Section Apparatus. An Aminco-Keirs spectrophosphorimeter equipped with the standard rotating can phosphoroscope, a 1P28-A photomultiplier tube, and an Aminco 10-222 photomultiplier photometer was used throughout for phosphorescence intensity measurements and excitation/emission spectra. The spectrophosphorimeter was adapted to utilize an Illumination Industries Inc. 150-W xenon short arc lamp with ellipsoidal condensing assembly as the excitation light source. An iron-constantan thermocouple connected to a strip chart recorder was inserted through the nitrogen inlet on the sample cell compartment for continuous monitoring of the cell temperature. For room temperature operation, the cell temperature was maintained at 301.5 f 0.5 K by means of two whisper fans mounted above the cell compartment which offset heat radiated by the light source. Higher operating temperatures of the cell could be obtained by means of a resistor heating element with thermistor control installed in the cell compartment. A sample holder (Figure 1)similar to one described by Winefordnerg was designed for use with samples on paper at room temperature and replaced the standard quartz Dewar assembly normally used at 77 K. Reagents. Reagent grade chemicals of the highest available commercial purity were used as received. Argon and oxygen (Liquid Air Inc., Columbia, S.C.) were dried over Drierite or humidified to 3.2,8.5,18.8,37.1,58.3,80.5, and 100% relative humidity at 298 K by passage through three gas wash bottles each containing 330 mL of aqueous sulfuric acid solution of the appropriate density to yield the given humidities.18 A flow rate of 500 mL/min was used and entrained H2S04aerosol was trapped by passage through a glass tube packed with glass wool. Solution Preparation. A basic methanol solvent system (0.1 M NaOH/5.56 M H,O/methanol) was prepared by dilution of 100 mL of 0.1 M NaOH (aqueous) to 1L with methanol. Sodium 4-biphenylcarboxylate (NaBPCA)
-
0.25"
A-
P I
5.4"
C
I
Flgure 1. Sample holder used in RTP studies. The holder allows mounting of paper circles with optimum geometry for exckation/emission measurements and replaces the standard quartz Dewar assembly. Note the hollow tube A that allows passage of desired gas directly over the face of the sample B which is held behind mask C exposing a 5 X 5 mm surface.
standard solutions were prepared by successive dilution of a stock solution of 10 mM 4-biphenylcarboxylic acid in the methanolic base with additional methanolic base solution. A neutralized solution of 1 mM sodium 1naphthoate (NaNA) was prepared by titrating 5 mL of 10 mM 1-naphthoic acid in 0.1 M NaOH (aqueous) with 0.1 M HC1 to a phenolphthalein end point and diluting the neutralized solution to 50 mL with distilled water. No emission is observed from phenolphthalein with the experimental conditions used. Preparation and Equilibration of Sample. Sample volumes of 5 pL and 9/32-in. circles of Whatman No. 1 filter paper were used for all spectral and quantitative measurements unless otherwise noted. The filter paper circle was mounted on the sample holder (Figure l), masked to expose a 5 X 5 mm surface, and the sample solution was applied to the paper. The sample was then placed in the sample compartment of the spectrophosphorimeter, and a dry or humidified gas of choice flushed directly over the surface. Analytical Calibration. A calibration plot was constructed (Figure 3) from four repetitive runs at each concentration of the standard NaBPCA solutions using excitation and emission maxima of 280 and 490 nm, respectively. Samples were desiccated under dry argon and relative intensity was monitored by a strip chart recorder for a period of 20 min from the time of sample introduction. Sample desiccation was complete within 15 min (for analytical purposes, 10-12 min in a warm cell is sufficient). The emission intensity was taken as the average reading between 16 and 20 min after sample introduction. Humidity Experiments. The phosphorescence intensity of a standard 0.5 mM NaBPCA solution was measured as a function of humidity and oxygen concentration by first equilibrating the sample under a flow of humidified argon for 20 rnin and then switching to a flow of oxygen at that same humidity for 30 min. Two minutes of flow through the gas handling train were necessary for the new gas to reach the cell compartment. Sample intensities were recorded on a strip chart recorder (typical plots are shown in Figure 4, regions A-C). Triplicate runs were made with The Journal of Physlcal Chemistry, Vol. 8 1, No. 20, 1977
E. M. Schulman and R. T. Parker
1934
TABLE I: Relative Intensities of Sodium 4Biphenylcarboxylate Samples in Ar and 0, as a Function of Relative Humidity
%Ha
Ikb
0 3.2 8.5 18.8 37.1 58.3 80.5 100
100 98.1 91.6 57.6 12.8 2.6 0.7 0.3
Rsd in Arc 1.2 1.1 2.4 1.2 2.6 7.1 4.3
11
1
loo]
Rsd in
IO,
0,‘
70.9 70.3 61.3 25.5 2.5 0.4 0.1 0.0
1.5 1.3 1.9 3.8 14 2.9 17
Q H , ~Qo,~
0 1.9 8.4 42.4 87.2 97.4 99.3 99.7
29.1 27.8 30.3 32.1 10.3 2.2 0.6 0.3
Percent relative humidity of gas at 298 K. Intensity, in argon relative to 0%humidity in argon ( Iko). Calculated from triplicate runs and given in percent. Intensity in oxygen relative to Z ~ O . e 1 ~ -0 I&. f I&-Io,. a
each humidified gas prepared, and emission intensities were noted after 20 min equilibration time in argon and 28 min of exposure to oxygen (Table I). Attempted Observation of Phosphorescence from Fluid Solution. A standard 0.5 mM NaBPCA solution was deoxygenated by 12 freeze-pumpthaw cycles Torr), and the resultant solution sealed off in a standard low temperature quartz sample tube and scanned for phosphorescence at 302 and 77 K. No emission was noted at the highest instrument sensitivity in the fluid solution; normal emission was seen in the frozen sample. Phosphorescence Intensity as a Function of Support Material. Aliquots of 1mM NaNA in neutralized and 0.1 M NaOH aqueous media, and 1 mM NaBPCA in 1 M NaOH (aqueous) were dried (overnight over Drierite) on each of the following supports: Whatman No. 1 paper, sucrose, starch, Whatman No. 1 (silanized), glass fiber paper (Reeve-Angle 100% borosilicate glass), SWP polyethylene fiber paper (a water wettable polymer “paper”), 50% SWP polyethylenedO% cellulose fiber paper,lg and a stainless steel plate. The supports were then illuminated with both the long and short UV wavelength settings of a hand held 15-W mineralight for a visual check of phosphorescence. Silanized Paper. Whatman No. 1filter paper was silanized by immersion in neat dimethyldichlorosilane for 1h, dried in the air, and allowed to stand for 30 days before use. Spectra and quantitative measurements were made on silanized and regular paper using 4 pL of neutralized 1mM NaNA solution on 1/4-in.circles of paper (masked to 4 X 5 mm) under dry argon. Excitation and emission maxima of 295 and 570 nm, respectively, were used; and the spectra are presented in Figure 8. Sucrose. Aliquots (0.2 mL) of 10 mM NaNA in 1 M aqueous NaOH were triturated with 0.5 g of sucrose, dried over Drierite for 24 h, and pulverized. Samples were placed in Kimax melting point tubes, sealed, attached to the paper sample holder, and spectra were measured at both 311 and 286 K using an Illuminations Industries 350-W mercury lamp. Comparison spectra of the sucrose sample adhering to adhesive tape mounted on the same holder, and of 1.6 p L of solution on paper were made under dry air (1L/ min). The samples sealed in Kimax showed wavelength shifts due to the absorption characteristics of the glass and were not compared with the paper. Comparison of 77 and 302 K NaBPCA Phosphorescence. A spectrum using the standard quartz Dewar assembly was obtained for a standard 1 mM NaBPCA solution at 77 K and identical instrument slits were used to obtain a 302 K spectrum of the sample on paper under The Journal of Physlcal Chemistry, Vol. 81, No. 20, 1977
.9.. .2......
”
0
10
I
20
30
TIME, Tin
Figure 2. Relative intensity vs. time plot for the desiccation of NaBPCA samples in Ar (-), air (----), and Os(. , , .) at 0% relative humidity. Emission intensities after 30 min of desiccation in Ar, air, and O2are 100, 76, and 46, respectively.
dry argon. The low temperature sample formed a cracked glass upon freezing and yielded an intensity comparison of 1.06 (480 nm) to 1.00 (485 nm) for the RTP sample.
Results Humidity and Oxygen Quenching Study. In order to investigate the specific effects of moisture and oxygen on the RTP of adsorbed organic molecules, a reproducible sample introduction system had to be developed. The sample holder shown in Figure 1was constructed to mount on an Aminco-Keirs spectrophosphorimeter in such a manner that sample solutions applied to paper circles mounted on the holder could be reproducibly introduced into the cell compartment and aligned with optimum geometry for front surface excitation and emission measurements. A sample could be dried or equilibrated in atmospheres humidified to various degrees while in the instrument cell compartment by passing a selected dry or humidified gas directly over the sample and into the cell compartment. Since phosphorescence intensity from samples adsorbed on paper is highly subject to quenching by m ~ i s t u r e , ~ ~ * drying or equilibration progress could be followed by monitoring the phosphorescence intensity of a sample vs. time spent in the cell compartment. Typical data plots of a sample drying in the sample compartment under flows of dry Ar, air, and O2 can be seen in Figure 2 for sodium 4-biphenylcarboxylate (NaBPCA) in methanolic base. The reproducibility of the sample preparation procedure was demonstrated by constructing a calibration plot (Figure 3) from replicate runs on a series of NaBPCA solutions. Four runs were made at each concentration, and the average of the relative standard deviations (rsd’s) for the data points in the concentration range of 0.01 to 2.0 mM was 2.4 % . This can be compared with the 5-7 % rsd reported for RTP of PABA on sodium acetate.13 The 0.5 mM NaBPCA solution (rsd = 1.9%) was found to lie in a slightly nonlinear, although quantitatively useful, section of the calibration plots and was selected as the concentration to be used in subsequent experiments. These experiments consisted of measuring the emission intensity of the compound under atmospheres of Ar and O2humidified to various degrees. The wet sample was first equilibrated under a flow of Ar at a given humidity (produced by equilibration of the flowing gas with aqueous sulfuric acid solutions) and then subjected to an atmosphere of O2at the same humidity while emission intensity was monitored vs. time. Typical data plots of runs at 3.2 and 80.5% relative humidity are shown in Figures 4a and
Room Temperature Phosphorescence
1935
\
75-1
>.
\
k
Lc
z W c z
- 50W
-
c
c c3
3 01
0'
33
02
CONCEIT9cT C h
13
20
clcl
I
Figure 3. Calibration plot for NaBPCA samples desiccated in the cell compartment under dry Ar. The error bars represent f l standard deviation of four runs.
i
I
3 , 0
\
\
I
20
40
63
80
Q E L A T I V E HLM I3 T V
Figure 5. Relative intensities of NaBPCA samples in Ar (curve A = IAr,-) and O2 (curve B = Io*, -) as a function of humldity.
---
Y n
A
0
1 0 1 15 22
C
I
D
52 T I M E , min
b)
T t M E , ri r
Figure 4. Relative intensity vs. time plots showing oxygen quenching effects on NaBPCA samples first equilibrated in Ar at (a) 3.2% and (b) 80.5% relative humidity: (region A) equilibrium region, 0-15 min; (region B) equilibrium intensity region, 15-22 min; (region C) oxygen quenching region, 22-52 min; (region D) regeneration of oxygen quenched sample, 52-62 min.
4b. Region A of Figure 4 (equilibration region: 0-15 min) shows the growth of emission intensity as the solvent
evaporates from the paper and the samples equilibrate with the Ar atmosphere. Region B (equilibrium intensity region: 15-22 min) shows the constant emission intensity obtained from samples which have reached equilibrium with the Ar atmosphere. Region C (oxygen quenching region: 22-52 min) shows the decline of emission intensity when the samples are exposed to the oxygen atmosphere. O2 quenching occurs to some degree for each different humidity used in this experiment including those at 0% relative humidity. Experimental runs were normally terminated after a 28-min exposure of the sample to the O2atmosphere; however, region D (regeneration of oxygen quenched sample: 52-62 min) shows the regeneration of O2quenched samples by flushing with the same Ar atmosphere used in the initial equilibration stages (regions A and B). Note that emission intensity increases continually throughout region D but does not, in the time of observation, reattain the intensity of region B. Table I presents the results obtained a t the various humidities in both oxygen and argon. Relative intensities are reported from the average of triplicate runs under the Ar atmosphere ( I A J and the oxygen atmosphere (lo,) at the stated humidity and are relative to an emission intensity of 100% for NaBPCA samples desiccated under dry argon (I*:). The emission intensities from samples under Ar were measured after 20 min of equilibration (region B, time = 20 min) while the intensities from the samples under O2 were taken after 28 min of exposure (region C, time = 50 min). It should be noted that emission intensities taken under Ar atmospheres represent equilibrium conditions and reflect the total amount of quenching due to moisture; however, emission intensities measured under O2 atmospheres are near but not at equilibrium conditions with respect to oxygen diffusion into the sample matrix. Experimental and time factors prohibited monitoring emission intensities until the slow diffusion of O2 into the sample matrix was complete. Figure 5 shows plots of the relative intensities from the NaBPCA samples in Ar ( I h , curve A) and in O2 (lo2,curve B) as a function of humidity. The relative intensity plot in Ar shows the quenching effect of moisture alone as the humidity increases while the relative intensity plot in O2 The Journal of Physlcal Chemistry, Vol. 81, No. 20, 1977
E. M. Schulman and R. T. Parker
1936
TABLE 11: Visible Phosphorescencea from Compounds on Various Supports NaNAb NaNAb (0.1 M NaBPCAC (neutralized NaOH (1M NaOH media) aqueous) aqueous)
Support materid
I
c , c
I
2c
I
a@ ?:-AT1
* -
-.--
- ,
-
- .-
..
.. I
60
62
-. -.
130
i L 4 u d Dl-Y
Figure 6, Percentage of total quenching in O2due to H20 (curve C = (lOOQHzO/Q,),-)and O2(curve D = (lOOQo,lQt), ----) vs. relative humidity for NaBPCA samples.
Whatman No. 1 paper Sucrose Starch 50/50 SWP polyethylene fiber-paper blend Whatman No. 1paper (silanized) Glass fiber paper (100% borosilicate) SWP polyethylene fiber “paper” Stainless steel
ttt ttt ttt ttt
ttt ttt ttt ttt
ttt ttt ttt ttt
+
t
t
-
-
-
-
-
-
-
I
-
a t t t indicates strong emission; t indicates weak but notably above background; - indicates background or no emission. Sodium 1-naphthoate. Sodium 4biphen ylcarboxylate.
’L’c-
r
,A
* A
5 EZS-r
,P
-r
Figure 8. Comparison of sodium 1-naphthoate spectra on paper (-) and on silanized paper (----) at 302 K. Figure 7. Degree of O2quenching represented by QozlIA,plotted as a function of relative humidity for NaBPCA samples.
shows the combined quenching effects of both moisture and O2 as a function of humidity. In order to sort out the effects of moisture and O2 quenching, a graph was constructed (Figure 6) showing the percentage of the total amount of quenching due to moisture (curve C ) and that caused by oxygen (curve D) as function of humidity. The amount of total quenching (QJ that occurred in O2 at stated humidity is given by eq 1. The amount of
quenching due only to moisture ( Q H z 0 ) at given humidity can be calculated by eq 2. The amount of quenching due
only to O2 (Q0J at stated humidity is given by eq 3. Figure Q02 = Qt -
QH2O
= IAr - ’02
6 shows the ratio of (QHZo/Qt)plotted as a function of humidity (curve C ) and the ratio of (QOz/Qt) plotted in the same manner (curve D). Increasing the humidity of the O2 atmosphere increases the amount of O2 penetration to the sample as noted by larger degrees of emission quenching at higher humidities in 02.Figure 7 presents this phenomenon graphically by showing a plot of Qo,/I& as a function of humidity. Support Material Study. In order to further investigate the nature of the support-phosphor interaction that results The Journal of Physical Chemistry, Vol. 81, No. 20, 1977
in observation of RTP, we varied the nature of the support used with carboxylate phosphors. Various supports were spotted with sodium 1-naphthoate (neutralized and 0.1 M NaOH solutions) and NaBPCA (1M NaOH solution) in aqueous media and desiccator dried overnight. The supports were then illuminated with both the long and short wavelength settings of a 15-W Mineralight UV source for a visual check of phosphorescence. The results were found to be the same for all solutions and are tabulated in Table 11. Strong visible phosphorescence is noted from all supports containing an abundance of surface hydroxyl groups such as paper, sucrose, starch, and 50% SWP polyethylene fiber-50% paper blend. No appreciable phosphorescence is observed from supports containing few or no surface hydroxyl groups such as SWP polyethylene fiber “paper”, borosilicate glass fiber paper, or stainless steel. Whatman No. 1 filter paper was treated to remove available hydroxyl groups by silanization in dimethyldichlorosilane and a 94% reduction in phosphorescence intensity for a neutralized sodium 1-naphthoate sample (NaNA) was realized compared to the same sample on untreated Whatman No. 1 under identical conditions, as shown by the spectral comparison in Figure 8. Spectra were recorded for NaNA on both sucrose and paper at 311 K under identical instrument resolution conditions and are shown in Figure 9. Resolution and maxima for excitation and emission are virtually identical for NaNA on each support at 311 K, but due to different geometrical factors for each support, no relevant intensity comparison can be made. Spectra of NaNA on sucrose were recorded at 311 and 286 K and are shown in Figure
Room Temperature Phosphorescence
!
1937
I
A
I
\
n
/ /
I 2 0c
40i
303
W‘AJ1,EUGTI
2C0
303
4
0c
/iA\EL_\\ST i
733
I
2cc
33:
\
I
I
1 5c3
ICC .lyAdt.E\uTr
,
EC3
6C3
-r-
Figure 9, Comparison of sodium 1-naphthoate spectra on paper (-) and sucrose (----) at 311 K. Intenslties are arbitrarily adjusted.
I
CC3 ‘171
533
r~
Flgure 10. Spectra of sodium I-naphthoate on sucrose (sample in a sealed Kimax capillary tube) at 31 1 (-) and 286 K (----). Intensities are arbitrarily adjusted. Since the sample was sealed in Kimax, the spectra are shifted with respect to those in Figure 9, due to the absorption characteristics of the glass.
10. It is interesting to note that the shoulder in the 560-575-nm range became much more resolved at the lower temperature, and an intensity gain of approximately 15% is realized at the lower temperature. Although NaNA on paper shows a similar intensity gain at the lower temperature, no additional resolution is observed for the shoulder in the 560-575-nm range. Figure 11 shows a comparison of the spectra of a frozen solution of NaBPCA at 77 K with an identical sample dried on Whatman No. 1at 302 K. The 77 K spectrum is better resolved; however, excitation and emission wavelengths are essentially the same with the RTP sample showing only a 5-nm red shift.
Discussion Our original hypothesis was that surface adsorption of the phosphor molecules on the support provides the rigidity necessary to prevent nonradiative collisional deactivation of the excited triplet state and also restricts O2 quenching when the sample is rigorously dried.7i8 We herein extend our hypothesis by offering hydrogen bonding of the ionic organic molecules to hydroxyl groups on the support as the primary mechanism of providing the rigid sample matrix for RTP; and we propose that moisture acts to disrupt the hydrogen bonding network (with subsequent loss of sample matrix rigidity) while aiding in the transport of 02 into the sample matrix. The experimental work of this study was designed to investigate the validity of our original hypothesis and help clarify the physical nature of RTP. The results give support to the hypothesis and insight into the RTP phenomenon.
Figure 11. Comparison of sodium 4-biphenylcarboxylate spectra at 77 (-) and 302 K (- --). Intensities are arbitrarlly adjusted.
-
Moisture and Oxygen Effects. These effects on RTP confirm that both can independently quench phosphorescence as evidenced by the relative intensity data in humidified Ar and O2 atmospheres presented in Table I and Figure 5. In the absence of 02,moisture acts alone as an extremely powerful quenching agent (Figure 5, curve A) showing a moderate quenching effect at low humidities (approximately 10% of the emission intensity is lost at 9% relative humidity for NaBPCA samples in Ar) while increasing to quite dramatic levels at humidities encountered under typical atmospheric conditions (more than 90% of the emission intensity is quenched at 40% relative humidity for NaBPCA samples in Ar). Increasing humidity decreases the RTP effect to the point that none can be observed as the atmosphere becomes saturated with water. In light of our results that only polar or ionic organic compounds p h o s p h o r e s ~ ewhen ~ ~ ~ adsorbed on supports containing an abundance of hydroxyl groups20(Table 11) and as a result of our surface studies, vide infra, we conclude that moisture must be acting to compete with the surface hydroxyl functions for hydrogen bonding to the phosphor molecules and/or to tie up hydroxyl groups on the support necessary to hold the phosphor molecules rigid; i.e., water “softens” the matrix, allowing collisional deactivation. Quenching by triplet ground state O2 is evident in the absence of moisture (29% of the emission intensity is quenched for NaBPCA samples desiccated under Ar and then exposed to dry O2 for 28 min), but the degree of O2 quenching is greatly facilitated by the presence of moisture (80% of the emission intensity is quenched for NaBPCA samples equilibrated under Ar at 37 % relative humidity and then exposed to O2 at the same humidity for 28 min) as shown in Figure 7. The sample matrix, when desiccated under Ar, is resistant to subsequent O2quenchings. Region C of Figure 4 shows the phosphorescence intensity from NaBPCA samples declining with time upon exposure to the 02 atmosphere. The emission intensity does not show an immediate drop to some level of O2quenching as would be expected if O2 diffused freely to the sample, thus the sample matrix must be resisting the penetration of O2to the phosphor molecules. Similar resistance to O2 penetration was observed by Jones for phosphors embedded in polymer film^.^^,^^ The rate of O2 diffusion was also shown to be much faster into triphenylmethylcarboxylic acid crystals than into the corresponding sodium salts by the ESR studies of Janzen, Johnson, and Ayers.2’ We have found that the organic acid phosphors show much weaker RTP than their corresponding sodium salts,’$* although both are capable of hydrogen bonding to the hydroxylic support. Furthermore a comparison of region C in Figures 4a (3.2% relative humidity) and 4b (80.5% relative huThe Journal of Physical Chemlstty, Vol. 81, No. 20, 1977
1930
midity) shows that both the rate of O2 penetration and degree of O2 quenching are increased tremendously at high humidity. Figure 7 illustrates this phenomenon graphically showing that the degree of O2 quenching increases progressively with increasing humidity, exhibiting unusually rapid increases in the degree of O2quenching on NaBPCA samples between 7 and 35% relative humidities. Oxygen not only diffuses into the sample matrix, but also can diffuse back out of the sample matrix as well. Sample emissions which have been partially quenched by exposure to humidified O2 for 30 min show increases in emission intensity after exposure to Ar at the same humidity as can be seen in region D of Figure 4. Although we have not been able to fully regenerate a sample from Q2 quenching in the time of observation, the presence of moisture greatly aids in the rate and degree of regeneration of sample emission as can be seen by comparing region D of Figures 4a and 4b. In the first 10 min after the O2 quenched sample is reexposed to Ar, a 44% recovery from the total amount of oxygen quenching is realized at 80.5% relative humidity while only a 19% recovery is made at 3.2% relative humidity. Inspection of the data indicates that moisture increases the permeability of the sample matrix to 02. The sample matrix shows resistance to penetration of O2 when prepared under Ar and then exposed to 0,; however, when samples are desiccated under O2or air, O2 is incorporated into the sample matrix. Such O2 incorporation effects probably were a large source of random error in early attempts to make analytical application of RTP. The preparation of samples by drying in the absence of Q2 greatly reduces such random incorporation of O2and greatly improves the reproducibility of the method, even when samples are run in air. Oxygen quenching effects are shown in the emission intensity vs. time plots of Figure 2 for NaBPCA samples desiccated under dry Ar, air and 02. Although the desiccation process is complete for samples dried in air and O2 within 15 min, emission intensities continue to increase slightly vs. time for the dry samples. It seems evident that O2 has been incorporated into the sample matrix upon desiccation and diffuses out of the sample matrix with time reducing the oxygen quenching effect observed for the dry samples. The degree of oxygen quenching is also found to be dependent on the concentration of oxygen present in the atmosphere under which the sample is dried. A 24% quenching of emission intensity is noted for a 30-min desiccation of the sample in air while 54% is quenched in O2 compared to a similar sample desiccated in Ar (Figure 2). The combined effects of moisture and O2 are quite detrimental to the excited triplet state emission process for RTP as can be seen in curve B of Figure 5. Figure 6 shows the relative contributions of oxygen and moisture to the overall quenching experienced by NaBPCA samples after 28 min of exposure to humid oxygen. Oxygen is responsible for the majority of the total quenching observed up to about 17% relative humidity. At 17% relative humidity, both effects contribute equally to the total quenching observed, and moisture (Le., disruption of the hydrogen bond matrix) becomes the most important quenching agent for humidities greater than 17%. Moisture must definitely be regarded as the most important quenching mechanism in RTP since it serves both to transport O2 into the sample matrix and to allow normal collisional deactivation to operate. A t relative humidities greater than about 20%, such internal conversion predominates even in pure O2 atmospheres. Fluid aqueous solutions of NaBPCA that are rigorously O2free show no The Journal of Physical Chemistry, Vol. 8 1, No. 20, 1977
E. M. Schulman and R. T. Parker
phosphorescence at all, as would be expected since samples on paper in water saturated argon show none above paper background emission. This result indicates that the high humidity quenching in Ar is not simply an artifact of traces of O2in the Ar used, which can also be seen by comparing the quenching by O2 in dry atmospheres containing 0,20, and 100% O2 (Figure 2). Support-Phosphor Interactions. The first studies of RTP reported that ionic organic compounds would phosphoresce when adsorbed on suitable supports such as paper, chopped cellulose, silica, or alumina and rigorously We noted that an abundance of surface hydroxyl groups seemed to be a general criterion for a suitable support for RTP, and our results confirmed that surfaces having many hydroxyl groups such as paper, sucrose, and starch provide excellent supports for observing RTP (Table 11). Attempts to observe RTP from supports containing few or no surface hydroxyl groups such as SWP polyethylene (a water wettable fiber) “paper”, glass fiber paper, or stainless steel treated with aqueous sodium carboxylate samples in both neutralized and alkaline media were not successful (Table 11) and removal of surface hydroxy functions from the best support, filter paper, by silanization reduced the observed intensity by more than 90% (Figure 8). Although hydroxylic surfaces seem to be a general requirement for suitable RTP supports, an anomalous case of phosphorescence from p-aminobenzoic acid adsorbed on sodium acetate precipitated from ethanol saturated in NaOH has been observed by Hurtubise13and confirmed by us. The ethanolic solvent medium is absolutely necessary in the sample penetration process. Attempts to observe this phenomenon using methanol, 2-propanol, 2-methyl-1-propanol, diethyl ether, acetone, or aqueous media yield little or no phosphorescence, and p-hydroxybenzoic acid, folic acid, and the benzamide of p-aminobenzoic acid are the only other compounds which yield appreciable phosphorescence by this method.13 We have observed similar anomalies with NaBPCA and NaNA in methanolic solvent systems on nonhydroxylic supports, and further studies of the “alcohol effect” on RTP support media are underway. These studies show the necessity for the organic compounds of interest to be polar or ionic in nature in order to exhibit RTP,7s8and other studies have verified this r e ~ u l t . ~The ~ ~combination -~~ of the necessity of a polar or ionic organic phosphor, with the general requirement of a support rich in hydroxyl groups, strongly supports the contention that hydrogen bonding of the phosphor molecules to the support is the primary mechanism of holding the phosphor molecules rigid enough to prevent collisional deactivation of the excited triplet state. The predominance of moisture quenching at high humidities agrees well with the concept of hydrogen bond formation preventing collisional deactivation, and with Lewis’ classic study17and the studies of tryptophan in PVA films,16vide supra. Additional evidence for a hydrogen bonding supportphosphor interaction can be seen with polar organic compounds which are prevented from hydrogen bonding to the support due to internal hydrogen bonding or chelation of the polar group with a metal ion, and with compounds which have steric restrictions to hydrogen bonding. We have found that, while sodium 2-naphthoxide phosphoresces brilliantly on paper at room temperature, the sodium salt of the chelating agent 1-nitroso-2-naphthol does not show RTP.22 Relative phosphorescence quantum yields of four ortho-substituted anilides, I(a-d) where a (R, = Rz = H), b (R, = H, R2 = CH3), c (R, = CH3, R2 = H), and d (R, = R2 = CH3), were measured in a variety
1939
Room Temperature Phosphorescence
@o
N/RI O A R ,
I
of hydroxylic solvents and isopentane at 77 K by Pfleni and Santus.23They found that, in general, higher quantum yields were obtained in the hydroxylic solvents. Compounds Ia and Ib, both of which are capable of internal hydrogen bonding, show a much smaller increase in quantum yield in going from isopentane to a hydroxylic solvent such as water-ethylene glycol than compounds IC and Id, which do not show internal hydrogen bonding. Sodium 2-biphenylcarboxylateexhibits much weaker RTP than sodium 4-biphenyl~arboxylate.~ Inspection of molecular models of sodium 2-biphenylcarboxylate reveaIs that the hydrogen bonding ability of this compound is sterically restricted by the ortho phenyl ring while the sodium 4-biphenylcarboxylate has no steric restrictions. Analytical Insights from this Study. Various analytical applications and advantages of RTP over low temperature phosphorimetry have been d e m ~ n s t r a t e d , ~ and - l ~ the qualitative similarity of 77 K and RTP spectra can be seen in Figure 11. RTP spectra are virtually identical when comparing hydroxylic supports of similar nature such as sucrose and paper (Figure 9,311 K), and slight increases in intensity and resolution (a resolution increase is seen for sucrose but not paper in going from 311 to 286 K, Figure 10) are observed with moderate lowering of the temperature (approximately a 15 % increase in emission intensity is noted for NaNA on sucrose in going from 311 to 286 K). For maximum quantitative reproducibility of the RTP method both moisture and oxygen should be excluded from the sample as evidenced by the very high rsd’s obtained for measurements in the presence of moisture and O2 (Table I). The increase in rsds for these samples is not solely an artifact of measuring a much weaker emission, since much better rsd’s were obtained for low concentration samples yielding similar emission intensities in dry Ar. Finally, samples can be prepared and desiccated in the instrument cell compartment in less than 15 min a t ambient temperatures (even faster with moderate warming of the cell) using a methanolic solvent system, thus, the sample desiccation period can be greatly reduced by our method as opposed to using aqueous solvent systems and dessicator or oven drying devices. Conclusion The results of this study indicate that hydrogen bonding of the phosphor to support is the most important interaction for holding the phosphor molecules rigid, thus
preventing collisional deactivation, and that the sample matrix intrinsically resists the penetration of oxygen, thus preventing triplet-triplet quenching of phosphor by 02. The presence of moisture serves to disrupt the hydrogen bonding network, “softening the matrix”, and allows both collisional deactivation of the excited phosphor and increased penetration of O2 into the sample matrix. The investigation of these effects has led to new and reproducible methods of sample handling and preparation for RTP.
Acknowledgment. Partial support of this research was provided by the Cottrell program of the Research Corporation and is gratefully acknowledged. We also thank P. F. Jones, P. G. Seybold, and J. D. Winefordner for stimulating conversations and suggestions; and the referees for helpful suggestions on the best presentation of our data.
References and Notes (1) Taken in part from the dissertation of R.T.P. lo be submitted to the Graduate School in partial fulfillment of the requirements for the Ph.D. Degree. Presented in part at the Symposium on Developments in Molecular Fluorescence and Phosphorescence Analysis, 173rd National Meeting of the American Chemical Society, New Orleans, La., March 21, 1977, ANAL 026. (2) R. B. Bonner, M. K. DeArmond, and G. H. Wahl, Jr., J . Am. Cbem. Soc.. 54. 988 (19721. (3) C. A.’Pa&er, “Pkotoluminescence of Solutions”, Elsevier, New York, N.Y., 1968, pp 45-46. (4) E. Vander Donckt, M. Matagne, and W. Sapir, Cbem. Pbys. Lett., 20, 81 (1973). (5) M. Zander, “F’hosphorlmetry”, Academic Press, New York, N.Y., 1968, p 117. (6) M. Roth, J. Chromatog., 30, 276 (1967). We are indebted to Dr. Roth for pointing this reference out to us. (7) E. M. Schulman and C. Walling, Science, 178, 53 (1972). (8) E. M. Schulman and C. Walling, J . Pbys. Cbem., 77, 902 (1973). (9) R. A. Paynter, S. L. Wellons, and J. D. Winefordner, Anal. Cbem., 48, 736 (1974). (10) S. L. Wellons, R. A. Paynter, and J. D. Winefordner, Spectrocbim. Acta, Part A , 30, 2133 (1974). (11) T. Vo-Dinh, E. L. Yen, and J. D. Winefordner, Anal. Cbem., 48, 1186 (1976). (12) P. G. Seybold and W. White, Anal. Cbem., 47, 1199 (1975). (13) R. M. A. Van Wandruszka and R. J. Hurtubise, Anal. Cbem., 48, 1784 (1976). (14) P. F. Jones, Polym. Lett., 6, 487 (1968). (15) P. F. Jones and S. Siege!, J. Cbem. Pbys., 50, 1134 (1969). (16) K. P. Ghiggino, C. H. Nicholls, and M. T. Pailthorpe, Pbotochem. Photobiol., 22, 169 (1975). (17) G. N. Lewis, D. Lipkin, and T. M. Magel, J . Am. Cbem. Soc.,63, 3005 (1941). (18) R. C. West, Ed., “The Handbook of Chemistry and Physics”, 52nd ed, The Chemical Rubber Co., Cleveland, Ohio, 1971, p E40. (19) The SWP polyethylene fiber (types E-400 and E-790) and a 50% SWP fiber-50% paper blend were gifts of the Crown Zellerbach Corp. (20) The specific counter example of p-aminobenzoic acid adsorbed on sodium acetate precipitated from ethanol is considered later. (21) E. G. Janzen, F. J. Johnson, and C. L. Ayers, J. Am. Cbem. Soc., 89, 1176 (1967). (22) E. M. Schulman and R. T. Parker, unpublished results. The analytical applications of this specificity is the subject of a forthcoming paper. (23) M. P. Pileni and R. Santus, J . Pbys. Cbem., 81, 755 (1977).
The Journal of Physical Chemistiy, Vol. 81, No. 20, 1977