Room temperature phosphorescence and delayed fluorescence of organic molecules trapped in silica sol—gel glasses

J. Photochem.

Photobiol.

A:

Chem.,

57

(1991)

41-63

41

Room temperature phosphorescence and delayed fluorescence of organic molecules trapped in silica sol-gel glasses David Lev~“-~ and David Avnirb “Institute

de Ciencia

bDeparrment The Hebrew

(Received

de Materiaies

de Madrid,

CSIC,

Serrano,

of Organic Chemistry and the F. Haber Research Universiry of Jerusalem, Jerusalem 91904 (Israel)

November

115

dpdo,

Centre

for

Madrid Molecular

28006

(Spain)

Dynamics,

6, 1990)

Abstract Room temperature phosphorescence (RTP) is obtained from a wide variety of organic polymerization process molecules when trapped in silica matrices prepared by the “sol-gel” of tetramethoxysilane. Examples include polycyclic hydrocarbons, aromatic acids, an aromatic base (quinine) and an organic dye (eosin-y). Lifetimes of up to several seconds are observed. Conditions for observing RTP vary. In some cases co-trapping of a heavy atom is needed, whereas in others RTP is observed even in wet gels without a heavy atom. A detailed study of the phenomenon was performed with 4-biphenylcarboxylic acid. It RTP and delayed fluorescence properties were studied as a function of base concentration, reaction time and temperature (Arrhenius analysis), revealing a multitude of emitting species (the acid and its anion, either adsorbed on the silica cage surface or “dissolved” in solvent-rich cages).

1. Introduction One of the major applications of the rapidly developing sol-gel process for low temperature glass preparation [l] is the ability to dope inorganic glasses with organic molecules, especially those which are photoactive [Z-7]. The room temperature polymerization of metal alkoxides has led to the preparation of materials of optical interest such as solid laser dye blocks [2, 4-61 and films [7], photochromic glasses [S, 91, hole-burning materials [3, 10, 111, light guides [12] and non-linear optical materials [13]. Recently, we have extended the application of this type of doped glass to the preparation of photocatalysts and have demonstrated their use in water reduction [14]. Not only have interesting photoactive materials been prepared in these studies, but doping with photoprobes has been intensively used for a detailed mechanistic study of the polymerization process of the metal alkoxide monomers involved [2-171. The method of doping sol-gel glasses with organic molecules has recently been applied for the preparation of chemical sensors [18, 191 and, for the first time, for the preparation of bioactive sol-glasses, by trapping of enzymes [18, 201. We have reported 1211 the observation that the efficient trapping of organic molecules in silica cages enables intense room temperature phosphorescence (RTP) to be observed. In this paper, we report a detailed study of the phenomenon, including

10X0-6030/91/$3.50

0 Elsevier Sequoia/Printed in The Netherlands

42

the properties of the final phosphorescent glasses and the use of phosphorescence to investigate the sol-gel process itself. RTP has been observed in a number of heterogeneous systems and in a number of homogeneous solid matrices. Skrilec and Cline Love [22] have demonstrate RTP in aqueous TI-Na lauryl sulphate micellar solutions; Alak and Vo-Dinh [23] have reported RTP from some polynuclear hydrocarbons adsorbed on cyclodextrin-treated filter paper in the presence of Tl-Pb acetate [23]; various amorphous matrices have been used such as glucose glasses [24], boric acid glass [25] and various plastics [26]. Of special relevance to our report are the intensive studies of RTP from adsorbed molecules [27-321. In these studies, RTP was observed from polar or ionic compounds [27-30, 33-351 adsorbed on thoroughly dried surfaces (of cellulose, silica and alumina) [33, 361, usually, but not always, in the presence of a heavy atom [37, 381. It was also found that, in order to observe RTP in these systems, the compounds should possess a reasonable phosphorescence quantum yield at low temperatures 139-421. As described in the following sections, the sol-gel inorganic Si02 matrix significantly relaxes the requirements for the observation of RTP: RTP is observed even from wet gels; in many instances a heavy atom is not needed; RTP is observed from a wide variety of molecules, both polar and non-polar; the exclusion of external quenchers is not necessary. After providing experimental detail (Section 2), we demonstrate the generality of the phenomenon of RTP from doped sol-gel SiOz glasses (Section 3.1). We then focus our attention on one molecule, 4-biphenylcarboxylic acid (BPH), for an in-depth study of a variety of properties and phenomena associated with RTP. This includes the following: (i) investigation of the effects of the added catalyst (NaOH) on the phosphorescence (P) and delayed fluorescence (DF) (Section 3.2) in both wet and dried gels; (ii) determination of the effect of prolonged drying of the gel (Section 3.2) and the testing of the isolation properties of the matrix; (iii) performance of an Arrhenius-type analysis of the two luminescence modes (P and DF, Section 3.3); (iv) investigation of the luminescence changes along the sol-gel-xerogel transitions (Section 3.4). In Section 4 we discuss the observations reported in Section 3, and provide a unified picture of the interplay between the properties of the silica cage under the various conditions studied and the luminescence performance of the trapped molecules.

2. Experimental

details

2.1. A typical procedure for the preparation of doped glasses Silica gel was prepared by hydrolysis of tetramethoxysilane (TMOS) in (tripledistilled) water-methanol solutions. The starting mixture was 5.00 ml TMOS, 2.40 ml HzO, 2.00 ml methanol and 4 ml of the dye solution (10e2 M) in methanol. For the study of NaOH effects, the water was replaced by 2.40 ml NaOH solutions of various concentrations (ranging from lop4 to 6 M). The samples gelled within several seconds (highest [NaOH]) or several hours (lowest [NaOH] and pure water). The gelation was carried out at room temperature in glass bottles covered with aluminium foil. After 3 days, evaporation of the solvents was made possible by punching small holes in the cover of the container. The samples were dried at 150 “C in vacuum (1 mmHg) for various periods of time (usually 18 or 48 h). Studies on the wet gels were carried out after they reached constant weight (500 h), For studies of the heavy atom effect, the water was replaced by 2.40 ml of a 1 M solution of NaBr or NaI.

43

2.2. Instrumenfation 2.2.1. Steady state luminescence measurements All measurements were carried out on a Perkin-Elmer luminescence spectrometer (model LS-5) in front-face geometry. The samples were placed in standard 1 mm x 10 mm quartz cells. The undoped glasses are not phosphorescent and no scattering problems were observed in the near-UV-visible range employed (see baselines in Fig. 1, Section 3.1). 2-2.2. Lifetime measurements Short RTP and DF lifetimes (milliseconds) were measured on a Perkin-Elmer LS-5 Auorometer. Long RTP lifetimes (seconds) were measured on an SLM spectrofluorometer. The long lifetime RTP decay curves obey simple single lifetime first-order kinetics. The decay curves of the samples exhibiting short RTP emissions (usually the wet gels) do not obey single lifetime first-order kinetics. For comparative purposes only, lifetimes are reported as the time required for the intensity to decay to 30% of its initial value.

3. Observations 3.1.

RTP of organic molecuies trapped in SiOz gel glasses: the phenomenon Trapping of a wide variety of organic molecules in SiOZ sol-gel glasses enables RTP to be observed from these trapped molecules. The generality of the phenomenon is demonstrated using the following molecules: phenanthrene, naphthalene, quinine, BPH, 1-naphthoic acid, eosin-y and pyrene. The RTP spectra of these organic molecules trapped in silica gel glasses are shown in Fig. 1 (pyrene spectrum is given in ref. 21). Most of the dyes emit either RTP or both DF and RTP under various gel preparation conditions. For example, under neutral conditions of gelation, phenanthrene, naphthalene and quinine show good RTP signals. Basic conditions of gelation are required to observe RTP emission for BPH, I-naphthoic acid and eosin-y, and the presence of a heavy atom perturber is necessary for the observation of phosphorescence for pyrene. In many studies [28, 29, 431, it has been observed that, in order to obtain RTP from adsorbed molecules, an ionic structure (or a high pH) is needed together with very dry conditions. In order to observe RTP from polycyclic hydrocarbons (PCH), it is necessary to add a heavy atom [4447]. Our study with three PCH molecules has revealed that RTP becomes more efficient with a decrease in the number of benzene rings, to the extent that, for naphthalene, RTP is observed even without a heavy atom of a heavy atom (NaBr) does not and in a neutral gel. In this case, the addition increase the RTP intensity. (An explanation for this trend can be offered by noting that, with increasing numbers of rings the T1 *S,, transition moves to longer wavelengths, i.e. the two states become closer and radiationless deactivation routes become more efficient [48].) Another remarkable observation is that naphthalene exhibits intense phosphorescence even in the wet gel (Fig. 2(A), (Table 1). By contrast, phenanthrene shows only weak RTP in the wet gels, which is significantly enhanced by addition of a heavy atom (Fig. 3(B)). RTP is observed for phenanthrene without a heavy atom after the gel is dried (Fig. 3(A), Table 1). A further increase in the size of the PCH molecule (pyrene) produces RTP that can only be observed in the dried gel and in the presence of a heavy atom (NaI). An explanation for this trend in the PCH series is given in Section 4.

44

As described in the following sections, most of the detailed analysis of the RTP phenomenon was carried out on 3PH. One of the glasses doped with this molecule (glass prepared with 0.5 M NaOH as catalyst and dried for 24 h at 150 “C and 1 mmHg) was used to demonstrate the special matrix isolation properties of this material

PHENANTHRENE w NAPHTHALENE

- _._ -._.I 400

- -.-GEL .- --GLASS - -._.__ I I I 450

500

/ /’ ___--__.___._____-.---------.-.--._GEL-GLASS --

_._ _.-._ I

550

600

I 400

1

61

350

1

1 450

I 500

--. I

550

6

30

X (nm)

350

400

450

500

550 X

Fig.

1.

600

650

700

7

(nml

(continued)

45

i5

__- - -- - --- __---_--___=---__

2

1

1

350

400

1

I

450

I

500

1

550

I

600

650

X (nml

Eosin-

y

1-L 1 \

c.7 E 3 a P .a G

\

\

\

: I’ / : : OF /-“\

5 z

/

/’ 500 Cdl

\

\

\

\

\

\

\

\

cy

\

\ \ \

/I \

I 550

\

/I \

1’

z

\ ‘1

/

-z

\

\

,’

I 600

I

I

1

650

700

750

x(nm)

Fig. 1. Emission spectra of various molecules trapped in silica sol-gel glass (P, room temperature phosphorescence; F, fluorescence; DF, delayed fluorescence; concentration of trapped molecule, approximately 10V3 mol 1-l glass). All excitations were in the range 290-320 nm, except for eosin which was excited at 483 nm. The glasses for phenanthrene, naphthalene and quinine were neutral. The glasses for naphthoic acid and eosin were prepared with 1 M NaOH, and that of BPH with 6 M NaOH.

(see also ref. 2). Exposure of the doped glass to either air or water (by immersion) caused an initial shortening of the phosphorescence lifetime by about 10% (from 1.10 to 1.00 s), and a reduction in intensity of about 20%. However, after this initial

decrease, both the lifetime and phosphorescence intensity remained unchanged for many weeks. The initial decrease is due to those trapped volecules which are very near to the pore surface or partially exposed to the pore volume. However, the bulk

0

350

400

450

500

X (nm) Fig. 2. Fluorescence (-) under neutral conditions see Table 1).

550

600

I

400

1

I

500

450 X

I

550

I

600

50

(nm)

and RTP (- - -) spectra of naphthalene trapped in gel glasses prepared (A) and with 1 M NaBr (B) (b.d., before drying; a.d., after drying;

Fig. 3. Fluorescence (-) and RTP (- --) spectra of phenanthrene trapped in gel glasses prepared under neutral conditions (A) and with 1 M NaBr (B) (b.d., before drying; a.d., after drying; see Table 1).

of the trapped molecules are embedded deep with the SiOz matrix, and are isolated not only from water but also from 02_ In this context, it should be mentioned here that most organic polymeric matrices used as hosts for photoactive molecules are permeable to O2 (49, 501.

47

TABLE

1

Relative emission intensities of trapped naphthalene Relative

(Naph)

and trapped phenanthrene

emission intensity (%)

I mar.

(A) (A) (B) (B)

Before drying After drying Before drying After drying

(A) Prepared

3.2.

(Phen)

I phorph.

Naph

Phen

Naph

Phen

100 59 41 18

23 14 100 3

80 100 6 100

8 19 16 100

under neutral conditions.

(B) With co-trapped

NaBr.

The effects of NaUH concentrations on the luminescence of BPff In this section we follow the changes in the phosphorescence and delayed fluorescence properties of BPH in the final xerogel, induced by changing the concentration of NaOH in the starting solution. It is well documented that the amount and nature of the catalyst have a great influence on the sol +gel transitions and on the surface properties of the porous gel formed [l, 51-531. As discussed below, we have found correlations between the features of the emission spectra (Amax, emission intensity and lifetime) and the surface or cage properties. We start with representative observations from three samples which were prepared with different NaOH concentrations (A, 10e4 M; B, 0.5 M; C, 6 M) according to the standard procedure given in Section 2. The samples were dried for 18 h in vacuum. The changes in the concentrations produce significant changes in the emission spectra of BPH. Figure 4 shows the emission spectra of the three samples. Case A (low [NaOH]) exhibits only DF, case B (medium [NaOH]) exhibits only RTP and case C (highest [NaOH]) exhibits both DF and RTP. Following these observations, the phenomenon was studied in detail. Figure 5 shows the effect of added NaOH on the fluorescence (F)(A), DF (B) and RTP (C) intensities in the dried gels (150 “C, 18 h, 1 mmHg vacuum). The RTP intensity passes through a maximum at a certain [NaOH] value which roughly coincides with a minimum in the DF. The F intensity decreases continuously to almost zero levels. These features are also apparent in the wet gels (Fig. 6), although RTP is weaker and DF is stronger (the intensity scales in Figs. 5 and 6 are the same). It should be noted that RTP is observed even from the wet gels. The changes in RTP intensity are accompanied by a similar trend in phosphorescence lifetimes. These pass through a maximum at about the same NaOH levels (Fig. 7 (dried gel), Fig. 8(B), wet gel). The very long lifetime of phosphorescence in the dry gels should be noted; it is of the order of seconds. In the wet gels it is of the order of milliseconds. Interestingly, the DF lifetime in the wet gels also passes through a maximum, in the millisecond range, but at a different [NaOH] value (Fig. 8(A)). The DF of the dried gel is more or less constant for all [NaOH] values (few milliseconds). values for the three types of emission as a function of The changes in the A,, [NaOH] are shown in Figs. 9 (dry) and 10 (wet). The trends in these changes, and their relationship to the trends observed for the intensities and lifetimes, are analysed in Section 4.

\

\

\ \

\

‘1

\

\

\

________---____c

L

.’

I

I

/ I

350

I

400

I

450

500

5x)

600

X(nm) Fig. 4. Phosphorescence and/or delayed fluorescence (-- -) and fluorescence (-) of BPH trapped in silica gel glasses prepared with different [NaOH] values: A, 10S4 M; B, 0.5 M; C, 6 M. h,, = 290 nm.

Finally, it should be noted that the phosphorescence intensities obtained for the dried gels above are not the maximum intensities. Heating beyond the standard 18 h has the effect shown in Table 2.

49

133

-

0

t

*

B 0.333.

0.167-

0 9.0

-

6.0

-

3.0

-

0.0

-3.00

-2.00

-1.00

t,=

t

0.00

f t.OO

log [NaOHl Fig. 5. Fluorescence (A), delayed fluorescence (B) and phosphorescence (C) intensities of BPH trapped in dried silica gels as a function of the NaOH concentration employed in the starting

solution. (Fluorescence intensity units are arbitrary.) 3.3.

Amhenius-rype

analysis of the phosphorescence

and delayed fluorescence

In order to understand some of the RTP characteristics of the BPH-doped gels, we carried out an Arrhenius-type analysis of the relationship between the emissions from the singlet (S,*) and triplet (T,*) states for a glass sample which emits both

50

A loo

50

0

a0

0.87

0.43

0.00

0.600

!

4.00

-2.75

4’55

-150

I

1.00

log WJOHI

Fig. 6. Fluorescence (A), delayed fluorescence (B) and phosphorescence (C) intensities of BPH trapped in wet silica gels as a function of the NaOH concentration employed in the starting solution. delayed

fluorescence

(5 M NaOH, The

and phosphorescence,

i.e. prepared

with high catalyst concentration

48 h, 150 “C).

relationship

between

the

emission

intensities

I is given

by IS41

51

log [NaOH] Fig. 7. Phosphorescence concentration the wet gels,

employed Fig. 8.)

lifetimes of trapped BPH in the starting solution.

in dried silica gels as a function of the NaOH (Decay to 30% is used for comparison with

in which QjDF and @+ are the quantum efficiencies of the DF and P respectively, A is a frequency factor, R is the Boltzmann constant and AE is the activation energy, which should be equal to the energy difference between S,* and Tr*. Equation (1) can be used to calculate AE if a plot of ln(1,-,,/1,) as a function of l/T is a straight line. As shown in Fig. 11, eqn. (1) is indeed obeyed and AE, calculated from the slope, is 1.85 kcal mol-‘, representing a low energy gap between the two states [54]. can also be calculated from the difference between the maxima AE(Tr* --, S,*) of the emission bands of DF and P. From this difference (Ah,, = 110 nm), AE is found to be 16.22 kcal mol-‘. The large discrepancy between these results is in our opinion, a clear indication that the DF and P are not correlated with each other. In other words, DF does not originate from the same molecules which are responsible for RTP. Under such conditions, eqn. (1) is obeyed if IDF and Ip obey separate Arrhenius plots

I oF =A> exp IP=A2

exp

- (AE,IRT) -(AEJRT)

(2) (3)

and the phosphorescence obeys DF obeys eqn. (2) with AEl 14.43 kcal mol-‘, eqn. (3) with AEz=2.70 kcal mol-‘, corresponding to the activation energies of the thermal deactivation processes_ Indeed AE =AEi -AEz = 1.73 kcal mol-’ is in close agreement with the value of 1.85 kcal mol-‘. 3.4.

DF and RTP changes during the sol+el-xerogel transition As mentioned in Section 1, photoprobes have been used successfully in an investigation of Si(OR), polymerization [16-201. We have observed that the long lifetime emissions (DF and RTP) are also sensitive probes of the structural changes which

52

0.00 i 4.00

r

-2.75

-1.50

-0.25

’ I

1.00

Fig. 8. Delayed fluorescence (A) and phosphorescence (B) lifetimes of trapped BPH in wet silica gels as a function of the NaOH concentration employed in the starting soIution.

occur during the sol-gel process. Neutral polymerization conditions were chosen because they provide a much slower reaction rate compared with basic conditions, thus enabling the gelation process to be followed. Under neutral conditions only DF is observed, but RTP can be induced by the addition of a heavy atom, NaBr (see Section 2). period (approximately 100 Figures 12(A) and 12(B) show that, after some “induction” h), there is an increase in the intensity of both the F and DF. However, they decrease as P (Fig. 12(C)) starts to appear. The two types of emission seem to be linked (as discussed below). However, Figs. 13(A) and 13(B) (the same system but without a heavy atom) show that an explanation of the maximum in the plots of intensity vs. reaction time may not be simple. These maxima also appear in the absence of RTP. Another indication that RTP and DF are linked comes from measurements of the lifetimes. Figure 14 shows that, in the absence of NaBr, the DF lifetime increases after an initial period (approximately 110 h). However, when NaBr is added to induce RTP (Fig. 15(A)), DF shows no increase, but the RTP lifetime increases (Fig. 15(B)). There is an important difference between the P lifetimes observed in these cases and those observed for the basic glasses. The lifetimes of DF and P are of the same order

53 375

1

362

-

A . . .

2 .

P 0

.

.

3-18f

335

1

I

4ocl

-_

---A

I

Kl

[NaOHl

delayed fluorescence (B) and phosphorescence (C) of trapped BPH in dried silica gels as a function of the NaOH concentration employed in the starting solution. Fig. 9. Changes in A,,

of the fluorescence (A),

of magnitude (milliseconds) for the neutral glasses, but they differ by a factor of lo3 for the basic glasses. This is referred to in Section 4. As can be seen in the analysis of intensity and lifetime vs. reaction time (Figs. 12-15), the 105-110 h point on the reaction time coordinate indicates a major structural

log [NaOH] Fig. 10. Changes in A,, of the fluorescence (A), delayed fluorescence (B) and phosphorescence (C) of trapped BPH in wet silica gels as a function of the NaOH concentration employed in the starting solution.

change in the polymerizing system. This can also be seen from the changes in A,, (Fig. 16). After 420 h reaction time, gelation was completed (3 days) and the final wet gels were then dried. A comparison of the spectral properties of the wet and dried gels,

5s TABLE

2

Changes in the emission intensities as a function of the drying time Absolute

Start of reaction End of gelation After 18 h drying After 24 h drying After 48 h drying

at h,,

of BPH

trapped

intensity

in silica gel glass ([NaOH]

= 0.5 M)

Amax (nm)

I Ruor.

ID-fluor.

I phosph.

F

D

P

4.01 5.79 20.53 22.08 34

0.010 0.018 -

0.026 0.009 1.16 9.38 16.92

358 358 350 353 353

= 363 = 354 -

490 = 485 485 485 483

Fig. 11. Arrhenius-type plot of the relative intensities of phosphorescence and delayed fluorescence as a function of temperature for trapped BPH in dried silica gel, prepared under basic conditions (5 M NaOH) (correlation coefficient, -0.986).

with and without NaBr, is given in Fig. 17. In the NaBr-doped glasses there is a substantial RTP even from the wet gels; in the dried glass, DF virtually disappears and the glass shows only RTP. The solution emission properties of BPH are given in Fig. 17(C) for comparison; F and DF are observed.

4. Discussion Most of the observations above can be rationalized. We first outline our conclusions and then apply them to the observations made. In general, the appearance of RTP, or its enhancement, is associated with changes in the properties of the (changing) sol-gel matrix and with cage properties. For BPH

IO.5

210

315

420

Time of Reaction(h)

Fig. 12. Changes in the fluorescence

(A), delayed fluorescence (B) and phosphorescence (C) intensities of trapped BPH during the sol-gel transition in polymerizing TMOS with co-trapped NaBr.

(and for many other molecules for which population of the triplet can be achieved), the following conditions are required: (i) conversion of the undissociated molecule to its anionic form, i.e. creation of basic conditions at the interface;

57

0.00

/. 0

I

105

210

I

315

I

420

Time of Reaction(II) Fig. 13. Changes in the fluorescence (A) and delayed fluorescence (B) intensities of trapped BPH without NaBr during the sol-gel transition in polymerizing TMOS. (ii) creation of interfacial properties which induce strong adsorption onto the cage surface walls; (iii) elimination of solvent molecules (water and methanol, in our case) from the immediate environment of the trapped molecule; (iv) increase in the rigidity of the trapping matrix. Depending on the degree to which these conditions are achieved in the gel preparation procedure, four possible species of BPH can exist each with its own characteristic luminescence properties. The first species, which gives rise to long-lived (seconds) high intensity RTP with little or no DF, is the strongly adsorbed (a), fully cage-entrapped anionic form of BPH (BP,-). The second, which gives rise to relatively short-lived (milliseconds) DF of moderate to weak intensity and relatively weak RTP with a similar lifetime, is BPd-, i.e. the anion which is dissolved (d) in the H,O-MeOH moIecules trapped with it in the same cage. The third species is the undissociated, dissolved BPH+ This species is the source of DF without P (Fig. 17(C)). Finally, there is the weakly absorbed, non-dissociated BPH, (under neutral or low basicity conditions of glass preparation}, which similarly gives rise mainly to DF.

58

0.00

0

105

210

315

I 420

Time of Reaction(h) Fig. 14. Delayed fluorescence

lifetime

of trapped

BPH during the polymerization process.

As can be seen in Figs. 5-8, phosphorescence reaches its maximum intensity and lifetime around 1 M NaOH. The optimal environmental properties in this case are the full dissociation of BPH to BP-, strong adsorption, small cages and few trapped solvent molecules in the dried gel. To the left and right of the maximum, the conditions for RTP worsen. To the right, i.e. at high [NaOH], the following changes occur: the silanols of the cage surface become fully ionized to Si-Ogroups which repel the negatively charged BP-, and the polymerization becomes very fast, resulting in larger cages and an increase in the amount of trapped solvent molecules (as is also observed from the DF properties, see below). The decrease in RTP intensity at the lower NaOH concentrations is associated with the decrease in the degree of BPH dissociation. Thus by sweeping the NaOH concentration from lop3 M to 6 M, the dominant BPH species passes from BPH, through BP,to BPd-, i.e. from non-phosphorescent species at the ends of this scale through phosphorescent BP,-. This interpretation is corroborated by the DF behaviour. Whereas the conditions for RTP are rather narrow, DF is observed for all BPH species outside this narrow range, namely from cage-dissolved BPH (Fig. 17(C) and low [NaOH] in Fig. 6(B)), in BP(high [NaOH] in Fig. 5(B)) and adsorbed BPH (low and medium [NaOH] Fig. 5(B)). These various DF species are also observed in the A,,, vs. [NaOH] behaviour (Figs. 9 and IO). The adsorbed species BPH, and BP,(blue-shifted A,,,) appear in the dry gels (Fig. 9(B)) in most of the [NaOH] range (except the very high [NaOH] vaiues), and in the wet gels (Fig. 10(B)) only at medium [NaOH] values, i.e, mostly BP,-. The dissolved species (red shifted by 40 nm compared with the adsorbed species) appear in the wet gels (Fig. 10(B)) at low (BPHd) and high (BPd-) [NaOH] values. A similar trend in A,,, is observed for the fluorescence and phosphorescence from the wet gels (Fig. 10(A)) and dry gels (Fig. 9(A)). The blue shifts observed at high [NaOH] values (Fig. 10) may reflect a shift from n,r* to r,# transitions. Such blue shifts are associated with the increase in the rigidity of the molecule on adsorption or cage trapping [SS], as recently observed by Zink, Dunn and coworkers [16]. A significant observation is the large decrease in fluorescence intensity as [NaOH] values at high [NaOH] in Figs. 9 and 10 are of increases (Figs. 5 and 6). (The A,, very low intensity.) Since this is accompanied by the appearance of phosphorescence

59

A

0.00

, 0

105

210

315

6

410

Time of Reaction(h) Fig. 1.5. Delayed fluorescence (A) and phosphorescence the polymerization process in the presence of NaBr.

at high [NaOH]

(B) lifetimes of trapped BPH during

and relatively strong DF at the highest [NaOH] value, we can conclude that the gradual transition from BPH to BP- is accompanied by an increase in the efficiency of the intersystem crossing between S1* and T,*. We must also comment on the striking resemblance between the dry gel curves and the wet gel curves (Figs. 5 and 6). We believe that, rather than reflecting a homogeneous situation in the wet gels, this similarity is indicative of the heterogeneous transition from the wet to the dry gels in the sense that the wet gel already contains cages which possess the properties of the final-dry cage. Indeed, the main difference between Figs. S(C) and 6(C) is the intensity, which is a factor of 100 weaker in the wet gel (the same scale is used). The Arrhenius analysis confirms the above observations. As mentioned in Section 3.3, a comparison of the results of this analysis with the spectral gap between h,,,(DF) and A,,(RTP) indicates that these two emissions originate from different species. Under the preparation conditions used, these species would be BPH, and BP,respectively. Furthermore, if DF and RTP originated from the same species, i.e. from

d

I&

2io

3 is

4io

Time of Reaction (h) Fig. 16. Changes in the maxima of fluorescence BPH during the polymerization process.

(A)

and delayed fluorescence

(B) of trapped

the same T, state, their lifetimes should be comparable. However, there is a factor of 1000 difference between the lifetimes, indicative of two emitting species. When the glasses are prepared under neutral conditions so that the only species available is BPH (and not BP-), RTP is induced only by a heavy atom. In this case both RTP and DF originate from the same species (BPH,) resulting, as expected, in similar lifetimes (millisecond scale, Fig. 15). We mentioned above that the reaction time of loo-120 indicates a unique change in the silica structure; Figs. 12 and 13 show that RTP starts to appear, the intensities of DF and F increase sharply, the lifetime of DF reaches a minimum (Figs. 14 and 15), the lifetime of RTP starts to increase and there is an abrupt change in the A,,, values of DF and F (Fig. 16). These observations can be interpreted as an indication that, around 110 h, the polymerizing material starts to behave as a strong adsorbent. It should be noted that, under neutral conditions, the hydrolysis and condensation-polymerization reactions operate much slower than under basic conditions [I]. The transition from BPHd to BPH, is apparently fast, comparable with a phase transition. The changes in Figs. 12-16 can be explained in a similar way. The increase in all

61

1 :

I I ;

d

; ,_

I

310

I

I

I

350

400

450

500

I

550

6

0

X (nm) Fig. 17. Fluorescence (-) and RTP (---) of BPH trapped in Si02 gel glasses prepared under neutral conditions (A) and when NaBr was present in the starting solution (B). The solution spectra of BPH (C) were recorded for comparison (1O-3 M, methanol).

three emission intensities (Figs. 12 and 13) reflects an increase in the rigidity imposed on the excited state BPH*; the decrease in the DF lifetime (Figs. 14 and 15(A)) reflects an increase in the efficiency of other decay routes of T1* (thermal deactivation in the case of Fig. 14 and deactivation plus RTP emission in Fig. 15). This reaches an extreme in the dried glass, where DF completely disappears (Fig. 17(B)).

Acknowledgments This research was supported by the Materials Division of the U.S. Army Research Development and Standardization Group (U.K.). D.L. thanks the MEC and the Institute de Ciencia de Materiales, CSIC, Madrid (Spain).

62 References 1 C. J. Brinker and G. W. Scherrer, Sol Gel Science, Academic Press, San Diego, 1990. L. L. Hench, Chem. Rev., 90 (1990) 33. S. Sakka and K. J. Kamiya, J. Non-C’qst. Solids, 42 (1980) 403. S. P. Mukhejee, J. Non-Cryst. Solids, 42 (1980) 477. C. J. Brinker and G. W. Scherrer, .7. Non-Ctysr. Solids, 70 (1985) 301. L. C. Klein, Annu. Rev. Mater. Sci., I5 (1985) 227. M. Yamane, S. Aso and T. Sakaino, J. Mater. Sci., Z3 (1978) 865. 2 D. Avnir, D. Levy and R. Reisfeld, J. Whys. Chem., 88 (1984) 5956. 3 T. Tani, H. Namikawa, K. Arai and A. Makishima, J. Appl. Phys., 58 (1985) 3559. 4 F. Salin, G. Le Saux, P. Georges, A. Brun, C. Bangall and J. Zarcycki, Opt. Lett., 14 (1989) 785. 5 R. Reisfeid, D. Brusilovsky, M. Eyal, E. Miron, Z. Burstein and J. It-vi, Chem. Phys. Lett., 160 (1989) 43, 6 E. T. Knobbe, B. Dunn, P. D. Fukua, F. Nishida and J. I. Zink, Proc. 4th Znt. Conf: on UItrastructures of Ceramics, Glasses and Composites, Wiley, New York, NY, 1989. 7 Y. Kobayashi, Y. Kurokawa and Y. Imai, .7. Non-Cyst. Solids, IO5 (1988) 198. 8 D. Levy and D. Avnir, J. Phys. Chem., 92 (1988) 4734. D. Levy, S. Einhorn and D. Avnir, J. Non-Clyst. Solids, 113 (1989) 137, and references cited therein. 9 K. Matsui, T. Morohoshi and S. Yoshida, Proc. ht. Meeting Adv. Materials, Tokyo I988. D. Preston ef al., J. Phys. Chem., 94 (1990) 4167. 10 A. Makishima and T. Tani, .I. Am. Ceram. Sot., 69 (1986) C-72. 11 H. Tanaka, J. Takahashi, J. Tsuchiya, Y. Kobayashi and Y. Kurokawa, J. Non-Cryst. Solids, IO9 (1989) 164. R. Lecher, A. Renn and W. P. Wild, Chem. Phys. Lett., 138 (1987) 405. Y. Kobayashi, Y. Imai and Y. Kurokawa, J. Muter. Sci. Left., 7 (1988) 1148. 12 D. Avnir, V. Kaufman and R. Reisfeld, J. Non-Cryst. Solids, 74 (1985) 395. 13 D. R. Ulrich, J. Non-Crysf. Solids, ZOO (1988) 174. 14 A. Slama-Schwok, D. Avnir and M. Ottolenghi, J. Phys. Chem., 93 (1989) 7544. 15 Y. Kobayashi. Y. Kurokawa, Y. Imai and S. Muto, .J. Non-Cyst. Solids, 205 (1988) 198. 16 J. Mckiernan, J. C. Pouxviel, B. Dunn and J. I. Zink, J. Phys. Chem., 93 (1989) 2129. J. C. Pouxviel, B. Dunn and J. I. Zink, J. Phys. Chem., 93 (1989) 2134. 17 D. Levy, Ph.D. Thesis, The Hebrew University of Jerusalem, 1989. 18 D. Avnir, M. Ottolenghi and R. Zusman, patent appIication pending, 1990. 19 R. Zusman, C. Rottman and M. Ottolenghi, J. Non-Cryst. Solids, 122 (1990) 107. 20 S. Braun, S. Rappoport, R. Zusman, D. Avnir and M. Ottolengthi, Muter. Lett., IO (1990) 1. 21 V. R. Kaufman, D. Levy and D. Avnir, J. Non-Cryst. Solids, 82 (1986) 103. 22 M. Skrilec and L. J. Cline Love, Anal. Chem., 52 (1982) 1559. S. Hamay. J. Am. Chem. Sot., ZZZ (1989) 3954. 23 A. M. Aiak and T. Vo-Dinh, Anal. Chem., 60 (1988) 596. 24 G. Oster, J. Joussot-Dubien and B. Broyde, J. Am. Chem. Sot., 81 (1959) 1869. 25 H. Buck, E. van der Lugt and L. Oosterhoff, Tetrahedron Suppl., 2 (1963) 173. 26 G. Oster, N. Geacintov and A. Khan, Nature, 196 (1962) 1089. 27 M. Roth, J. Chromatogr., 30 (1967) 276. 28 E. M. Schulman and C. Walling, Science, 53 (1972) 178. 29 E. M. Schulman and C. Walling, J. Phys. Chem., 77 (1973) 902. 30 R. J. Hurtubise, Solid Swface Luminescence Analysis, Marcel Dekker, New York,, 1981. 31 R. J. Hurtubise, SoIid Surface Luminescence Analysis, Marcel Dekker, New York, 1981, pp. 93-l 14. 32 R. T. Parker, R. S. Freedlander and R. B. Dunlap, Anal. Chem., 120 (1980) 1. 33 E. M. Schulman and R. T. Parker, J. Phys. Chem., 81 (1977) 1932. 34 R. A_ Paynter, S. L. Wellons and J. D. Winefordner, Anal. Chem.. 46 (1974) 736.

63 35 36 37 38 39 40 41 42 43 44 45 46

47 48

49 50 51 52 53

54 55

R. R. S. K. J. S. C. J. H. E. T. T. E. E.

M. A. von Wandruszka and R. J. Hurtubise, Anal. Chem., 49 (1977) 2164. J. Hurtubise, Mol. Cumin. Spectrosc., 2 (1988) 1. J. Hilsenbeck and C. J. Haustein, Iowa Acad. Sci, 95 (1988) 38. F. Harbaugh, C. M. O’Donnell and J. D. Winefordner, Anal. Chem., 45 (1973) 381. B. F. Lloyd, Analyst, 103 (1978) 775. K. Lower and M. A. El-Sayed, Chem. Rev., 66 (1966) 199. A. Parker and C. G. Hatchard, Analyst, 87 (1963) 664. J. Aaron and J. D. Winefordner, Tuiantu, 22 (1975) 707. C. Hollifreld and J. D. Winefordner, Anal. Chem., 40 (1978) 1759. L. Y. Bower and J. D. Winefordner, Anal. Chim. Acta, 101 (1978) 319. vo-Dinh, E. Lue Yen and J. D. Winefordner, Talanta, 24 (1977) 146. vo-Dinh, G, L. Walden and J. D. Winefordner, Anal. Chem., 49 (1977) 1126. L. Y. Bower and J. D. Winefordner, Anal. Chim. Acata, 202 (1978) 1. L. Y. Bower and J. D. Winefordner, Appl. Spectrosc., 33 (1979) 9. G. W. Robinson and R. P. Frosh, J. Chem. Phys., 38 (1963) 1187. P. F. Jones, Polym. Lett., 6 (1968) 487. P. F. Jones and S. Siegel, J. Chem. Phys., 50 (1969) 1134. B. E. Yoldas, .J. Non-Cry&. Solids, 63 (1984) 145. M. Yamane, S. Inoue and A. Yasumori, J. Non-Cryst. Solids, 63 (1984) 13. C. J. Brinker, K. D. Keefer, D. W. Schaefer and C. S. Ashley, J. Non-Cryst. Solids, 48 (1982) 47. M. Nogami and Y. Moriya, J. Non-Cyst. Solids, 37 (1980) 191. J. Zarzycki, M. Prassas and J. Phalippou, J. Mater. Sci., 17 (1982) 3371. C. A, Parker and C. G. Hatchard, Trans. Faraday Sot., 57 (1961) 1894. C. G. Guilbault, Practical FIuorescence, Marcel Dekker, New York, 1973, Chapter 3, p. 79. W. L. Bhaumik and R. Hardwick, .I. Chem. Phys., 39 (1963) 1595. R. A. Passwater, Fluorescence News, 5 (1971) 4.