University of Southern California Dissertations and Theses

A photophysical study of pyrene adsorbed onto silicas of variable surface area and porosity

(USC Thesis Other)

A photophysical study of pyrene adsorbed onto silicas of variable surface area and porosity

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content A PHOTOPHYSICAL STUDY O F PYREN E A D SO R B ED O N TO SILICAS O F VARIABLE SURFACE A R EA A N D POROSITY by Herman T. Yee A Dissertation Presented to the FACULTY O F THE G R A D U A TE S C H O O L UNIVERSITY O F SO U TH ER N CALIFORNIA In Partial Fulfillment of the Requirements for the Degree D O C TO R O F PHILOSOPHY (Chemistry) August 1988 UMI Number: DP21981 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. Dissertation Publishing UMI DP21981 Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code ProQuest ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES, CALIFORNIA 90089 c % This dissertation, written by Herman T. Yee under the direction of h....i.s..... Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School, in partial fulfillment of re quirements for the degree of DO CTO R OF PHILO SO PHY Dean of Graduate Studies June 29, 1988 DISSERTATION COMMITTEE Chairperson To M y Parents To Sharon, Dale and John To Ann Elizabeth and To Ralph A C K N O W LE D G EM EN TS I would like to thank the Chemistry Department for being very supportive of m e during m y training. There are som e individuals who are dear to m e that I wish to convey m y feelings of gratitude. Michele Dea is the hardest working individual that I know and without her help and friendship in a ll the years I have been here, I s t il l might be here! Thanks Michele; you're the BEST! Dr. Lawrence Singer is the most guiding advisor anyone could ask for in a director. Without his encouragement, advice and his willingness to allow his students to be fa ir ly autonomous, it would have been d iffic u lt to develop essential research s k ills . Thanks Dr. Singer for keeping an open ear and mind to m e. Jack Tseng cam e to the Singer group in the pre-computer year of 1984 and saw m y struggles with a handheld programmable calculator. He provided our lab with the fir s t true "PC" (even if it is from Taiwan!). Without this addition, m y calculations and graphing would s t i l l be done by hand. Thanks Jack and long liv e Lotus! Ted Yamada has always been there when anyone needed his help or company. W e are old comrades and w ill be until w e die or until one of us gets m ad at the other! Thanks Ted for being a friend and a lab partner. TABLE O F CONTENTS Page Dedication ii Acknowledgements i i i List of Tables vii List of Figures v iii Abstract xiv Chapter 1. Introduction 1 1.1 Introduction 1 1.2 Surface Area and Porosity Introduction 2 1.2.1 Other Methods of Surface Characterizations 3 1.3 Spectroscopy of Surfaces 4 1.3.1 Fluorescence Spectroscopy Introduction 4 1.3.2 Pyrene Solution Behavior 5 1.3.3 Pyrene Monomer-Excimer Emission Analysis 8 1.4 Introduction to a Family of Silicas 17 1.5 Modification of Silica and Adsorbed Pyrene 22 1.6 References 24 Chapter 2. Surface Area Charaterization of Native Silicas and Silylated Silicas 29 2.1 Introduction 29 2.2 BET Method of Surface Area Determinations 29 2.3 Results and Discussion 35 2.3.1 Native Silica Results 35 2.3.2 Silylated Silica Results 45 2.3.3 Discussion of Results 45 2.4 Experimental 55 2.4.1 Materials 55 2.4.2 Equipment 56 2.4.3 Methods 56 2.4.3.1 Silylation Procedure 56 2.4.3.2 BET Measurments 57 2.5 References 59 Chapter 3. Porosity Distribution of Native Silicas.and Silylated Silicas 60 3.1 Introduction 60 3.2 "Modelless" Method 62 3.3 Results and Discussion 65 3.3.1 Native Silica Results 65 3.3.2 Silvlated Silica Results 3.3.3 Discussion Page 3.4 Experimental 79 3.4.1 Materials 79 3.4.2 Equipment 79 3.4.3 Methods 79 3.5 References 80 Chapter 4. Photophysical Study of Pyrene Adsorbed onto Native Si 1ica 82 4.1 Introduction 82 4.2 Results of Pyrene o ;n Native Silica 83 4.2.1.0 Variable Temperature Time Resolved Spectra of Pyrene in Hexane 84 4.2.1.1 Emission Lifetime of Pyrene in Hexane 89 4.2.1.2 Vibrational Analysis of the Monomer Emission (Pyrene in Hexane) 89 4.2.1.3 Pyrene Mobility in Hexane 93 4.2.1.4 Activation Energy of Pyrene Excimer Formation in Hexane 95 4.2.2.0 Variable Temperature Time Resolved Spectra of Pyrene on S24 97 4.2.2.1 Decay Measurments of Pyrene Emission on S24 116 4.2.2.2 Vibrational Analysis of the Monomer Emission (Pyrene on S24) 127 4.2.2.3 Pyrene Mobility on S24 129 4.2.2.4 Activation Energy of Pyrene Excimer Formation on S24 130 4.2.3.0 Variable Temperature Time Resolved Spectra of Pyrene on S72 133 4.2.3.1 Decay Measurments of Pyrene Emission on S72 147 4.2.3.2 Vibrational Analysis of the Monomer Emission and Pyrene Mobility (Pyrene on S 72) 151 4.2.3.3 Activation Energy of Pyrene Excimer Formation on S72 157 4.2.4.0 Variable Temperature Time Resolved Spectra of Pyrene on S63 160 4.2.4.1 Decay Measurments of Pyrene Emission on S63 179 4.2.4.2 Vibrational Analysis of the Monomer Emission and Pyrene Mobility (Pyrene on S63) 189 4.2.4.3 Activation Energy of Pyrene Excimer Formation on S63 189 v Page 4.3 Discussion 191 4.4 Experimental 205 4.4.1 Materials 205 4.4.2 Equipment 205 4.4.3 Methods 206 4.5 References 208 Chapter 5. Photophysical Study of Pyrene Adsorbed onto Si lylated Si 1 ica 211 5.1 Introduction 211 5.2 Results of Silylated Silica 212 5.2.1.0 Variable Temperature Time Resolved Spectra of Pyrene on SS24 212 5.2.1.1 Decay Measurments of Pyrene Emission and Vibrational Analysis of the Monomer - Emission (Pyrene on SS24) 223 5.2.1.2 Pyrene Mobility and Activation Energy of Pyrene Excimer Formation on SS24 229 5.2.2.0 Variable Temperature Time Resol ved Spectra of Pyrene on SS72 233 5.2.2.1 Decay Measurements of Pyrene Emission and Vibrational Analysis of the Monomer Emission (Pyrene on SS72) 246 5.2.2.2 Pyrene Mobility and Activation Energy of Pyrene Excimer Formation on SS72 253 5.2.3.0 Variable Temperature Time Resolved Spectra of Pyrene on SS63 257 5.2.3.1 Decay Measurments of Pyrene Emission and Vibrational Analysis of the Monomer Emission (Pyrene on SS63) 266 5.2.3.2 Pyrene Mobility and Activation Energy of Pyrene Excimer Formation on SS63 273 5.3 Discussion 276 5.4 Experimental 281 5.4.1 Materials 281 5.4.2 Equipment 281 5.4.3 Methods 281 5.5 References 282 Selected Bibliography 284 vi L is t of Tables Table 2.1 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 5.1 5.2 5.3 5.4 5.5 5.6 Page Summary of Surface Areas and Porosity Data for Native and Silylated Silicas 36 Pyrene Monomer Emission: Vibrational Data at Various Temperatures ( 6.5 X 10_3M Hexane Solution) 90 Normal ized Pyrene Band Intens.ities in Vari ous Solvents 94 Pyrene Monomer Lifetimes: Various Concentrations at Various Temperatures on S24 117 Pyrene Monomer Emission: Vibrational Data at Various Temperatures (S24) 128 Pyrene Monomer Lifetimes: Various Concentrations at Various Temperatures on S72 148 Pyrene Monomer Emission: Vibrational Data at Various Temperatures (S72) 152 Pyrene Monomer Lifetimes: Various Concentrations at Various Temperatures on S63 180 Pyrene Monomer Emission: Vibrational Data at Various Temperatures (S63) 190 Pyrene Monomer Lifetimes: Various Concentrations at Various Temperatures on SS24 224 Pyrene Monomer Emission: Vibrational Data at Various Temperatures (SS24) 230 Pyrene Monomer Lifetimes: Various Concentrations at Various Temperatures on SS72 247 Pyrene Monomer Emission: Vibrational Data at Various Temperatures (SS72) 248 Pyrene Monomer Lifetimes: Various Concentrations at Various Temperatures on SS63 267 Pyrene Monomer Emission: Vibrational Data at Various Temperatures (SS63) 268 vn 6 7 IQ 14 16 18 20 32 37 38 39 40 41 42 43 44 46 47 48 49 L is t of Figures Potential Energy Diagram of Pyrene Monomer and Excimer Various Concentrations of Pyrene versus Excimer Development Pyrene Absorption Spectrum (Hexane Solution) Pyrene Emission in Solvents of Variable Polarities Scheme for Pyrene Monomer-Excimer Formation and Decay Representative S ilica Surface Different Types Silanols Brunauer, Deming, Deming, T e lle r Isotherm Classifications Adsorption Isotherm for S24 Adsorption Isotherm for S24 (unheated) Adsorption Isotherm for S72 Adsorption Isotherm for S63 BET Isotherm for S24 BET Isotherm for S24 (unheated) BET Isotherm for S72 BET Isotherm for S63 Adsorption Isotherm for SS24 Adsorption Isotherm for SS24 (unheated) Adsorption Isotherm for SS72 Adsorption Isotherm for SS63 Figure Page 2.11 BET Isotherm for SS24 50 2.11a BET Isotherm for SS24 (unheated) 51 2.12 BET Isotherm for SS72 52 2.13 BET Isotherm for SS63 53 3.1 Porosity Distribution for S24 66 3.1a Porosity Distribution for S24 (unheated) 67 3.2 Porosity Distribution for S72 69 3.3 Porosity Distribution for S63 71 3.4 Porosity Distribution for SS24 72 3.4a Porosity Distribution for SS24 (unheated) 73 3.5 Porosity Distribution for SS72 75 3.6 Porosity Distribution for SS63 76 4.1 Variable Temperature Spectra of 6.5 X 10“3M Pyrene in Hexane at t=0 86 4.1a Variable Temperature Spectra of 6.5 X 10“3M Pyrene in Hexane at t=150 88 4.2 Variable Temperature Pyrene Monomer Emission Decay (6.5 X 10“3M in hexane) 91 4.3 Variable Temperature Pyrene Excimer Emission Decay (6.5 X 10~3M in hexane) 92 4.4 Arrhenius Plot for 6.5 X 10’ 3M Pyrene in Hexane 96 4.5 Variable Temperature Spectra of 8.91 umole/gm Pyrene on S24 at t=0 99 4.5a Variable Temperature Spectra of 8.91 umole/gm Pyrene on S24 at t=150 101 4.6 Variable Temperature Spectra of 14.5 umole/gm Pyrene on S24 at t=0 103 4.6a Variable Temperature Spectra of 14.5 umole/gm Pyrene on S24 at t=150 4.7 Variable Temperature Spectra of 36.5 umole/gm Pyrene on S24 at t=0 4.7a Variable Temperature Spectra of 36.5 umole/gm Pyrene on S24 at t=150 4.8 Variable Temperature Spectra of 91.7 umole/gm Pyrene on S24 at t=0 4.8a Variable Temperature Spectra of 91.7 umole/gm Pyrene on S24 at t=150 4.9 Monomer Decay for 8.91 umole/gm Pyrene on S24 4.10 Monomer Decay for 14.5 umole/gm Pyrene on S24 4.10a Excimer Decay for 14.5 umole/gm Pyrene on S24 4.11 Monomer Decay for 36.5 umole/gm Pyrene on S24 4.11a Excimer Decay for 36.5 umole/gm Pyrene on S24 4.12 Monomer Decay for 91.7 umole/gm Pyrene on S24 4.12a Excimer Decay for 91.7 umole/gm Pyrene on S24 4.13 Arrhenius Plot for 14.5 umole/gm Pyrene on S24 4.14 Arrhenius Plot for 36.5 umole/gm Pyrene on S24 4.15 Variable Temperature Spectra of 10 umole/gm Pyrene on S72 at t=0 4.15a Variable Temperature Spectra of 10 umole/gm Pyrene on S72 at t=150 4.16 Variable Temperature Spectra of 20 umole/gm Pyrene on S72 at t=0 4.16a Variable Temperature Spectra of 20 umole/gm Pyrene on S72 at t=150 4.17 Variable Temperature Spectra of 60 umole/gm Pyrene on S72 at t=0 4.17a Variable Temperature Spectra of 60 umole/gm Pyrene on S72 at t=150 4.18 Monomer Decay for 10 umole/gm Pyrene on S72 4.18a Excimer Decay for 10 umole/gm Pyrene on S72 4.19 Monomer Decay for 20 umole/gm Pyrene on S72 4.19a Excimer Decay for 20 umole/gm Pyrene on S72 4.20 Monomer Decay for 60 umole/gm Pyrene on S72 4.20a Excimer Decay for 60 umole/gm Pyrene on S72 4.21 Arrhenius Plot for 10 umole/gm Pyrene on S72 4.22 Arrhenius Plot for 20 umole/gm Pyrene on S72 4.23 Variable Temperature Spectra of 6 umole/gm Pyrene on S63 at t=0 4.23a Variable Temperature Spectra of 6 umole/gm Pyrene on S63 at t=150 4.24 Variable Temperature Spectra of 11.6 umole/gm Pyrene on S63 at t=0 4.24a Variable Temperature Spectra of 11.6 umole/gm Pyrene on S63 at t=150 4.25 Variable Temperature Spectra of 49.8 umole/gm Pyrene on S63 at t=0 4.25a Variable Temperature Spectra of 49.8 umole/gm Pyrene on S63 at t=150 4.26 Variable Temperature Spectra of 60 umole/gm Pyrene on S63 at t=0 4.26a Variable Temperature Spectra of 60 umole/gm Pyrene on S63 at t=150 4.27 Monomer Decay for 6 umole/gm Pyrene on S63 4.27a Excimer Decay for 6 umole/gm Pyrene on S63 4.28 Monomer Decay for 11.6 umole/gm Pyrene on S63 Figure Page 4.28a Excimer Decay for 11.6 umole/gm Pyrene on S63 186 4.29 Monomer Decay for 60 umole/gm Pyrene on S63 187 4.29a Excimer Decay for 60 umole/gm Pyrene on S63 188 4.30 Arrhenius Plot for 6 umole/gm Pyrene on S63 192 4.31 Arrhenius Plot for 11.6 umole/gm Pyrene on S63 193 4.32 Arrhenius Plot for 60 umole/gm on S63 194 5.1 Variable Temperature Spectra of 1.78 umole/gm Pyrene on SS24 at t=0 214 5.1a Variable Temperature Spectra of 1.78 umole/gm Pyrene on SS24 at t=150 216 5.2 Variable Temperature Spectra of 13.8 umole/gm Pyrene on SS24 at t=0 218 5.2a Variable Temperature Spectra of 13.8 umole/gm Pyrene on SS24 at t=150 220 5.3 Monomer Decay for 1.78 umole/gm Pyrene on SS24 225 5.3a Excimer Decay for 1.78 umole/gm Pyrene on SS24 226 5.4 Monomer Decay for 13.8 umole/gm Pyrene on SS24 227 5.4a Excimer Decay for 13.8 umole/gm Pyrene on SS24 228 5.5 Arrhenius Plot for 1.78 umole/gm Pyrene on SS24 231 5.6 Arrhenius Plot for 13.8 umole/gm Pyrene on SS24 232 5.7 Variable Temperature Spectra of 1.75 umole/gm Pyrene on SS72 at t=0 235 5.7a Variable Temperature Spectra of 1.75 umole/gm Pyrene on SS72 at t=150 237 5.8 Variable Temperature Spectra of 26.8 umole/gm Pyrene on SS72 at t=0 239 5.8a Variable Temperature Spectra of 26.8 umole/gm Pyrene on SS72 at t=150 241 x ii Figure Page 5.9 Variable Temperature Spectra of 28.8 umole/gm Pyrene on SS72 at t=0 243 5.9a Variable Temperature Spectra of 28.8 umole/gm Pyrene on SS72 at t=150 245 5.10 Monomer Decay for 1.75 umole/gm Pyrene on SS72 249 5.10a Excimer Decay for 1.75 umole/gm Pyrene on SS72 250 5.11 Monomer Decay for 26.8 umole/gm Pyrene on SS72 251 5.11a Excimer Decay for 26.8 umole/gm Pyrene on SS72 252 5.12 Arrhenius Plot for 1.75 umole/gm Pyrene on SS72 254 5.13 Arrhenius Plot for 26.8 umole/gm Pyrene on SS72 256 5.14 Variable Temperature Spectra of 2.7 umole/gm Pyrene on SS63 at t=0 259 5.14a Variable Temperature Spectra of 2.7 umole/gm Pyrene on SS63 at t=150 261 5.15 Variable Temperature Spectra of 14.6 umole/gm Pyrene on SS63 at t=0 263 5.15a Variable Temperature Spectra of 14.6 umole/gm Pyrene on SS63 at t=150 265 5.16 Monomer Decay for 2.7 umole/gm Pyrene on SS63 269 5.16a Excimer Decay for 2.7 umole/gm Pyrene on SS63 270 5.17 Monomer Decay for 14.6 umole/gm Pyrene on SS63 271 5.17a Excimer Decay for 14.6 umole/gm Pyrene on SS63 272 5.18 Arrhenius Plot for 2.7 umole/gm Pyrene on SS63 274 5.19 Arrhenius Plot for 14.6 umole/gm Pyrene on SS63 275 x i i i ABSTRACT The photophysical characterization of solid surfaces using adsorbed hydrocarbons, along with conventional surface analysis, is providing a greater understanding of important surface processes. For example, in catalysis, a knowledge of the adsorption sites and surface reactivities is important for the development of new and novel synthetic methods. To this end, three native s ilic a surfaces and their silylated analogs were investigated by conventional surface area and porosity determinations and by the pyrene monomer-excimer photophysical probe. Variable temperature studies of the adsorbed pyrene, probe mobility and activation energies for excimer formation were investigated. The native silicas were distinct; ranging from a s ilic a with high surface area and pore volume located predominantly in small pores to a s ilic a with lower surface area and pore volume more evenly distributed. The effect of sily latio n was to decrease the surface area on a ll silicas except for one s ilic a. Pore distributions show a shift of porosity towards smaller sized pores. Variable temperature work with the pyrene revealed a unique pattern of photophysical behavior on each of the silicas. In general, the pyrene adsorbed onto native silicas showed an increase in excimer emission intensity with decreasing temperature to -100°C which is contrary to solution behavior. Analysis of the monomer-excimer decays x iv shows evidence for at least two kinds of excimer; diffusively formed and static dimer. Arrhenius plots show an activation energy for excimer formation less than the previously reported 4 kcal/mole. Pyrene adsorbed onto the silylated silicas showed solution-like behavior and aggregation of the probe as opposed to dispersion of the probe as ea rlier reported. An increase in activation energy is associated with the onset of an additional segmental rotational barrier due to the bulky trimethyl s ily l groups. xv CHAPTER 1 INTRODUCTION 1.1 Introduction The photophysics and photochemistry of adsorbed hydrocarbons on solid surfaces is rapidly becoming a useful method of analysing surface characteristics (1-15). Although greater understanding of these surfaces is leading to new and novel surface synthetic reactions (16-19), m uch of the research on the microscopic and molecular level is s t i l l primitive and not well defined as compared with similar solution studies. Diffusional processes, molecular geometries and specific surface characteristics lead to important differences in comparison with solution phase studies such as in micelles (14,20,21). In general, surfaces are important for processes such as physical adsorption, chemisorption, catalysis and molecular diffusion. A n understanding of a particular surface can potentially give unique solutions in controlling chemical reactivities with respect to surface polarities, molecular aggregations and surface energies (22-24). The practical importance of surfaces is most readily seen in chromatography in which various surfaces w ill preferentially adsorb som e reaction products to allow for their purification. Recently, s ilic a gel has been used as a unique environment for synthetic reactions (16-19). 1 1.2 Surface Area and Porosity Introduction H istorically, surfaces have been characterized by adsorption, chemical and spectroscopic techniques (25). N o one technique has had an advantage over the other; rather these techniques are complementary and supplementary to each other. The predominant method of surface characterizations has been the determination of surface area and pore distributions. Surface area measurements are usually done employing the Brunauer-Emmett-Tel ler (BET) method although other methods are available but not widely used (26,48,51). The BET method has been recognized as the standard method for surface area determination. Pore distibution determinations, in general, are varied and there is no wide agreement with regards to a particular method. The problem in establishing a method for porosity analysis has been that surfaces, in general, do not have regular pore shapes. There are som e surfaces in which the pores have been m ade to be regular but most surfaces have a wide range of pore shapes and sizes. Given the v a ria b ility of surface pores, the theoretical methods are equally variable. Classical methods u tiliz e a pore model at the outset of a determination (27,28) while recent methods only apply a model after som e in itia l analysis (29,30). The fir s t known "model less" method was proposed by Brunauer, Mikhail and Bodor (31). This method m ade use of the Kelvin equation and related a quantity called the hydraulic radius to the volume and surface area of a pore. Application of a pore model would ultimately be related to the hydraulic radius and the size of a pore as well as the pore 2 distribution could be determined. A more concise explanation of this method follows. A more direct method of obtaining porosity data is to use mercury porosimetry. The method involves intruding the solid with mercury and the volume of mercury taken up by the solid is then related to a pore distribution. Mercury porosimetry does assume a pore model a priori and the other methods, including those which also assume a pore model a p rio ri, use data easily obtained during a BET determination and the pore characteristics from these methods do not significantly d iffer from the analysis done with the mercury method. A third method of surface characterization by adsorption techniques is the generation of adsorption/desorption isotherms. Various molecules are adsorbed and desorbed on the surfaces to generate characteristic isotherms. The isotherms can then be classified based on the Brunauer, Deming, Deming, T e lle r classification scheme in which five types of isotherms have been distinguished (32) (Figure 2.0). 1.2.1 Other Methods of Surface Characterization Chemical methods make up another large category of surface investigations (25). Most of these methods have centered mainly on the determination of active hydrogen on that particular surface. These methods usually involve isotope labelling/exchange and/or chemical reactions which use reducing agents such as lithium aluminum hydride to generate hydrogen gas which is used as a measure of "surface 3 activity." Additionally, surface polarity can be determined by potentiometric titra tio n of the surface via certain characterization reactions and/or colorimetric methods (33). 1.3 Spectroscopy of Surfaces Spectroscopic surface characterizations include infra-red (34), Ram an (35), u ltra v io le t-v is ib le (26), electron spin resonance (36), nuclear magnetic resonance (37), X-ray photoelectron spectroscopy (38) and thermal desorption spectroscopy (39). A ll of these methods have had limited success in terms of the depth of information that can be derived. Fluorescence spectroscopy has been a recent addition to this arsenal of methods. 1.3.1 Fluorescence Spectroscopy Introduction As with a ll spectroscopic methods, fluorescence spectroscopy depends on a probe molecule. Pyrene was selected as the probe of choice in this study for several reasons. First, the interesting monomer-excimer system is very wel 1 documented and characterized (40) from early solution studies and more recent, m icellar studies (20,21,23,24,52). Second, kinetic studies can be done with this probe since the lifetimes of monomer and excimer are very distinct, about 450 and 65 nanoseconds respectively, making the two molecular entities easily identifiable and accessible with available instrumentation. F in ally, the monomer fine structure yields valuable and useful information with regards to the nature of the microenvironment (41,42). 1.3.2 Pyrene Solution Behavior The pyrene excimer system is one of the best characterized metastable photochemical intermediates known (40). The excimer, E, is formed when an electronically excited pyrene interacts with another pyrene in its ground state. Fluorescence results from this excited state dimer returning to the ground state. Specifical ly, the sequence of events can be described by the following scheme (Scheme 1.1) Scheme 1.1 A) P + hv ------------ » - P* (excitation) B) P* + P ------------^ P*— P (E) (excimer formation) C) P = * P (monomer fir s t order decays) D) P*— P (E) P + P (excimer decay) It should be noted that excimers are not exclusive to pyrene but also have been documented with other arenes (40). This excimer has a proposed "sandwich" structure in which the p arallel planes of pyrene are separated by about three angstroms (43). Figure 1.1 is a potential energy diagram outlining the energetics of the monomer-excimer system. Excimer formation is a diffusion controlled reaction and has been characterized in many different solvents (40,43). Since excimer formation is a bimolecular process, it is very dependent on the concentration of pyrene. Figure 1.2 shows that at concentrations of about 10“^M or less, the observed fluorescence is concentration independent, showing only the structured monomer emission at 395 nanometers. As the concentration of pyrene increases, the monomer 5 (excimer formation) pi—pi--------------------------- excimer decay L x J P— P Figure 1.1 Potential Energy Diagram of Pyrene Monom er and Excimer. r — * T & ,« » 0 i e 400 500 Figure 1.2 Various Concentrations of Pyrene versus Excimer Development. emission intensity decreases while a red shifted, re la tiv e ly structureless emission peak begins to increase. This new emission at 465 nanometers is the excimer emission. The structure in the monomer emission comes from transitions between the zeroth vibrational level of the excited state to discrete vibrational levels in the ground state (Figure 1.1). The excimer, an excited state dimer, is a minimum in the excited state potential curve. The emission from this minimum in the excited state results in a vertical transition, according to Frank-Condon rules, to a repulsive part of the ground state potential curve because the dimer state is dissociative (unstable) in the ground state. The dissociative ground state is a continuum of close lying vibrational states so that the emission appears as a structureless band. The absorption spectrum of pyrene in a hexane solvent, Figure 1.3, shows three characteristic peaks at 333, at 317 and at 303 nanometers. It is necessary to know the absorption spectrum of pyrene in order to (1) know what wavelength is needed to excite the molecule and (2) to be able to characterize and measure the amount of pyrene on the silicas. 1.3.3 Pyrene Monomer-Excimer Emission Analyses The analysis of the pyrene monomer-excimer fluorescence can be divided into (1) kinetic, (2) vibrational, (3) diffusion/probe mobility and (4) energetic analyses. 8 Figure 1.3 Pyrene Absorption Spectrum (Hexane Solution). 9 RELATIVE INTENSITY (ARBITRARY UNITS) o ov O s o The solution kinetics of the pyrene monomer-excimer system has been successfully analysed by employing a diffusive-encounter model developed from studies of m icellar systems (21). An early paper by Singer, Francis and Lin, (12) developed equations to analyse pyrene emissions on silylated and native s ilic a based on a successful micellar kinetic model (21) proposed by Singer and Atik. The kinetic equation, equation 1.1, describes the monomer behavior on the silicas where I/I g is the re la tiv e intensity, ke is the rate of the diffusive excimer formation, n is the average number of neighboring pyrenes and k^ is the usual decay of the monomer emission. At large values of t, equation 1.1 can be approximated as equation 1.2. Equation 1.1 Ln(I/Ig) = n(exp(-k0t ) - l ) - k i t (monomer decay) Equation 1.2 Ln(I / I q) = -(n + k^t) (at large t) A plot of equation 1.2, Ln(I / I q ) versus t, should be a linear relationship where the intercept would represent the number of neighboring pyrenes, n, and the inverse of the slope would yield the 1 ifetim e, 1 /k j = t . (53) In the analysis of the excimer decay, Singer, Francis and Lin were able to distinguish static dimers, those dimers already existing in the ground state from diffusive dimers. The time dependence equation for the excimer, equation 1.3, had to consider these static excimers. Equation 1.3 IN(t ) = Cx'exp(-k'^t) + xA(exp(-nk0t) - exp(-k'^t) (excimer decay) 11 IN(t) is the relative excimer emission intensity, C is an experimental constant, x 1 and x are the re la tiv e number of static and diffusive dimer sites respectively, k'i is the excimer decay rate and A is equal to nke/(k'i-nke). SFL were able to f i t their experimental data to equations 1.2 and 1.3. I t was proposed that the surface had particular sites which were responsible for the spectroscopic observation. This model permits a kinetic interpretation of pyrene on the surface. Others have developed similar equations to account for pyrene behavior on their particular surfaces (11,44,45). Vibrational analysis of the fine structure of the monomer emission gives information related to the microenvironment that the pyrene senses (24). Pyrene has been used successfully as a probe of the microenvironment in solution studies (20,21,24,41,42). Figure 1.4 shows the emission of the monomer in various solvents of different polarities. Pyrene has fiv e vibronic bands, two of them are forbidden. The 0-0 band is an orbital ly forbidden transition and is weak in nonpolar solvents. The band at 384 nanometers, band I I , is strong and is allowed. Band II also shows very l i t t l e variation with solvent polarity. The ratio of the 0-0 band/band I I is used as a measure of the microenvironment. As the ratio increases, the pyrene "senses" a more hydrophobic environment (42). Thus it seems possible to use pyrene to probe the microenvironment of a surface at low (monomeric) loadings. A com m on practice with pyrene studies is to measure the ratio between excimer and monomer emissions, I e/ I m * This ratio is used as a 1.2 Figure 1.4 Pyrene Emission in Solvents of Variable Polarities, (A) Water, (B) NaLS, (C) NaLS/dodecane/pentanol microemulsion, (D) dodecane/pentanol (24). 440 330 (nm) (n m ) M 440 (n m ) measure of the ease of excimer formation (21). In solution, this ratio varies inversely with the viscosity since excimer formation is a diffusive process. Obviously, this ratio is very useful in determining the effect of viscosity on excimer formation in solution, but its interpretation on surfaces is not quite as clear. I t may represent just surface diffusion of pyrene neglecting any other factor, but more lik e ly , i t probably represents diffusion of pyrene as a summation of al 1 forces; e.g., surface and pore effects. Related to the I e/ I m ratio is the activation energy of excimer formation. Equation 1.4 relates the excimer and monomer emission intensities to an expression for the activation energy of pyrene excimer formation at low temperature (23,40) (Figure 1.5). Equation 1.4 I e/ I m = (k'dm /kfm)exp(-Wdin/kT) k'd m and kfm are the rates of excimer formation and monomer decay respectively, W d m is the activation energy of excimer formation, k is the Boltzmann constant and T is the temperature. A similar equation, equation 1.5, is given for excimer formation at high temperatures. Equation 1.5 I e/ I m = (kfd*k'dm * ( 1M)/kfm*k 'md)exp(B/kT) B is the excimer binding energy, kfd is the rate of excimer emission decay, (^M) is the monomer concentration, and k '^ is the rate of excimer dissociation. In both equations, a plot of Ln(Ie/ I m ) vs. 1/T should be linear with deviations representing c ritic a l temperatures. Values of the activation energy for excimer formation are obtained from the slope of these plots. From these various methods of analyses borrowed from solution studies of pyrene, som e kind of interpretation can be made for pyrene on surfaces. kdm M + h i / k m d M 2 M + h i / Figure 1.5 Schem e for Pyrene Monomer-Excimer Formation and Decay. 1.4 Introduction to a t Family of Si 1 icas This dissertation w ill concern its e lf mainly with the fluorescence spectroscopy of pyrene on a family of silicas. The work presented w ill examine the relationship between spectroscopic and adsorption data. The family of silicas consists of an "amorphous"(S24), a "mesoporous"(S72), and a "microporous"(S63) s ilic a , a ll obtained from a supplier, Grace-Davidson, and their silylated counterparts which were synthesized in this work. The adjectives used were the supplier's. These porous silicas are only one form of a family derived from the polycondensation of orthosilicic acid, Si(0H)4 (46,47). Figure 1.6 shows a representation of what constitues a si 1ica.surface; siloxane bridges interspersed between tetracovalent si Icon atoms. The method of preparation dictates largely the physical properties of the s ilic a . For example, native s ilic a is re la tiv e ly hydrophilic but, an ethoxy derived s ilic a shows re la tiv e hydrophobicity. The pH at which the condensation is done w ill dictate the formation of small pores, usually an acidic pH optimizing at a pH equal to two, or large pores, usually at a basic pH. Two surface functions exist on the surface, siloxane bridges and surface hydroxyls (silanols). Silanols are further divided into geminal, vicinal and isolated categories as shown in Figure 1.7. The surface population of these silanols can vary from 100% for unheated, hydrated s ilic a to 0% for silicas exposed to heat at 1100 degrees Centigrade (46). Silanols serve as an important surface and pore species, contribute to the structural framework of the s ilic a and ________________________ 17 Figure 1.6 Representative Silica Surface. 00 Figure 1.7 (B) Siloxane Different Types Silanols, (A) Isolated, Bridge, (C) Geminal, (D) Reactive Silanol, (E) Physisorbed Water. X ° \ / 5 o ^ I x .1 1 / CD I i f ) X i P - (75 t / / X LU ''O - x o ' / X * * * \ / o o I ^ ° i f ) X O ( f ) - “ < I I ^ 20 provide important sites of hydrophilic reactions; namely adsorption of water. I t has been demonstrated that heat treatment, particularly above 400 degrees Centigrade, w ill cause severe population changes in the types of surface hydroxylations (46). As these changes occur, the native characteristics also change; i.e. as s ilic a is exposed to higher temperatures, the population w ill shift to favor an increase in the number of siloxane bridges because of the increase in dehydroxylati ons. These surface groups provide the main interaction between the surface and molecules near the surface. I t is seen to be particularly important with regards to water adsorption on s ilic a . Water can be physisorbed or chemisorbed. Figure 1.7 shows an example of physisorption of water. Chemisorption is a more permanent effect and is a stronger interaction between the surface and water. Water adsorbed in this manner can generate new silanol groups on the surface. In fact, this is what happens to a surface re la tiv e ly devoid of surface hydroxyls when it is in the presence of a hydrated environment (46,47). Physisorbed water has very weak interactions and is usually eliminated from the s ilic a surface with heating at 150 degrees centigrade. This kind of adsorption also does not change the population of surface hydroxyls (46,47, 49). In the pore, selective silanol types are found and water plays an important role here because the heat of adsorption is greater than on the surface (46,47,49). There is a group of silanols which have been called "reactive" in the sense that because of their close proximity 21 towards each other, hydrogen bonding can occur between the hydroxyls (46-49). I t has been demonstrated that the heat of adsorption of aromatics near these hydroxyls is greater when compared with the "usual" surface hydroxyl and another observation was that silicas with surface areas less than 500 rrr/gm had small populations of these hydroxyls. Thus, it seems lik e ly that such "reactive" hydroxyls are in operation in the pores. A contiuum of events can occur on the surface as well as in the pores because of the varieties of silanols. 1.5 Modification of Si 1ica and Pyrene on Si 1ica I t has already been noted that modification of the silanols w ill result in a s ilic a with different properties from the native s ilic a depending on the type of substituent. A n interesting modification uses a s ily la tin g agent to react with the hydroxyls, in particular a reaction which substitutes a bulky trimethyl si ly l group. Theoretically, such a modification should leave a surface re la tiv e ly hydrophobic compared with the original s ilic a since the s ily la tio n should have homogenized the surface and the trimethyl s ily l function is hydrophobic. An interesting claim is that hydroxyls deep in som e fine pores are not accessible to s ily la tio n (49,50). S ilylatio n is an interesting modification because of its potential effect on diffusion of pyrene. Access to pores may be blocked because of the size of the trimethyl si ly l function so that pore effects could be negligible (49). Surface diffusion of pyrene on s ilic a has been demonstrated (8,12) and the presence of preferential sites of 22 adsorption have been shown in which pyrenes are statistical ly distributed onto these surface sites. From these sites, excited pyrenes can diffuse and encounter ground state pyrenes to form excimers or there may be som e effect in operation to allow for som e ground state associated complexes to exist which can lead to excimer emission. Conventional studies (45,8) treat'the diffusing pyrenes as simply a diffusing species with no barrier for its migration. These studies also use two parameters in their analysis of pyrene kinetics. Complications arise because diffusing pyrenes may not necessarily have com e from nearest neighbors in the sam e site. Studies of pyrene on s ilic a have been extensive (1-15) but the depth of the studies has been limited to a few silicas. There has never been a systematic study on a series of native silicas and their silylated counterparts. I t is hoped that this work w ill help elucidate the importance of pore and surface effects. Variable temperature work w ill explore the energetics of excimer formation and diffusion on s ilic a surfaces. From this, a model w ill be proposed in which the importance of both surface and pore effects contribute to the overall behavior of pyrene on s ilica. 23 I REFERENCES 1) Weis, L.D., Evans, T.R., Leermakers, P.A., JACS, 90(22), (1968)6109 2) de Boer, J.H., Z. Phys. Chem., B14, (1931)163 3) Hara, K., de Mayo, P., Ware, W., Weedon, A., Wong, G., Wu, K., Chem. Phys. L e tt., 69(1), (1980)105 4) Oelkrug, D., Radjaipour, M., Z. Phys. Chem., NF88, (1974)23 5) Ishida, H., Takahashi, H., Tsubomura, K., Bui 1. Chem. Soc. Jpn., 43, (1970)3130 Oelkrug, D., Radjaipour, M., Z. fu r Phys. Chem. Neue Folge Bd, 123S, (1980)163 de Mayo, P., Bauer, R., Ware, W., J. Phys. Chem., 86, (1982)3781 de Mayo, P., Bauer, R., Ware, W., Natajaran, L., JACS, 104, (1982)4635 Thomas, J.K., Chandrasekaran, K., J. of Col 1oid & Interfacial Sci., 100, (1984)116 10) Lochmuller, C.H., Colborn, A.S., Hunnicutt, M.L., Harris, J.M., JACS, 106, (1984)4077 11) Bauer, R., de Mayo, P., Ware, W., Natarajan, L., Can. J. | Chem., 62, (1984)1279 i % 12) Singer, L., Francis, C., Lin, J., Chem. Phys. L e tt., 94, (1983)162-167 24 t I 13) Avnir, D., Daniel, H., Grauer, Z., J. Colloid & Interfacial ! | Sci., 1983 | ! 14) Turro, N., Cheng, C.C., JACS, 106, (1984)5022 i | 15) Drake, J.M., K1 a fte r, J., J. of Lumin. 31-32, (1984) 642. ! 16) Farwaha, R., deMayo, P., Schauble, J.H., Toong, Y.C., J. Org. I | Chem. 50, (1985) 245 j 17) Johnston, L., Wong, S.K., Can, J. Chem. 62 (1984), 1999 118) Donati, D., Fiorenza, M., Sarti-Fantoni, P., J. Heterocycl. Chem., 16, (1979)253 19) Kotzias, D., Hustert, K., Pari ar, H., Kork, F., Naturwissen- i I schaften, 70, (1983)413 | 20) Singer, L.A., Soln. Behavior of Surfactants: Theo. & App1ied | Aspects, K.L. M ittal & E.J. Fendler eds., Plenum Press, New | | York, 1982 21) Singer, L.A., A ttik , S., Nam, M., Chem. Phys. L e tt., 67, (1979)75 22) Csicsery, S.M., Chem . in Britain (May,1985) 23) Aoudia, M., Rodgers, M., Wade, W.H., J. Phys. Chem. 88, (1984) 5008 24) Thomas, J.K., Chem. Reviews, 80(4), (1980)283 | 25) Boehm, H.P., Knozinger, S., Catalysis-Sci. and Tech., Vol.4, M. Boudart, J.R. Anderson eds., Springer-Veri ag, New York, 1982 26) Brunauer, S., Emmett, P.H., Tel le r , E., JACS, 60, (1938)309 27) Cohan, L.H., JACS, 60, (1938)433 Rao, K.S., J. Phys. Chem., 45, (1941)506 Do! 1 imore, D., H eal, G.R., J. Appl ied Chem., 14^, (1964)109 Roberts, B.F., J. Colloid Interface Sci., 23t (1967)266 Brunauer, S., M ik h a il, R.S.H., Bodor, E., J. Col loid & Interfacial Sci. 24, (1967) 45131) Brunauer, S., Deming, L., Deming, W., Teller", E., JACS, 62, (1940)1723 Drushel, H.V., Sommers, A.L., Anal .Chem., 2!8, (1966)1723 Ibach, H., Mul le r , J.E., C atalyst Charact. S ci., ACS Symp. Ser. #288, (1985)464 and references therein Maulhardt, H., Kunath, D., Appl. Spectrscopy, 34, (1980)383 and references therein Flockhart, B.D., Surface and Defect. Prop, of Sol ids, Vol. 2, The Chemical Society, London, (1973)69 Derouane, E,G., Fraissard, J., F r ip ia t, J.J., Stone, W., C atal. Rev., 7, (1973)121 Roberts, M.W., Adv. C atal., 29, (1980)55 Czanderna, A.W., Vacuum Microbal. Tech., 6, (1967)129 Birks, J. B., Photophysics of Aromatic Molecules, Wiley- Interscience, New York, 1970 and references therein Nakajima, A., Bui 1. Chem. Soc. Opn., 44, (1971)3272 Dong, D.C., Winnik, M.A., Photochem. and Photobio., 35, (1982)17-21 43) Turro, N., Modern Molecular Photochemistry, The Benjamin/Cummings Co., Menlo Park, 1978 | 44) Suib, S.L., Kostapapas, A., JACS, 106, (1984)7705 | 45) de Mayo, P., Organic Phototransformations in Nonhomoqenous ! Media, ACS Symp. Ser. #278, M.A. Fox ed., (1984)1-19 i 46) Synder, L.R., Separation Science, 1(2 & 3), (1966)191-218 I j 47) Okkrese, C., Phys. and Chem . Asp. of Adsorbants and Catalysts, j j B.G. Linsen ed., Academic Press, New York, (1970) | 48) Adamson, A.W., Phys. Chem. of Surfaces, 4th ed., Wiley and I Sons, N ew York, (1982) j j 49) Synder, L.R., Ward, J.W., J. Phys. Chem., 70(12), (1966)3941 j 50) Nishioka, G.M., Schranke, J.A., J. Colloid & Interfacial Sci. ! 105(1), (1985) 102 ; i ii ] 51) Gregg, S.J., Sing, K.S.W., Adsorption, Surface Area and Porosity, j 2nd ed., Academic Press, New York, (1982) | 52) Castano, F., M artinez, E., Spect. L e tt., 17(8), (1984)477 | I 53) In the model for monomer decay on s ilic a , an adsorption site is defined to be an area which is encompassed by a diffusing excited state pyrene molecule. In the following reaction scheme, SPPn is a pyrene molecule surrounded by n neighboring pyrenes within a given adsorption site. The fraction of total 1ig h t(Ia) absorbed by P is given by anI a where a is the Poisson weighting factor, e.g., the fraction of excited pyrene molecules residing in a site SP*Pn. The reaction pathways available to P* are those defined by k^, the usual radiative and non-radiative j decays, and the diffusive excimer formation which is given by nke. 27 SCHEM E FO R M O N O M ER DECAY: (1) SPPn ---------► SP*P„ (2) SP*P„ — SPP„ (3) SP*P„--------► SPV(n-l) CHAPTER 2 SURFACE AREA CHARACTERIZATION O F NATIVE SILICAS A N D SILYLATED SILICAS 2.1 Introduction With classical methods of solid characterizations, the adsorption isotherm is an important relationship describing the interaction between an adsorbate and an adsorbent. In this dissertation and many other typical solid characterizations (1-3), the method involves the adsorption of a gas at fixed temperature and pressure to the solid. This forms the essential basis for the Brunauer, Emmett and T e lle r (BET) method for the determination of surface area (4). 2.2 The BET Method of Surface Area Determinations The BET method of surface area determination is recognized as the standard against which other methods of surface area determinations are matched. The model its e lf is a sim plification, but when applied in discriminating situations, i t can quickly and easily obtain surface areas which do not vary m uch from the determinations by other methods. This method involves the adsorption of a gas at a fixed temperature and pressure to a particular solid. The mathematical relationship describing this interaction is given by Equation 2.1. Equation 2.1 n = f(p/p°) T, gas, solid where n represents the amount of gas adsorbed, p® is the saturation 29 vapor pressure of the adsorptive, p is the pressure of the adsorptive, T is the temperature for a particular gas and solid. The adsorption of the gas onto a solid is the resultant of the forces of attraction between atoms of the solid and gas. These forces f a l l into two categories; physical and chemisorption. Physical adsorption forces are best described by the general formula relating the energy of two atoms; namely the Lennard-Jones Potential as given in Equation 2.2 (3). Equation 2.2 e(r) = -Cr"^ + Br"*^ where c(r) is the total energy of two atoms, C is a dispersion constant associated with the instantaneous dipole-dipole, dipole- quadrapole, and quadrapole-quadrapole interaction with dipole-dipole interaction being the most significant factor, r is the distance between atoms and B is a constant which expresses the effect of short range repulsive forces. More specifically, the interaction of the gas to a solid can be represented by the potential function given in Equation 2.3. Equation 2.3 ^(r^j) = _r . -Ty*. ® + R. . f r • •” ij^ i j Dij£ r i j where i and j are the gas and solid respectively and in particular r^-j being the distance between gas and solid. Isotherms can be conveniently classified by the scheme of Brunauer, Deming, Deming and T e lle r (BDDT) (5) into five distinct categories as shown in Figure 2.1. Each classification has specific features according to the interaction between adsorbate and adsorbent. Types IV and V possess a 30 Figure 2.1 Brunauer, Deming, Deming, T e lle r Isotherm Classifications. 31 Relative pressure 9/p Am ount of gas adsorbed M 9 C O ro hysteresis loop in which the lower part of the loop represents measurements taken at progressive additions of gas and the upper loop representing progressive withdrawals of gas. I t is the type IV isotherm which w ill be dealt with extensively in this dissertation. The BET model is based upon a kinetic model proposed by Langmuir (6) in which the surface was regarded as having an array of adsorption sites. The condensation rate was related to the rate of evaporation at equilibrium for this array of adsorption sites. Equation 2.3 defines the condensation rate of a unit area of surface. Equation 2.3 a^kOgp where p is the pressure, k is the gas constant from the kinetic theory of gases, a^ is the condensation coefficient (the fraction of incident molecules which condense on the surface) and © q is the fraction of bare sites. (0q + 6^ =1 where is the fraction of occupied sites.) Equation 2.4 defines the evaporation rate. Equation 2.4 zm ep-jexpC-qj/RT) where zm ©^ defines the number of adsorbed molecules, zm is the number of sites per unit area, q^ is the heat of adsorption and is the frequency of oscillation of the molecule orthogonally to the surface. At equilibrium, equations 2.3 and 2.4 are equal to each other and are represented by Equation 2.5. Equation 2.5 ajK0p: = zm 9jv^exp(-q^/RT) or rearranging and solving f o r 0 j Equation 2.6 0^ = a^p/(a^«p + zm v^exp(-q^/RT)). An alternative form for is to allow it to equal the ratio of the 33 number of moles adsorbed, n, to the number of moles to achieve a monolayer, nm . Equation 2.7 = n/ nm = B P/(1 + BP) where B is equal to (a^K/zmv^)exp(q^/RT). Equation 2.7 is the Langmuir equation for monolayer adsorption. The BET method applies this equation to a m ultilayer situation and obtains the working form of the BET equation. The application to a m ultilayer situation describes the rate of condensation of a gas in the fir s t monolayer which is equal to the rate of evaporation of gas from the second layer at equilibrium, Equation 2.8. Equation 2.8 a2*<pSi = z^^v^expC-q/RT) For the ith layer, the equation would be given as Equation 2.9 a-jKpe^-jj = z^Q^v^expC-q/RT) Further rearrangement of equation 2.9 and solving for zm leads to the number of molecules adsorbed, Z, for a particular surface, A, as represented by Equation 2.10. Equation 2.10 Z = Azm(© ^ + 2 02 +...+ i'e'^) The number of moles adsorbed, n, w ill be represented by Equation 2.11. Equation 2.11 n = (Azm ?:(ie1 -))/Nav where Nav is Avogodro's number. The application of the BET method assumes that (1) a ll layers except the fir s t have a heat of adsorption equal to the heat of condensation, (2) in a ll layers except the fir s t, a ll evaporation/condensation conditions are identical and (3) when p = pO then the gas condenses to a bulk 1iquid(infinite layers). 34 Summation, substitution of the original terms and rearrangement of equation 2.11 yields the working BET equation. Equation 2.12 (p/p°)/n( l- p/p°) = l/n m c + (c -l/n m c)p/p0 In determinations of the surface area, plots are constructed of (p/p^)/n(l-p/p^) versus p/p^. This is a linear plot with deviations from lin e a rity at values of p/p^ > 0.05 or p/p^ s 0.03. From the slope and intercept of this plot, the values of nm and c can be calculated. The term c is related to the heat of absorption although som e (3) believe that its use is a crude estimate. The BET method is used because of its simplicity in approximating the surface area in an actual determination. The values generated do not vary m uch compared with the use of other elaborate methods. Som e of these other methods claim less "subjectivity" and have fewer assumptions for finding the surface area (4,7). 2.3 Results and Discussion 2.3.1 Native Si 1 ica Results Surface area analyses were done on three unmodified (native) samples, an amorphous (S24), a mesoporous (S72), and a microporous (S63) silicas. The isotherms, Figures 2.2 to 2.4, are of the type IV isotherm as classified by the BDDT scheme (5). Figures 2.5 to 2.7 are plots of x /v (l-x ) versus x where x is p/p^. These plots are linear for the range 0.05 > x > 0.03. Table 2.1 summarizes the data and compares the surface area obtained in this work, for the native and silylated silicas, with values supplied by the manufacturer. I t can be seen from 35 Table 2.1 Summary of Surface Areas and Porosity Distributions of Native and Silylated Silicas Surface Area Given Surface Percentage of Pores Silica (m2/qm) Area (m2/gm) in Pore Range S24 415 m m m m ^ 13% less than 8 A 38% 12 to 32 A 49% greater than 32 A S72 398 340 20% less than 14 A 22% 14 to 32 A 58% greater than 32 A S63 925 675 83% less than 6 A 17% greater than 10 A SS24 453 25% less than 18 A 28% 18 to 40 A 47% greater than 40 A SS/2 258 m m — _ 16% less than 5 A 35% 5 to 20 A 49% greater than 20 A SS63 462 _ _ _ 49% less than 7 A 45% 7 to 30 A 6% greater than 30 A 36 80 70 - 6 0 - 5 0 - 4 0 - 3 0 - 20 1 0 - o H 1 ------------------- 1 ------------------- 1 ------------------- 1 ------------------- 1 ------------------- 1 ------------------- 1 ------------------- 1 ------------------ 1 ------------------- 0 0 2 0 .4 0 £ 0.8 1 R E L A T IV E PRESSURE ( F T ) Figure 2.2 Adsorption Isotherm for S24. 2 60 260 2 4 0 220 200 180 160 HO 1 20 1 0 0 80 6 0 40 20 0 9 ( I » I 02 0£ R E L A T IV E PRESSURE (P /P ) Figure 2.2a Adsorption Isotherm for S24 (unheated). 90 8 0 70 60 5 0 40 3 0 20 10 0 02 R E L A T IV E PRESSURE (P /P ) Figure 2.3 Adsorption Isotherm for S72. VOLUME Al 6 0 - 5 0 - 40 - 3 0 - 20 - 0 0 2 0.4 0J6 1 R E L A T IV E PRESSURE (P /P ) Figure 2.4 Adsorption Isotherm for S63. ■ P * O 0. 004 0.0035 - 0.003 - 0.0025 - 0.002 - O jO O IS - 0 . 0 0 1 - 0.0005 - 0 - 0.02 0.04 0.06 0 0 8 0.1 0.12 014 0.16 R E L A T IV E PRESSURE (P /P > □ E X P E R IM E N T A L + BEST F IT Figure 2.5 BET Isotherm for S24. 0. 01 0.009 0.008 - 0.007 0.006 - 0.004 0.003 - 0.002 - 0 . 0 0 1 - 0 Q6 0.2 R E L A T IV E PRESSURE (P /P ) □ E X P E R IM E N T A L + BEST F IT Figure 2.5a BET Isotherm for S24 (unheated). X/fVQ-XM 0. 01 0 X 0 9 - 0.007 - 0.006 0.004 - 0.003 - 0.002 - 0 . 6 0 02 0.4 R E L A T IV E PRESSURE (P /P ) □ E X P E R IM E N T A L + BEST F IT Figure 2.6 BET Isotherm for S72. C O HX-DA1/X 0 . 0 1 0 .0 0 9 0 .0 0 8 0 .0 0 7 H 0 .0 0 6 a o o s 0 .0 0 4 0 .0 0 3 0.002 0 . 0 0 1 0 □ 02 I 0 .4 a e R E L A T IV E PRESSURE (P /P ) E X P E R IM E N T A L + B E S T F IT Figure 2.7 BET Isotherm for S63. 0 . 8 4 S » ■ P * Table 2.1 that there are slight discrepancies between the values obtained in this work and the values supplied. In order to c la rify the discrepancy, three determinations were done for each s ilic a with the end result the same or very close. 2.3.2 Silylated Si 1 ica Results The three native silicas were silylated and then analysed to obtain a surface area measurement. The silylated silicas are designated as SS24, SS72 and SS63 after their parent s ilic a . Figures 2.8 to 2.10 are the isotherms for the si ly l ated silicas and again, they are of the type IV class. Figures 2.11 to 2.14 are the plots of x /v(l-x) versus x for these silicas. There is lin e a rity for the range 0.05 >j x > 0.03 and from the slope and intercept, surface areas were calculated. Table 2.1 summarizes the surface areas determined. 2.3.3 Discussion of Results W hen the results for the in itia l determination were done, the discrepancies, between the manufacturer's specifications and these results, were fir s t thought to have com e from having not pre-dried the silicas prior to the determination because surface water would have occupied sites and could have potentially given a false surface area (1,3). (M ultiple determinations yielded similar results and this work was also corroborated by another group. (P. Radford))Surface adsorbed water is a contaminant p a rtic u la rly for these silicas because of the hydrophilic surface hydroxyls. Figures 3.1a and 3.4a are porosity 45 VO LUM E 50 - 4 0 - 3 0 - 20 - 1 0 - OB 0 Q < 6 02 0 .4 R E L A T IV E PRESSURE (P /P ) Figure 2.8 Adsorption Isotherm for SS24. - P » c r > 240 220 - 200 - 160 - MO - 1 2 0 - 1 0 0 - 8 0 - 6 0 - 4 0 - 20 - T T 0 0 2 0 .4 0 5 OB 1 R E L A T IV E PRESSURE (P /P > Figure 2.8a Adsorption Isotherm for SS24 (unheated). 5 0 - 4 0 - 3 0 - 1 0 j 6 0 .4 02 0 R E L A T IV E PRESSURE (P /P ) Figure 2.9 Adsorption Isotherm for SS72. VO LUM E 6 0 - 5 0 - 4 0 - 3 0 - 20 - 1 0 - 0J6 QA 02 R E L A T IV E PRESSURE (P P » Figure 2.10 Adsorption Isotherm for SS63. -p » ((X-DAl/X 0 . 0 1 0 .0 0 9 0X107 - 0 .0 0 6 - 0XJ05 0 .0 0 4 - 0 .0 0 3 - 0.002 - 0 . 00 1 - 0 .4 02 0 R E L A T IV E PRESSURE IP /P ) □ E X P E R IM E N T A L + B E S T F IT Figure 2.11 BET Isotherm for SS24. tn o ((X-UAi/X 0.006 0 .0 0 5 5 - 0.0G 5 - 0 .0 0 4 5 0 .0 0 4 0 .0 0 3 5 - 0 .0 0 2 5 - 0.002 0£ 0 0 .4 0. 2 R E L A T IV E PRESSURE ( P / P > □ E X P E R IM E N T A L + B E S T F IT Figure 2.11a BET Isotherm for SS24 (unheated). 0. 02 0.019 0.018 0.017 0.016 0.015 0.014 0.013 0 . 0 1 2 § o x r a £ 0 .0 1 x 0 .0 0 9 0 .0 0 8 0 .007 0 .0 0 6 0 .0 0 5 0 .0 0 4 0 .0 0 3 0. 002 0 . 0 0 1 0 0 02 0 .4 a 6 R E L A T IV E PRESSURE (P /P > Figure 2.12 BET Isotherm for SS72. □a1 c n ro X /IV fl-X M 0 .0 1 0 .0 0 9 - 0 .0 0 8 - 0 .007 - 0 .0 0 6 - ttOOS - 0 .0 0 4 - 0.003 0.002 - 0 . 0 0 1 - 0£ 0 .4 R E L A T IV E PRESSURE (P /P ) □ E X P E R IM E N T A L + BEST F IT Figure 2.13 BET Isotherm for SS63. cn c o distributions which show the effect of not having a "dry" sample, but the overall effect on the surface area is not noticeable. The surface area from those "wet" samples are not appreciably different from the dried samples; 443 m 2/gm versus 415 m 2/gm respectively for native s ilic a S24 and 485 m^/gm versus 455 rn^/gm respectively for the "wet" and "dry" si lylated SS24. I f a trend is present, then i t would be a smal 1 increase of the surface area for the "wet" surface. There does seem to be a dramatic reduction effect by s ily latio n of the surface; in the S63/SS63 case where the surface area of the former has been reduced by about 50% and S72/SS72 has a surface area reduction of about 61%. The amorphous S24/SS24 results are d iffic u lt to interpret because the surface is very inhomogenous. O n the native S24 surface, i t seems that the effect of s ily latio n was to have generated new surfaces. With S63 and S72, these two surfaces are m uch more regular with S63 having the most regularity. The property of regularity can be defined by looking at a pore distribution curve. S72 has a greater range of pores than does S63 and S24 has the greatest range of pores am ong a ll surfaces studied. Much of the surface can be found in these pores. S ilylatio n has been known to block these pores (9,10); in effect chemically f il l i n g in many gaps (9). In doing so, the surface area should decrease as it does here. I t is interesting that the two surfaces with a defined regularity would show such a dramatic decrease, while the amorphous S24 would not behave along the sam e lines. A possible reason for this may be because of the different types of hydroxyls that are available for reaction with 54 hexamethyldisi1azane (HMDS); namely the "reactive" hydroxyls mentioned in the Introduction chapter. O n the S24 surface, the pores vary widely and consequently the hydroxyls on this surface may be f a ir ly isolated from one another. O n the other two surfaces, the pores are wel 1 defined and are generally of small dimensions as w ill be seen in the following chapter. I f most of the surface area of S72 and S63 are in their pores, then it may be concluded that s ily la tio n w ill block entry into these pores, especially since the concentration of hydroxyls w ill be greater than for S24. A further enhancement is that since the pores are of small dimensions, the likelihood of "reactive" hydroxyls is increased. Scanning electron microscopy (SEM) studies done by Moldenhauer- Kopach (11) show that the S24 and S72 surfaces are "lumpy" while the S63 surface is "rock-like" with many fracture planes. These descriptions characterize the "gross" features of these surfaces. It was not evident from the SEM studies what s ily latio n did to the surface because the SEM did not have sufficient resolution to show any gross changes from the parent s ilic a . Pores were also not seen in those studies so m uch of the microscopic discussion is limited. The surfaces were d e fin itely altered because the si lylated product tended to be very hydrophobic compared with the parent s ilic a . This was shown with wetting experiments of the si lylated and parent silicas (11). 2.4 Experimental 2.4.1 Material s Three silicas, Syloids 24, 72 and 63 were obtained from Grace- ________________________________________________________________ 51 Davidson Chemical Division. Nitrogen gas was the adsorbate used in the BET analysis and was prepurified, 99.99% pure. Hexane solvent (Mallinkrodt) was used in the s ily la tio n procedures and was stirred and le ft standing over magnesium sulfate prior to use. Hexamethyldisi1azane (HMDS) was purchased from Aldrich and used as received. 2.4.2 Equipment The BET analysis was done on a high vacuum line fitte d with gas burettes and a digital manometer, accurate to + /- 0.5 torr, in order to volumetrical ly determine the surface areas. This apparatus was the normal equipment used in an undergraduate physical chemistry laboratory at the University of Southern California. Since nitrogen gas was the adsorbate gas, liquid nitrogen was the coolant used to condense the nitrogen gas. Once data points had been recorded, the data was processed and calculations were done using an IBM PC/PC clone using the Lotus spreadsheet as a template. 2.4.3 Methods 2.4.3.1 S ily la tio n Procedure About four grams of s ilic a dried, overnight at 150 Celsius, were placed into a three necked 500 ml round bottom flask. Then 300 ml of dry hexane, stirred and le ft standing over magnesium sulfate, was introduced into the flask. Dry nitrogen was bubbled in for about one hour while the mixture was stirred and heated to a reflux. About 20 ml ____________ 56 of hexamethyldisi1azane (HMDS) was slowly injected by syringe while making sure the reaction did not overheat excessively as the s ily la tio n reaction (9) is an extremely exothermic reaction. After the HM D S was added, the mixture was allowed to reflux for six hours while continuously being stirred. The reaction could be monitored by pH paper since ammonia was a secondary product and could be trapped out in a water trap. After the flask was cooled, the s ilic a was decanted and centrifuged. The decantate was poured into a large reservoir of water in order to effect destruction of any unreacted HMDS. The centrifuged s ilic a was washed many times with fresh hexane and placed back into the reacting flask. A second s ily la tio n was done because i t has been reported that there were s t i l l unreacted surface hydroxyls after s ily la tio n procedures (10). Furthermore, there was the possibility of water contamination either in the HMDS, capped and stored in the refrigerator, or som e water that was strongly adsorbed onto s ilic a which had not com e off during the heating or during the fir s t reaction (10,12). After the s ilic a was recovered from the second s ily la tio n , it was placed under vacuum overnight to remove the hexane solvent and dry the s ilic a . The resulting product was weighed and stored in a sealed v ia l. This procedure was repeated for each of the original silicas. 2.4.3.2 BET Measurements BET measurements were done on an apparatus described in the Equipment section. Samples were weighed accurately on an analytical 57 balance. Dead volume measurements were done before each BET run. After the measurements were made, the sample was exposed to high vacuum until the manometer did not record a reading when the sample was exposed to a heat gun. While the sample was degassing, i t was important to shake the sample to prevent gas pockets from developing in between the s ilic a . Once there was no reading on the manometer, the sample was exposed to the high vacuum for about two hours. Liquid nitrogen in a dewar was placed around the sample; making sure that the sample was well within the dewar to ensure that it was at liquid nitrogen temperature. The dewar was topped off when som e liquid nitrogen had evaporated. Measurements are made f ir s t ly by introducing 15 torr aliquots while accurately recording the manometer readings and allowing enough time for equilibration and then recording the pressure after equi1bration. Equilibration is the most important factor in this experiment because ca p illa ry condensation takes a certain amount of time to occur. Therefore, pressure reading must be f u lly stable before they are recorded. At a total pressure of 100 torr, aliquots of 50 torr were introduced until a total pressure of 300 torr was reached. Aliquots of 100 torr were then given until near ambient pressure was reached. Ambient temperature and pressure were recorded. Data points were then transferred to a worksheet programmed into the Lotus program. Plots of x /v (l-x ) versus x were then plotted and the region of 0.05 > x > 0.03 was then fitte d for the best f i t straight line. The slope and intercept were calculated as well as the surface area. _________________________________________________________________________ 58. REFERENCES Gregg, S.J., Sing, K.S.W., Adsorption, Surface Area and Porosity, 2nd ed., Academic Press, New York, (1982) Ponec, V., Knor, Z., Cerny, S., Adsorption on S o lid s , CRC Press, Cleveland, (1974)44-107 Adamson, A.W., Phys. Chem. of Surfaces, 4th ed., Wiley and Sons, New York, (1982) Brunauer, S., Emmett, P.H., T e lle r , E., JACS, 60, (1938)309 Brunauer, S., Deming, L., Deming, W., T e lle r , E., JACS, 62, i (1940)1723 j Langmuir, I., JACS, 38, (1916)2219, 40, (1918)1368 j Halsey, G.D., J. Chem. Phys., Jj6, (1948)931 j H i l l , T.L., Adv. Catalysis, XV, (1952)236 j Synder, L.R., Ward, J.W., J. Phys. Chem., 70(12), (1966)3941 j Nishioka, G.M., Schranke, J.A., J. Colloid & Interfacial Sci. j 105(1), (1985) 102 | Moldenhauer-Kopach, K., Dissertation, University of Southern j California, (1987) j i Synder, L.R., Separation Science, 1(2 & 3|, (1966)191-218 j CHAPTER 3 POROSITY DISTRIBUTIONS O F NATIVE SILICAS A N D SILYLATED SILICAS 3.1 Introduction Unlike surface area determinations which principally use the BET method, porosity determinations have been approached by a number of different methods (1-5). Part of the contribution to surface phenomena comes from the existence of pores. These pores can contribute not only in terms of surface area but they can also represent specialized domains which may behave atypical ly compared to surface behavior. Part of the specialization comes from the pore's defined borders; dimensions which can determine what special property or behavior a particular functional group may have compared to being on the surface where this functional group may not behave in any particular manner or have any enhanced properties. The determination of the pore distribution is done d irectly (6-8) or indirectly (1-5). Both types of methods are developed around and from equations relating to capillary condensation. Currently, the most direct method of determining porosity is by mercury porosimetry (7) where mercury is intruded into the pores. The pore size is then related to the pressure necessary to intrude the mercury. This relationship is represented by the Washburn equation. Equation 3.1 is a special case of the Young-Laplace equation which describes the 60 interaction between the pressures on opposite sides of a liquid-vapor interface; namely the meniscus of cap illary condensation. Equation 3.1 r p = -2ycos'0/Ap where r p is the radius of the pore, Ap is the difference in pressure between the gaseous and liquid mercury, y is a constant of integration and 2cos6 is a function relating geometrically the contact angle with the radius of the meniscus. A drawback with this method is that an assumption of a cylindrical pore shape is made. Indirect methods involve the use of the Kelvin equation (9,10,11), Equation 3.2, in the evaluation of porosity. Equation 3.2 Ln p/p° = (2yV -j/RT)l/rm where V-j is the molar volume of the liquid adsorptive, p^ is the saturation vapor pressure, rm is the m ean radius of curvature and would be equal to in fin ity when the liquid adsorbate is at its saturation vapor pressure. Equation 3.2 can also describe the phase change between liquid and vapor at a specific p for som e specific rm of a pore. The problem with a ll porosimetry methods is the unknown and variable geometry of the pore. Pores are not s tric tly a ll c y lin d ric a lly closed off tubes, wedged shaped ink bottles or any other geometric variation. Investigators (1-4) have modified the rm term to account for the variations in geometry. They have developed model equations s tric tly for cylinders, ink bottles and other geometries. The rationale behind a ll the modeling is that for som e surfaces, the predominant pore model may be a cylinder or an ink bottle, but the models do not exclude the 61 possibility of other pore shapes. Empirical data has shown that pore distributions do not vary to a great extent between models (5,12). One method, by Brunauer, Mikhail and Bodor (BMB), is called a "model less method" because i t does not assume any pore shape prior to analysis (12). 3.2 The "Model less Method" B M B has shown with this method that there is not a marked difference in choosing a cylindrical model or p a ra lle l plate model over any other model. In fact, they came to the conclusion that a ll the other models may have correction terms and complicated calculations but the end result with a simple p arallel plate or cylindrical model makes the other methods cumbersome because the distributions generated do not vary m uch from each other. What sets this method apart from a s tric tly cylindrical or p a ra lle l plate model is that the choice of model can be applied after som e in itia l analysis of the pore data. In other words, a preliminary distribution is generated to which a pore model can be applied. The calculations are simple and m uch of the data is borrowed from a BET determination. The basis of this method is a term called the hydraulic radius, r h. This term is defined as the ratio of a volume element to its corresponding surface area as in the following equation. Equation 3.3 r h = dV^/dS^ where the superscript k denotes a core element since no model has been 62 applied. The values of V and S are derived from adsorption-desorption isotherms such as those generated during a BET determination. V, the volume of gas adsorbed/desorbed, is easily determined, but S, the surface area of the core, is found by a method based on Kiselev's method of surface area determination (13). Kiselev was able to derive a general thermodynamic equation, Equation 3.4, for capillary condensation. Equation 3.4 yds = Aydn where y is the surface tension of the liquid adsorbate, ds is the surface that disappears when a pore is successively f il l e d by capillary condensation, Ay is the change in chemical potential and dn is the number of moles of liquid taken up by the pore. The ds term is not the surface of the pore but rather the surface of the core since the walls of the pore have an adsorbed film on them and thus, the core surface is s lig h tly less than the actual pore surface area. Ay can be approximated, by Equation 3.5, if the adsorbate gas is assumed to behave as a perfect gas. Substituting equation 3.5 into equation 3.4 yields equation 3.6 which is the integral form of the working "model less" method. Equation 3.5 Ay = RT Ln(p/p^) Equation 3.6 dSk = -(RT/y')(dn)(Ln(p/p°) where p/p^ is the re la tiv e pressure at the midpoint of the cycle and the volume of the core can be represented by Equation 3.7. Equation 3.7 d v ^ = (n^ - _i))Vi = dnV1 63 Substitution of equation 3.7 into equation 3.6 yields the working form of the "model less" method, Equation 3.8. Equation 3.8 dv^/dS = y V-|/RTLn(p/p®) where the lefthand side of the equation is equal to the hydraulic radius, r^. A plot of dv^/dr^ versus mean r^ is a core distribution. Up to this point of the analysis, no pore shape has been assumed or used. To convert from a core distribution to a pore distribution, a pore model is now applied. The hydraulic radius is equal to one half the radius of a cylinder and for a parallel plate model, the hydraulic radius is equal to one half the distance between the plates (12). The hydraulic radius is re a lly a special case of equation 3.2. ( r^ = rn/^) Corrections for the thickness of the adsorbed film should be done but do not add any new information to the distribution curve. The distribution curve without corrections tend to have s lig h tly greater values for the core volumes than for a corrected distribution curve. The hydraulic radius is the sam e value whether corrections are applied or not (12). It is interesting that the choice of pore model seems not to influence the pore distribution very much. This is probably because the models are too simplistic and not sophisticated enough to show dramatic differences. I t is quite obvious that pores are a continuum of shapes, sizes; and for any one model to describe this continuum would require a f a ir ly sophisticated set of equations. According to the literatu re (10-12), the simple cylindrical and p a ra lle l plate models seem to offer the best approximation of porosity. More complex 64 corrections and models seem not to change the porosity in any dramatic fashion. B M B suggests that instead of trying to apply a pore model, perhaps it might do just to consider the core distribution since no assumptions are made as to what the actual pore shape may be (12). I t should be noted that this analysis does not include very fine pores which require another method of analysis (14). (A fine pore is defined as a pore with a radius less than about 15 Angstroms.) Surfaces with fine pores usually exhibit an isotherm of the type I variety, Figure 2.1. This does not preclude the possibility that a surface could contain both fine pores as well as the larger variety. Most surfaces do have both kinds of pores with the resulting isotherm a composite of type I and som e other type isotherm. For the work here, micropore analysis is necessary to the extent that it can be demonstrated that micropores are present but their actual dimensions need not be known since the pyrene molecular dimensions, 7.2 Angstroms by 13.0 Angstroms, are such that i t would not f i t in the micropore. 3*3 Results and Discussion 3.3.1 Native Si 1 ica Results Three unmodified (native) s ilicas, S24, S72 and S63 were analysed for their pore distributions. Figures 3.1 to 3.3 are the calculated core distributions. Figure 3.1a is a core distribution when the sample of S24 was not heated. The surface area had only increased by about 6% , but the effect on the core distribution is dramatic. Surface water would have become another matrix surface for the nitrogen molecules to 65 P O R E DISTRIBUTION (DELTA V/DELTA r) 0.016 0.015 0.014 0.013 0 . 01 2 O j o n 0 . 0 1 0 .0 0 9 0 .0 0 8 0 .0 0 7 0 .0 0 6 0 X 0 5 0 .0 0 4 0 .0 0 3 0.002 0 . 0 0 1 4 0 3 0 0 1 0 20 A V E R A G E H Y D R A U L IC R A D IU S (A N O S T R O M S ) Figure 3.1 Porosity Distribution for S24. < y > oi Por e D istribatioalD eita Vofennp/Ddta r) 004 5 0 .0 4 - 0 0 3 5 - 0 .0 3 - 0 0 2 5 - 0.02 - 0.015 - 0.0! - 0 0 0 5 - 0 - 0 2 0 4 0 A v e n g e H y d ra u lic R a d iu s (a n g stro m s! Figure 3.1a Porosity Distribution for S24 (unheated). c n • v j s 0 ' to 6 -er'' probe and formed false surfaces for capillary condensation to occur in. I t is interesting that only large sized "pores" are formed by the surface water, from about a r^ of 20 angstroms to about 40 angstroms. The distributions for this hydrated surface indicates that a large portion of the volume is located in these "pores". O n the dehydrated S24, Figure 3.1, the distribution curve shows a surface which is re la tiv e ly well defined in spite of it supposedly being amorphous. The average hydraulic radius is about 10 angstroms which when applied to a cylindrical model yields a radius of about 20 angstroms or when applied to a p arallel plate model yields a distance between the plates of about 20 angstroms. Figure 3.2 is the distribution for S72 and i t has a greater range of pores than does S24 which is suprising. Furthermore, the supplier, Grace-Davidson quoted an average radius on S72 of 75 angstroms while this work calculates an average hydraulic radius of about 20 angstroms; cylindrical and p a rallel plate models yield a radius and distance between plates of 40 angstroms respectively. S72 has a greater volume in the smaller pores, of about 10 angstroms and smaller, than does S24. This was expected because S72 is supposed to be a surface with regular pores. The discrepancy in the average porosity of S72 by this work and Grace-Davidson may be an experimental error by this work. However, since multiple determinations in our laboratories gave similar results, this is not a lik e ly explanation. Other possibilities are errors by the supplier (15) or inadvertant alteration of the sample through handling. 68 P O R E DISTRIBUTION (DELTA V/DELTA rt 0.014 0.013 - 0 J O tt - 0 . 0 1 - 0 .0 0 9 - 0 .0 0 7 - 0 .0 0 6 a o o s - 0 .0 0 4 - 0 .0 0 3 - 0.002 - 0 . 00 1 4 0 3 0 0 1 0 20 A V E R A G E H Y D R A U L IC R A D IU S (A N O S TR O M S ) Figure 3.2 Porosity Distribution for S72. cn Figure 3.3 is the distribution curve for S63 and i t c le a rly shows a pore distribution dominated by small pores, with most of the volume being located in pores of about 4 angstroms radius or distance between parallel plates. S63 has an average hydraulic radius of about 10 angstroms or about 20 angstroms with cylindrical and parallel plate models. This was specified by the supplier, Grace-Davidson. I t is important to note the smaller pores on S63 with their well defined regularity when considering s ily la tio n effects since the smaller pores probably contain a large population of "reactive" hydroxyls (16,17). 3.2.2 Si 1 y 1 ated Si 1 ica Resul ts Three silylated silicas were analysed for their porosity. The three silicas were silylated versions of S24, S72 and S63. The silylated versions are designated as SS24, SS72 and SS63. Figures 3.4 to 3.6 are the core distributions for these silicas. Additionally, Figure 3.4a is a core distribution of SS24 without heating prior to the BET determination. This particular SS24 should have som e surface water. The surface area of this "wet" surface has about a 7 % greater surface area than the "dry" surface but as with the S24 "wet/dry" study, the core distribution is dramatically altered. In fact, most of the volume of the pores is located in the region of a hydraulic radius of 20 angstroms and greater. I t seems that on this surface, surface water has aggregated and formed a rtific ia l pores in which m uch of the c a p illary condensation is f illin g . The "dry" SS24, Figure 3.4, shows a surface in which the average pore is about 20 angstroms by cylindrical 70 P O R E DCSTRBB I ITIO NIDELTA. VOLUM E/DELTA r) 0.06 0 j 0 5 - 0 .0 4 - 0 .0 3 - 0.02 - 0 . 0 1 - ^ — a . ■ B —• 20 18 16 1 4 1 2 1 0 8 6 2 4 0 A V E R A G E H Y D R A U L IC R A D IU S (A N G S T R O M S ) Figure 3.3 Porosity Distribution for S63. P O R E DtSTRlBUTfON(DELTA VOLUME/DELTA ri 0.017 0.016 - 0.015 - 0.014 - 0.013 - 0. 012 - 0 . 0 1 - 0 .0 0 9 - 0 .0 0 8 - 0 .0 0 7 - 0 .0 0 6 - 0 ,0 0 5 0 .0 0 4 - 0 .0 0 3 0.002 - 0 20 A V E R A G E H Y D R A U L IC R A D IU S (A N O S TR O M S ) Figure 3.4 Porosity Distribution for SS24. ro P o re DistribotionfDelta Vohnne/Deita rl 0 .0 3 2 0 .0 3 0 .0 2 8 0 .0 2 6 0 .0 2 4 0. 022 0, 02 0.018 0.016 0.014 0 . 0 1 2 0 .0 1 0 .0 0 8 0 .0 0 6 0 .0 0 4 0 10 2 0 3 0 4 0 A v e ra g e H y d ra u lic R a d iu s (a n g stro m s) Figure 3.4a Porosity Distribution for SS24 (unheated). - ■ j G O and p arallel plate models. Much of the volume is found in pores of about 10 angstroms and less. Figure 3.5 is the distribution for SS72 and shows an average pore of about 20 angstroms as w ell. The difference here is that the range of pores 20 angstroms and smaller has a greater distribution than for SS24. Most of the pore volume in SS72 seems to be in pores of the range 6 to 20 angstroms, with som e volume in very fine pores. Figure 3.6 is the distribution of S63 and it cle a rly shows that only small and fine pores are present and account for most of the porosity. This surface shows an average hydraulic radius of about 10 angstroms, 20 angstroms by cylindrical and parallel plate models, which is the sam e as the other surfaces but again the main difference lies in the distribution of the pores. O n this surface, the major volume resides in the fine pores, of about 8 angstroms and smaller. 3.3.3 Pi scussion I t should be noted that the porosity determinations were calculated from a single BET isotherm on each s ilic a and that the structure seen in the porosity distributions may come from variations between data points on taking the derivative; e.g. the BET isotherm was not corrected for "noise". A more re lia b le distribution should have been determined by combining data from m ultiple BET isotherms or to use curve smoothing techniques to eliminate apparent structure. In comparing the core(pore) distributions between the native and silylated s ilicas, i t is evident that s ily la tio n does change the 74 P O R E DtSTRDEMJTfONfDELTA VOLUM E/DELTA d 0 .0 1 5 0 .0 1 4 0 .0 1 3 0. 012 - 0.009 - 0.007 - 0.006 - 0.004 0.003 0. 002 - 0 . 0 0 1 0 4 8 1 2 1 6 20 24 A V E R A G E H Y D R A U L IC R A D IU S (AN G STR O M S) Figure 3.5 Porosity Distribution for SS72. • ' j CJI PO R E DJSTRflBi/riO NtDELTA. VOLUM E/DELTA r) 0.04 0.035 - 0.03 - 0.025 0 .0 15 - 0 . 0 1 0.005 - 24 20 1 6 0 8 4 A V E R A G E H Y D R A U L IC R A D IU S (AN G STR O M S) Figure 3.6 Porosity Distribution fo r SS63. on overall porosity from a distribution of large pores to a distribution of smaller pores. This effect is best seen in the S24/SS24 and S72/SS72 s ilic a series. Comparing Figures 3.1 and 3.4 for the S24/SS24 series, the average pore radius stays about the same, 20 angstroms, but the distribution is shifted towards smaller pores on the silylated surface although there appears to be an increase in the 20 angstrom range as w ell. This effect is seen also for S72/SS72 series where the average pore radius decreased from about 40 angstroms to about 20 angstroms and is reflected in the shift of distribution to smaller pores, Figures 3.2 and 3.5. For S63/SS63 the extent of s ily la tio n is minor by comparison to the other surfaces. The volume of pores in both native and silylated versions is largely concentrated in the fine pores, but som e of the volume, on SS63, of s lig h tly larger pores have decreased. There is very l i t t l e change in the distribution between S63 and SS63. This observation is supported by a study by Nishioka and Schramke (18) which concluded that the role of a s ily la tin g agent once bound to the surface was to cover high energy sites and that small pores s t i l l have unreacted hydroxyls which can adsorb water in spite of having been silylated. This means that the observed decreased volume in larger pores is caused by sily latio n . The second interesting effect of s ily latio n is the homogenization of a ll of the surfaces. The silylated versions a ll retain som e characteristic of the parent; S72 has a range of pores as does SS72, but most striking is the similar average pore radius for a ll of the surfaces. A possible explanation may be that the pattern of s ily la tio n 77 mimics the re la tiv e positions of the silanols, on an ordered surface such as S72, with the result that the silylated surface retains a similar average pore radius as the original s ilic a . This might explain the S24/SS24 results in that S24, being an amorphous surface, has silanols randomly arranged. S ilylatio n of these randomly arranged silanols would result in randomly arranged trimethyl s ily l groups. Since they are random, they are ineffective as pore blockers with the result of keeping the same average radius. Silanols in smaller pores may be accessible to a certain extent; i t seems only silanols close to the surface can be silylated and that deep silanols are not readily accessible. S ilylatio n of the silanols close to the surface seems to be the principal cause of the shift in the distribution to smaller pores. The S72/SS72 series show a decrease in average pore radius and this can be explained by observing the S72 behavior coming from a more ordered set of silanols. S ta tis tic a lly , the silanols are randomly dispersed on the surface but the surface has more ordered pores than S24 and s ily la tio n of these more ordered pores results ip their being blocked which is reflected in the distribution shift and decreased pore radius. S63 is the most ordered of the surfaces, containing small pores which s ily la tio n is effective on the periphery which roughly keeps the average pore radius constant. A minor but interesting effect is surface water adsorption. Figures 3.1a and 3.4a show that the surface water at low temperatures becomes a matrix of its own; capable of expressing a contiuum of false pores which can be f i l l e d by capillary condensation. SEM studies by 78 Moldenhauer-Kopach (19), do not show a microscopic difference between native and silylated silicas. The only way to distinquish the two is that the native surface is hydrophilic and wets easily while the silylated surface is hydrophobic and is not easily wetted. 3.4 Experimental 3.4.1 M aterial s (20) 3.4.2 Equipment (20) 3.4.3 Methods The method of determining porosity has been discussed. The gas data from the BET determinations were manipulated into distribution curves by an IBM PC/PC-clone using a programmed worksheet (Lotus software). REFERENCES 1) Foster,A.G., Trans. Faraday Soc., 28, (1932)645 2) Do 1 1 imore, D., H eal, G.R., J. Appl ied Chem., JL4, (1964)109 3) Wheeler, A., Catalysis, vol. I I , Rheinhold, New York, (1955) 4) Cranston, R.W ., Inkley, F.A., Adv. in Catalysis, vol. 9, Academic Press, New York, (1957) 5) Roberts, B.F., 0. Colloid Interface Sci., 23, (1967)266 6) B artel 1, F.E., Walton, C.W., J. Phys. Chem., 38, (1934)503 7) Drake, L.C., Ind. Eng. Chem., 41, (1949)780 8) de W it, L., Scholten, J.J., J. C atalysis, 36, (1975)36 9) Thomson, W.T., P h il. Mag., 42, (1871)448 10) Adamson, A.W., Phys. Chem. of Surfaces, 4th ed., Wiley and Sons, New York, (1982) 11) Gregg, S.J., Sing, K.S.W., Adsorption, Surface Area and Porosity, 2nd ed., Academic Press, New York, (1982) 12) Brunauer, S., M ik h a il, R.S.H., Bodor, E., J. Col loid & Interfacial Sci. 24, (1967) 45131) 13) K is e le v , A.V., Usp. Khim. (in Russian), L4, (1945)367 14) M ik h a il, R.SH., Brunauer, S., Bodor, E.E., J. of Col loid Interface Sci., 26, (1968)45-53 15) I t is not known how Grace-Davidson evaluates their samples, but the results here are reproducible and previous work by Singer, Francis and Lin also noted this discrepancy. 80 Synder, L.R., Separation Science, JL[2 & 3^, (1966)191-218 Synder, L.R., Ward, J.W., J. Phys. Chem., 70(12), (1966)3941 Nishioka, G.M., Schranke, J.A., J. Colloid & Interfacial Sci. 105(1), (1985) 102 Moldenhauer-Kopach, K., Disseration, University of Southern California, (1987) The materials, equipment are the sam e as described in the Experimental section of Chapter 2. The method of calculation has already been discussed in this chapter. CHAPTER 4 PHOTOPHYSICAL STUDY O F PYRENE A D SO RB ED O N TO NATIVE SILICA 4.1 Introduction The use of pyrene as a probe on s ilic a has been demonstrated in a number of studies (1-15). These surfaces are not homogeneous (16-19) since each surface has its ow n properties which w ill affect the manner in which pyrene w ill react. One of the most challenging and interesting approaches uses the pyrene monomer-excimer system as a probe of surface structure. This study examines the response of the pyrene probe system to the varied s ilic a surfaces described in e a rlie r chapters. The effect of hydration is especially important because it determines the type of silanols (16,19) present which in turn influences the interactions with pyrene. An important consideration in doing these studies is how m uch surface water is present and how to eliminate it as an unwanted variable. The most com m on method for removing surface water is by heating, but excessive heating w ill cause the surface to restructure (16). Studies by Okkrese (20) and Nishioka (21) have suggested that the s ilic a surface does not restructure appreciably i f it is heated at or below 150 Celsius. The present work is an extensive study of a family of silicas, native and silylated , having variable amounts of adsorbed pyrene. Variable temperature studies on the pyrene monomer-excimer system, 82 from room temperature to liquid nitrogen temperature, were used to explore the importance of surface characteristics on probe binding, distribution and mobility. There have been only three other similar studies which were reported after initiatio n of this project (22-24). 4*2 Results of Pyrene on Native Si 1ica Pyrene was adsorbed onto three native silicas, S24, S72 and S63 by the method described in the Experimental Section (4.4). The surface area and pore distributions of these silicas were discussed in previous sections. A summary of their properties is provided in Table 2.1. Variable temperature time resolved (transient) spectra were recorded for the following samples: 6.50 X 10-3 M pyrene in hexane and at least three samples, "low", "medium" and "high", of pyrene adsorbed (loaded) onto the native silicas. (The low, medium and high terms are qualitative and refer to the amount of rela tiv e excimer emission which could be detected; e.g. an indication of probe distribution on the surface. The exact loadings are presented in the text.) The spectra of these samples are represented in Figures 4.x to 4.xx, gated at t=0 ns, and in Figures 4.xa to 4.xxa, gated at t=150 ns. ("x" denotes a whole number.) The term "gated" refers to when the spectra were recorded re la tiv e to the pulse excitation; e.g., t=150 ns spectra are those spectra taken approximately 150 nanoseconds after pulse excitation of the sample by the laser. (The t=0 is not actually at the same time of pulse excitation but closer to approximately 2 nanoseconds following triggering of the laser (25).) 4.2.1.0 Variable Temperature Time Resol ved Spectra of Pyrene in Hexane Figures 4.1 and 4.1a are the spectra of a 6.50 X 10“^ M pyrene in hexane solution. This sample was used as a control sample for comparison with the surface studies. The variable temperature work was restricted to the range from room temperature to -90° Celsius which is the temperature hexane solid ifies. At room temperature, this sample showed predominantly the characteristic broad, structureless excimer emission with only a small amount of structured monomer emission. The excimer emission has maximum intensity at 474 nm . The monomer, when it becomes prominent at the lower temperatures, has emission peaks at 374, 380, 384 (maximum intensity emission), 390 and 394 nm . The spectra at t=150 show that the excimer emission is more intense than at t=0 at a ll temperatures. These results are summarized in Table 4.1 along with the data on the re la tiv e monomer vibrational intensities, and the ratio of excimer emission to monomer emission (I e/ I m ). The tabulated results on the re la tiv e amounts of monomer and excimer emission reflect the importance of diffusion. Excited pyrenes must have som e time to diffuse in the solvent to encounter a ground state pyrene which is reflected by the increased emission at a longer time. Of interest are the spectra at -50°C where at t=0, the monomer and excimer emissions are almost equal intensities. In contrast at t=150, the excimer emission is the dominant emission. In the frozen sample at -90°C, monomer emission becomes the dominant emission with a small amount of excimer s t i l l being formed. 84 Figure 4. 1 Variable Temperature Spectra of Pyrene in Hexane at t=0. FLUORESCENCE INTENSITY (ARBITRARY UNITS) Figure 4. a Variable Temperature Spectra of 6.5 X 10' Pyrene in Hexane at t=150. RELATIVE INTENSITY (ARBITRARY UNITS) o - bo O ' 4.2.1.1 Emission Lifetimes Emission lifetimes were calculated for the monomer and excimer decays. i J Monomer lifetimes were calculated from the lim iting slope of the ! fluorescence decay of the monomer emission. A monomer lifetim e of 345 ! ! ns was determined which is consistent with literatu re values (26). I ! Figure 4.2 is a plot of the monomer emission decay for various i temperatures. The plot show a biexponential curve which gradually j converts into a near exponential curve on changing the temperature j J from room temperature to -90° C. Monomer lifetimes were calculated j I from the lim iting slope at large time and did not require deconvolution since the monomer lifetim e is sufficiently long to be i distinct from a ll other processes (27, 28). j I Figure 4.3 is the decay plot for the excimer emission. The decay j shows a nonexponential decay which converts into a near exponential j * decay on changing the temperature from room temperature to -90° C. Thej early nonexponential behavior represents the rise time for excimer j formation after excitation (26,27). As the temperature decreases, fewer excimers are formed because of slower probe diffusion in the j increasingly viscous medium. Monomer fluorescence dominates more and ‘ more. j * 1 j j 4.2.1.2 Vibrational Analysis of the Monomer Emission in Hexane -\ Table 4.1 contains vibrational information of bands I, I I , and ! * ’ [ \ I I I of the monomer emission. Vibrational information gives informations Table 4.1 Monomer Vibrational Intensities (6.5 X 10-3M Pyrene in Hexane) Normalized Monomer Band Intensities Sample I (374nm) I I (384nm) I I I (395nm) (465nm)/(395nm) 0, R.T. 1.00 3.11 2.61 3.70 0 C 1.00 2.43 2.39 3.25 -50 C 1.00 2.23 1.69 1.23 -90 C 1.00 2.12 1.62 0.25 150, R.T. 1.00 3.31 3.38 6.77 0 C 1.00 3.38 3.33 7.24 -50 C 1.00 2.48 2.18 3.68 -90 C 1.00 2.39 1.88 0.48 90 -05 □ t □+ -1 - -IS - - 2 5 - -3 - 280 200 240 1 2 0 1 6 0 TIME < N S > □ R.T. + ■ 0 C 0 -50 C A -90 C -3 figure O Variable Temperature Pyrene Monom er Emission Decay (6.5 X 10 M in hexane) •+ a -02 - ■0.4 C H - -0 6 - -12 - -L 4 - -18 - 22 - - 2.4 - - 2£ - 0 200 400 TIME ( N S I □ R.T. + 0 C o -SOC £ -90 C -3 Figure 4.3 Variable Temperature Pyrene Excimer Emission Decay (6.5 X 10" M in hexane). about the microenvironment (29,30) that the pyrene senses as was discussed in Chapter 1. Table 4.2 summarizes literatu re values (30) for solvents of different polarities. The pyrene monomer fine structure is reported by Winnik and Dong (30) to be insensitive to hydrogen bonding effects which they conclude makes the monomer an excellent probe of micropolarity. The analysis uses the ratio of band Ill/band I since band I I I is a strong and an allowed transition which does not vary with solvent polarity while band I is a weak and a forbidden transition which is weak in nonpolar solvents. Thus, an increase in the ratio indicates an increase in hydrophobic character of the microenvironment. A decrease in the ratio means an increase in the hydrophi1ic ity of the microenvironment. The observed vibrational band intensities of Table 4.1 correspond to hydrophobic environments. The value of 1.88 is greater than the literatu re value of 1.72 for hexane probably because of major contributions from excimer emission to the longer wavelength bands in the monomer emission. The ratio is significantly reduced at liquid nitrogen temperature, as expected, because of the decreased excimer emission. 4.2.1.3 Pyrene Mobi 1 ity in Hexane A related quantity included in Table 4.1 is the I e/ I m ratio. This particular ratio has been used to measure probe mobility and is inversely related to the solvent viscosity (29). Table 4.1 cle a rly shows, as the temperature decreases, the solvent becomes more viscous 93 Table 4.2 Normalized Pyrene Band Intensities in Various Solvents ; Normal ized Band Intensity Medium I II I I I Perfluorodecalin 1.00 1.92 Hexane 1.00 1.72 Cyclohexane 1.00 1.72 T etramethy1s i 1ane(28) 1.00 1.14 1.71 Dodecane 1.00 1.69 Hexadecane 1.00 1.67 Polydimethyl siloxane(28) 1.00 0.98 1.64 n-Butyl ether 1.00 1.19 n-Octanoic acid 1.00 1.08 t-Amyl alcohol 1.00 1.06 1,2,4-Trichlorobenzene 1.00 1.03 0-Dicholorobenzene 1.00 0.98 I Chloroform 1.00 0.80 Glycerol 1.00 0.63 Water(28) 1.00 0.60 0.63 Acetonitrile 1.00 0.56 Dimethyl sulphoxide 1.00 0.51 (Table from ref. 30 except for entries from ref. 28) until it freezes and the I e/ I m ratio also correspondingly decreases. The time dependence of the d iffu sively formed excimer is revealed by a smaller ratio at t=0 than at t=150 at a ll temperatures. i | j | 4.2.1.4 Activation Energy of Excimer Formation in Hexane j j The activation energy of excimer formation can be determined from I j the variable temperature work in conjunction with the theory j i introduced in Section 1.3.3 and equations 1.4 and 1.5. The | experiments here do not contain enough data points to derive precise j |activation energies, but do provide a meaningful estimate. Arrhenius j I (plots of the natural logarithm of I e/ I m versus the inverse of the i |temperature, are constructed. The photophysical behavior of pyrene has I I been documented to show a "high" and "low" temperature behavior (26,32). Each of these behaviors has been described according to equations 1.4 and 1.5. "Low" temperature behavior corresponds to a negative slope which yields the activation energy of excimer .formation, W^, while "high" temperature behavior corresponds to a I |positive slope which provides the excimer binding energy, B. The I [theory predicts that at "low" temperatures, the I e/ I m ratio w ill increase with increasing temperature while at "high" temperatures, the I e/ I m ratio w ill decrease with increasing temperature. Figure 4.4 is the activation energy plot for pyrene in hexane. It is known that pyrene has a c ritic a l temperature at or above room temperature depending on the solvent. The c ritic a l temperature is the point at which the pyrene changes from "high" temperature behavior to 18 - 16 - L 4 - L2 - 08 - 06 - 04 - 02 - -02 -04 - -0 6 - -08 - - L 2 - -L4 5 5 4.7 4 9 4 . 1 3.7 33 05 1/T X 1 ( 3 0 0 00 □ W-0 + ®t-l50 -3 Figure 4.4 Arrhenius Plot for 6.5 X 10 M Pyrene in Hexane. v O cn "low" temperature behavior. This plot shows a re la tiv e ly linear I relationship and is consistent with the low temperature behavior of i jpyrene, equation 1.4, in which the I e/ I m ratio should increase with | | increasing temperature. Had variable temperature been extended above I room temperature ranges, the plot should have shown another linear ! !region with a positive slope that intersects this linear plot at the |c ritic a l temperature. Using equation 1.4, the slope of this plot yields the activation energy of excimer formation of 0.12 eV which is comparable to the literatu re value of 0.11 eV (26). The la tte r has i j |been added to the plot for comparison. 4.2.2.0 Variable Temperature Time Resolved Spectra of Pyrene on S24 Figures 4.5, 4.5a to 4.8, 4.8a are spectra for various amounts of j pyrene adsorbed onto S24. The temperature range was from room jtemperature to liquid nitrogen temperature. Figure 4.5, 8.91 umole/gm |pyrene, shows predominantly monomer emission. The t=0 spectra show a j hint of an emission t a il at the end of the monomer emission. This t a il is probably a very low concentration of excimer since the t=150 spectra, Figure 4.5a show this ta il to have decayed to baseline. The * nature of this "ta il" is d iffic u lt to ascertain because i t persists in the t=0 spectra at -196°C. Som e of the excimer may be derived from diffusive encounters but it is more lik e ly that this excimer originates from a static dimer; e.g. a ground state dimer. As the temperature decreases, the spectra do not change appreciably in either time frame. The only changes noted are a greater definition and 97 Figure 4.5 Variable Temperature Spectra of 8.91 umole/gm Pyrene on S24 at t=0. 98 R T tn > I ^ 1 5 0 C 1 9 6 C ________ nm 4 6 0 6?0 Figure 4. Variable Temperature Pyrene on S24 at Spectra of 8.91 umole/gm t=150. (ARBITRARY UNITS) RELATIVE INTENSITY o * intensity of the emission peaks because vibrational levels become better defined at lower temperatures. The monomer emisson peaks are at 373, 383, 395(maximum emission peak)nm with a broad 414 n m peak. The 373 n m peak becomes maximal at -150°C. Figures 4.6 and 4.6a, 14.5 umole/gm pyrene, show approximately equivalent amounts of monomer emission to excimer emission at room temperature. Monomer emission peaks are s t i l l at the sam e wavelengths as in the lower loaded sample, but the excimer emission is broad, with a maximum at 474 nm . A comparison of the t=0 spectra with the t=150 spectra shows that excimer emission at t=0 greater than excimer emission at t=150. This is contrary to what was seen for pyrene in hexane. This effect has been seen by Yamazaki, Tamai and Yamazaki (33) who studied Langmuir-Blodgett films of pyrene. Their study and studies by Singer, Francis and Lin (12) and de Mayo (22,34-36) suggest that static dimers are responsible for this effect because these static dimers have a short lifetim e. The spectra show that decreasing the temperature in itia l ly causes .a slight increase in the amount of excimer emission, but after -100°C, the excimer emission begins to decrease. In an e a rlie r study by Singer, Francis and Lin (SFL) (12), two types of excimers, diffusive and static, were proposed to be on the s ilic a surface. At room temperature, both types of excimers are observed but can not be distinguished from each other. However, by decreasing the temperature, the diffusive population will decrease so that the static dimer population, proposed to be ground state dimers (12,22,34-36), will begin to dominate the excimer emission. Figure 4.6 Variable Temperature Spectra of 14.5 umole/gm Pyrene on S24 at t=0. RELATIVE INTENSITY (ARBITRARY UNITS) \ I Figure 4.6a Variable Temperature Spectra of 14.5 umole/gm Pyrene on S24 at t=150. 105 (ARBITRARY UNITS) VE INTENSITY L O R EL ATI o- o o c r > Figures 4.7, 4.7a and 4.8, 4.8a, 36.5 and 91.7 umole/gm of adsorbed pyrene, respectively, show large amounts of excimer emission with a small contribution from monomer emission. Again, the monomer emission intensity seems to be greater at t=150 for a ll temperatures than at t=0. Excimer emission has a maximum at 474 n m and the only prominent monomer peaks are the 373 and 395 n m peaks. Decreasing the temperature causes in it ia lly a slight increase in the excimer emission until -100°C. After -100°C, there is a decrease in the excimer emission with an increase in the monomer emission but the excimer concentration is so high that i t is s t i l l the dominant emission at -196°C. The detailed nature of the excimer, static versus diffusive, at these loadings is largely unknown. However, both kinds of excimer population seem to be present since the excimer emission dominates more in the early time frames after excitation which is consistent with static dimers. The excimer emission envelope appears to be asymmetrical and is more so at lower temperatures. This asymmetry may imply several complexes contributing to the excimer emission. Such asymmetry has been discussed by Yamazaki, Tamai and Yamazaki (33) who studied Langmuir-Blodgett films of pyrene and concluded that som e of the excimer emission they observed came from different kinds of dimers including the diffusive and static dimers. (Another interpretation would be that diffusive and static dimers should have the same spectra i f their geometries were the same, so that different spectra imply excimers with different geometries.) 107 Figure 4.7 Variable Temperature Spectra of 36.5 umole/gm Pyrene on S24 at t=& 108 RELATIVE INTENSITY (ARBITRARY UNITS} o o Figure 4. Variable Temperature Spectra of 36.5 umole/gm Pyrene on S24 at t=150. t 110 -196 C r e l a t iv e in t e n s i t y (ARBITRARY UNITS) Figure 4.8 Variable Temperature Spectra of 91.7 umole/gm Pyrene on S24 at t=0. 112 RELATIVE INTENSITY (ARBITRARY UNITS) Figure 4.8a Variable Temperature Spectra of 91.7 umole/gm Pyrene on S24 at t=150. 114 RELATIVE INTENSITY (ARBITRARY UNITS) 4.2.2.1 Decay Measurements of Pyrene on S24 The monomer lifetimes for the various amounts of adsorbed pyrene at various temperatures on S24 are given in Table 4.3. These lifetimes were derived from the fluorescence decays monitored at 395 nm , the monomer emission maximum. Figures 4.9 to 4.12 are the variable temperature monomer decay plots with various amounts of adsorbed pyrene. Figure 4.9, the monomer decay plot at room temperature of 8.91 umole/gm, shows near exponential decay with a very small hint of curvature which changes at about 50 ns to exponential decay. (The curvature w ill be related to the excimer decay on the samples with greater amounts of pyrene. At greater loadings of pyrene, i t w ill be more apparent that this curvature is related to the excimer decay.) The lim iting slope of the exponential decay corresponds to a lifetim e of 254 ns. As the temperature decreases, the decay curvature becomes greater and stays the same at a ll of the temperatures lower than room temperature. The lim iting slope changes and becomes smaller meaning a longer lifetim e for the monomer as the temperature decreases. The point of change from the unchanging early, faster decay to the la te r, slower decay occurs about 200 ns. Figure 4.10, 14.5 umole/gm pyrene, shows a near exponential monomer decay at room temperature which is very similar to the lower loaded S24. The lim iting slope yields a lifetim e of 228 ns. O n cooling this sample, the curvature becomes greater but does not change m uch with decreasing temperatures while the lim iting slope becomes smaller 116 Table 4.3 Pyrene Monomer Lifetimes of Variable Concentrations at Variable Temperatures (S24) Sample (umole/gm) Monomer Lifetime (ns) Temperature (C) 8.91 188.3 R.T. 2/6.9 -50 330.3 -100 3/2.9 -196 14.5 253./ R.T. 2/2.2 -50 303.2 -100 360.1 -196 36.5 136.3 R.T. 210.9 -50 218.8 -100 269.3 -196 91./ 114.3 R.T. 100.2 -50 280.8 -100 351.3 -196 -0 2 - -0 6 - -08 - - 1 - - 1 2 - -L4 - -18 - □ o -22 - -2.4 - - 2£ - - 2& - 600 400 0 200 TIME < N S R.T. + -50 C o 400 C d -196 077 K > Figure 4.9 Monomer Decay for 8.91 umole/gm Pyrene on S24. u -0 5 -1 -L5 -2 -2S -3 -3 S -4 0 200 400 600 B O O TIME ( N S ) R.T. + -50 C o -100 C & -196 077 K ) Figure 4.10 Monom er Decay for 14.5 umole/gm Pyrene on S24. “ o LN( R E LA TIV E U ¥ -OS - C -LS - -2 - - 2 5 -3 - V . a + aD 4^0 + io □ +<i. _ •£*p + £j + £o o + A + T 3 □ < 1 + A o + □ A o + □ A □ A 0 O + + □ + A □ A o & □ + £ ro o 1 0 0 200 3 0 0 400 T IM E (N S -1 0 0 c □ R .T . + -5 0 C < > -100 C a -196 0 7 7 K ) Figure 4.10a Excimer Decay for 14.5 umole/gm Pyrene on S24. which means a longer lifetim e for the monomer as was seen previously. Figure 4.10a is the excimer decay plot for this sample at various temperatures. I t shows a near exponential decay at room temperature, but on cooling, the slope at longer times becomes smaller, e.g. the decay becomes nonexponential. I t is interesting to note that the early part of the decay stays constant with decreasing temperature until it reaches about 150 ns. According to Figure 4.10, the monomer decay shows a similar effect at a time s lig h tly less than 200 ns. The intensity at which the excimer decay begins to deviate corresponds to about 25% of the in itia l excimer emission intensity. The corresponding part on the monomer decay is about 30% of the in itia l monomer emission intensity. A comparison with the emission spectra for this sample reveals that the excimer emission intensity at t=0 seems to be greater than at t=150. At t=0, the excimer emission intensity is at maximum while at t=150, the excimer emission decays to an intensity of about 22% of the original intensity. The monomer emission intensity is a maximum at t=0 but has decayed to about 50% of its original intensity at t=150. These observations suggest that the curvature seen in the monomer decays is caused by excimer decay and that at longer times, the decay becomes more representative of pure monomer decay. Figure 4.11, 36.5 umole/gm of pyrene, shows a monomer emission decay which is nonexponential at a ll temperatures. The most striking feature of this decay is that the faster decaying early component decays at the same rate with decreasing temperature. (There may even 121 LM RELATTVE -02 - - 0 4 - - 0 6 - -0 8 - - I - - 1 2 - -L 4 - -L 6 - -1 8 - -2 - -22 - -2 .4 - -2J6 - 0 00 200 3 0 0 4 0 0 T IM E (N S □ R .T . + -5 0 C o 4 0 0 C a -196 0 7 7 1 0 Figure 4.11 Monomer Decay for 36.5 umole/gm Pyrene on S24. K > ro LN IR ELA TfVE -02 - -04 - -OS - -08 - -1 - -L4 - -L 6 - -L 8 - -2 - -02 - -2.4 - - 0 6 - -08 2 4 0 200 0 8 0 CO 160 4 0 T IM E (N S □ R .T . + -5 0 C o -1 0 0 C 4 -196 0 7 7 1 0 ^ Figure 4.11a Excimer Decay for 36.5 umole/gm Pyrene on S24. n o u > be a faster decay at -196°C on closer examination.) The longer decaying component changes toward a smaller slope meaning an increased lifetim e for the monomer. The point of deviation between the two parts of the decay occurs at about 100 ns. In Figure 4.11a, the excimer decay decays nonexponential ly with the point of deviation at about 30 ns. The excimer decay also shows a com m on early decay with decreasing temperature until 30 ns. The longer decay component then has a smaller slope with decreasing temperature which is analogous to the monomer decays for this sample. In an e a rlie r study, Singer, Francis and Lin (12) were able to assign the faster decay component to static dimer and the longer decaying component to the diffusive excimer. I t is expected that the diffusive excimer lifetim e should be longer at decreasing temperatures because the monomer lifetim e is extended due to slower diffusion. The excimer intensity at 150 ns ranges from about 10% to 25% of the original excimer intensity on cooling from room temperature to -196°C while the monomer intensity at 150 ns is about 22% of the original monomer emission intensity at a ll temperatures. A comparison of these observations with the spectra for this sample helps to explain why the excimer emission intensity decreases at t=150. Figure 4.12 shows that the monomer decay at room temperature of 91.7 umole/gm of pyrene is nonexponential. I t also shows that the lifetim e of the early, faster decaying component does not change with decreasing temperature until s lig h tly almost 25 ns. After 25 ns, the decays, with decreasing temperature, w ill tend to have smaller 124 1 l - » l\3 cn 6 0 0 R .T . + -5 0 C o -100 C 4 -196 0 7 7 K ) Figure 4.12 Monomer Decay for 91.7 umole/gm Pyrene on S24. T IM E (NS) -100 C B O O -02 - -0 6 - -08 - - I - -12 - -L4 - -L 6 - -L 8 - -2 - -22 - -2.4 - - 2& - 300 0 1 0 0 200 400 TIME t N S I RT. + -50 C o -100 C a -196 077 K > Figure 4.12a Excimer Decay fo r 91.7 umole/gm Pyrene on S24. lim iting slopes meaning that the monomer lifetim e is being extended. The curvature also tends to be greater with decreasing temperature. ( It is noteworthy that only the early, faster decaying component and the lim iting slope behave in an organized manner compared with the decays that are occurring between these two lim its.) Figure 4.12a is the excimer emission decay plot for this sample and i t shows a nearly exponential decay at room temperature. With cooling, there is a com m on decay until 25 to 30 ns followed by a gradual shifting to smaller lim iting slopes which means a longer lifetim e for the diffusive excimer. The monomer lifetim e can be obtained re la tiv e ly easily from the lim iting slope of the monomer emission decay. Usually more consistent lifetimes are obtained with smaller amounts of adsorbed pyrene. The kinetic model for the excimer formation proposed by SFL (12) that considers diffusional and static contributions could not be used because the monomer lifetimes of the various samples at the sam e temperature were not consistent. This result excludes determining the excimer lifetimes by the model. 4.2.2.2 Vibrational Analysis of Monomer Emission on S24 Table 4.4 summarizes the data of the vibrational band intensities for pyrene on S24. At 8.91 umole/gm pyrene, the data indicates that the pyrene molecules sense a more hydrophilic environment compared to pyrene in hexane. Interestingly, at t=0, the emitting pyrene molecules are seeing a rel a tiv e ly more hydrophobic surrounding than those at 127 Table 4.4 Monomer Vibrational Intensities (Pyrene on S24) Normalized Monomer Band Intensities W Im Sample I (374nm) II (384nm) I I I (395nm) (465nm)/(395nm) i 8.91 umole/gm I t=0, R.T. ! -50 C 1..00 I -100 C 1,00 ! -196 C 1.00 . t=150, R.T. 1.00 ! -50 C 1.00 i -100 C 1.00 -196 C ------ 0.77 0.60 0.50 0.66 0.59 0.42 1.15 1.09 1.11 1.12 1.12 0.94 Sample had no measurable excimer. '14.5 umole/gm lt= 0 , R.T. : -50 c i . oo I -100 C 1.00 j -196 C 1.00 t=150, R.T. 1.00 -50 C 1.00 -100 C 1.00 I -196 C 1.00 0.81 0.61 0.60 0.87 0.69 0.71 0.66 1.42 1.22 1.17 1.33 1.38 1.24 1.31 1.00 0.93 0.79 0.40 0.48 0.67 0.39 I 36.5 umole/gm 11=0, R.T. i -50 C 1.00 1 -100 C 1.00 | -196 C 1.00 |t=150, R.T. 1.00 ] -50 C 1.00 : -loo c i.oo -196 C 1.00 1.00 1.00 0.83 0.87 0.88 1.53 1.37 1.62 1.84 1.41 1.70 1.68 6.61 6.93 4.53 3.51 4.29 5.88 3.14 91.7 umole/gm t=0, R.T. | -50 C ! -loo c ! -196 C !t=150, R.T. -50 C -100 C -196 C A ll excimer; no measurable monomer. 1.00 2.20 6.50 All excimer; no measurable monomer. All excimer; no measurable monomer. 1.00 1.36 3.21 1.00 1.18 1.82 4.55 t=150. These band intensities can be compared with the SFL (12) work where it was concluded that interactions with silanols were responsible for the observed ratio. Such silanol interaction must not be very strong since there is ample evidence for surface m obility by pyrene. Studies by de Mayo (8,22) also show that adsorbed molecules such as pyrene tend to have very rapid motion on the surface (8). These band intensities can be compared with the band intensities in Table 4.2, derived from the work of Winnik and Dong (30), for a description of the polarity of the surface in terms of liquid phase standards. For example, S24 seems to be a f a ir ly polar surface, with a polarity sim ilar to t-amyl alcohol. 4.2.2.3 Pyrene Mobi 1 ity on S24 Table 4.4 also contains the ratio I e/ I m for the variably loaded S24. This ratio in solution measures the ease of excimer formation. O n surfaces, this ratio may represent factors, surface effects and pore effects, which affect the mobility of pyrene. This ratio neglects the effect of static (ground state) dimers since there is no objective manner in detecting them using this method. At best, this ratio can give a qualitative view of diffusion on the surface since a ll excimers do not come from a diffusive mechanism. The analysis that can be done seems to indicate a trend towards decreasing diffusion with decreasing temperature which is to be expected. The d iffic u lty is in assessing how m uch diffusion has been shut off. 1'29 4.2.2.4 Activation Energy of Excimer Formation on $24 Figures 4.13 and 4.14 are Arrhenius plots for 14.5 and 36.5 umole/gm of adsorbed pyrene. These were the only plots done because the emission spectra had measureable amounts of monomer and excimer emission. In each of these plots, at t=150 ns, there is a c ritic a l temperature at -100°C. In solution phase studies, the c ritic a l temperature of pyrene was seen at or above room temperature (26,32). At t=0 ns, the plot is more complex with two points of deviation at -50°C and at -100°C. The deviation at -100°C shifts from a small negative slope to a greater negative slope. This negative slope region corresponds to the "low" temperature behavior seen in solution phase studies. The deviation at -50°C appears to be a "true" c ritic a l temperature, assuming the definition from solution phase studies, and shows a positive slope which is consistent with "high" temperature behavior seen in solution phase studies. de Mayo (22) has estimated a surface activation energy of diffusion of 4 kilocalories per mole in studies of the quenching of pyrene emission by halonapthalenes on s ilic a surfaces. The binding energy of the pyrene excimer in solution is known to be 0.37 eV (8.53 kcal) and is characteristic of the molecular species but independent of the environment (26,32). The positive slope of the "high" temperature region is related to the binding energy but is c learly smaller than the solution value of 0.37 eV. From solution studies, i t is also known that structurally hindered excimers tend to have smaller binding energies (32,37). 130 0 2 01 - - 0 1 -02 - -02 - -0 .4 — - 0 5 - -0 .7 - -08 - -09 - 13 U 9 7 5 l/T X W O O ( K ) □ a t t- 0 + « t t-1 5 0 S lm o ltt* d Figure 4.13 Arrhenius Plot for 14.5 umole/gm Pyrene on S24. (ja« 2 L 9 18 17 L6 L5 U 13 L2 U 1 3 U 9 7 5 3 l/ T X W OO (K) □ l i t t- 0 + « t t-1 5 0 S im u k fe d Figure 4.14 Arrhenius Plot for 36.5 umole/gm Pyrene on S24. C O no An analysis of the emission spectra of these samples and these Arrhenius plots shows that the re la tiv e excimer emission does s lig h tly increase to a maximum at -100° C, but decreases between -100° C and - 196° C. 4.2.3.0 Variable Temperature Time Resolved Spectra of Pyrene on S72 The most striking feature of this series is that the emission spectra at t=0 show the excimer emission to be the dominant emission, but at t=150, monomer emission becomes the dominant emission. This observation is true for a ll temperatures and is similar to how pyrene behaves on S24 except the degree of change is greater on S72 than on S24. The second most striking feature is that as the temperature decreases, the excimer emission stays re la tiv e ly constant down to -196°C. The monomer and excimer peaks emit at the same energies as for the S24 series. As the samples are cooled, the excimer emission intensity increases very s lig h tly to a maximum between 0°C and -50°C, but decreases to an intensity s lig h tly less than the maximum and stays re la tiv e ly constant with further cooling to -196°C. A 10 umole/gm sample, Figure 4.15, shows in it i a ll y equivalent intensities for monomer and excimer emission at t=0. At t=150 ns, Figure 4.15a, the emission is predominantly monomer emission with a smaller contribution from the excimer emission. A comparison of the spectra at t=0 with the spectra at t=150 shows that cooling has a more dramatic effect at t=0 than at t=150. The t=0 spectra show an almost imperceptible increase in excimer emission to -50° C followed by a 133 Figure 4.15 Variable Temperature Spectra of Pyrene on S72 at t=0. 10 umole/gm 134 n m 460 1 500 60 0 RELATIVE INTENSITY (ARBITRARY UNITS) Figure 4.15a Variable Temperature Spectra of Pyrene on S72 at t=150. 10 umole/gm n m 460 560 RELATIVE INTENSITY (ARBITRARY UNITS) r more apparent decrease in this emission through -196°C. However, the t=150 spectra show almost constant excimer emission intensity at a ll temperatures. O n a 20 umole/gm sample of adsorbed pyrene, Figures 4.16 and 4.16a, shows spectral behavior similar to the 10 umole/gm sample at a ll temperatures. This sample displays a very dominant excimer emission at t=0 but there is a dramatic increase of monomer emission at t=150. Again, on cooling the sample, there is an almost imperceptible increase in the excimer emission intensity to 0°C followed by l i t t l e change with further cooling to -196°C. Figures 4.17 and 4.17a are the spectra of the sample with 60 umole/gm of adsorbed pyrene. Spectral changes are not evident because the amount of pyrene on this surface results in an overwhelming amount of excimer emission. There are no visible changes from t=0 to t=150 and the only temperature effect seems to be a gradual sharpening of the excimer emission envelope. An important difference between S24 and S72 is that S24 is an amorphous substance while S72 has been made to be f a ir ly regular, at least more regular than S24. In terms of important interactions, silanols provide most of the interaction and as stated in Chapter 1, there are many different kinds of silanols with different properties. "Reactive" silanols are very important as a surface species because of their great reactivity and adsorptivity. I t is known that less than 5% of the silanols are the "reactive" type with surfaces having surface areas under 500 m^/gm (16). S24 has a surface area of 415 m^/gm while 138 Figure 4.16 Variable Temperature Spectra of Pyrene on S72 at t=0. 20 umole/gm RELATIVE INTENSITY (ARBITRARY ONTTS) i i H w g o Figure 4.16a i Variable Temperature Spectra of Pyrene on S72 at t=150. 20 umole/gm 141 RELATIVE INTENSITY (ARBITRARY UNITS) Figure 4. Variable Temperature Spectra of Pyrene on S72 at t=0. 60 umole/gm 143 RELATIVE INTENSITY (ARBITRARY UNITS) Figure 4.17a Variable Temperature Spectra of Pyrene on S72 at t=150. 60 umole/gm RELATIVE INTENSITY (ARBITRARY UNITS) Q . ; o C D a S72 has a surface area of 398 m^/gm. Although S72 has a smaller surface area, it has greater ordering and greater volume in smaller pores. Thus, S72 would be more lik e ly to have "reactive" silanols than S24, especially in the pores since these kinds of silanols are formed by close lying silanols. This might be the environment that readily adsorbs pyrene; e.g., the pores would be the preferential sites of adsorption because of these "reactive" hydroxyls. 4.2.3.X Decay Measurements of Pyrene on S72 The emission decay characteristics of monomer and excimer are different compared with the S24 decays. The S72 and S24 decays are sim ilar in that both show nonexponential behavior with a com m on faster decay component at a ll temperatures and a change to a smaller lim iting slope for the monomer decays with the decreasing temperature. The S72 excimer decays are completely different from the S24 excimer decays. Figure 4.18 shows nonexponential decay for the monomer emission of 10 umole/gm of adsorbed pyrene. The com m on faster decay component deviates at about 100 ns to the longer decaying component. At 20 umole/gm and 60 umole/gm of adsorbed pyrene, the decays, Figures 4.19 and 4.20, are nonexponential and the point of deviation occurs between 50 and 100 ns. The monomer lifetim es, determined from the lim iting slope, are shown in Table 4.5 for a ll of the S72 samples and at the various temperatures. Figure 4.18a, 10 umole/gm, is the excimer decay and shows nonexponential decay at a ll temperatures. The faster decay component 147 Table 4.5 Pyrene Monomer Lifetim es of V ariab le Concentrations at V ariab le Temperatures (S72) Sample (umole/gm) Monomer Lifetime (ns) Temperature (C) 10 220.3 R.T. 339.8 -50 353.2 -100 366.7 -196 20 189.4 R.T. 258.9 -50 371.1 -100 380.1 -196 50 109.6 R.T. 229.0 -50 238.2 -100 270.3 -196 i i t 148 \ + * □ & + O □ A+ O ^ 0 *+ O 0 A « - Do * * 0 ° 0 * 0 □ +4 o □ + £ □ ^ ° 4- O u +4 0 □ + O 0 A 4+0 ^ U % □ -hu 0 □ □ □ □ □ T 0 200 400 600 800 T IM E (NS) R.T. + -50 C o 400 C & -196 077 K > Figure 4.18 Monomer Decay for 10 umole/gm Pyrene on S72. -0 5 - -1 - -15 - -2 - -2 5 - -3 - 0 200 280 4 0 80 2 4 0 1 2 0 I S O TIME ( N S I R.T. + -50 C o -WOC A -196 077 K ) Figure 4.18a Excimer Decay for 10 umole/gm Pyrene on S72. has a com m on decay at a ll temperatures but more interesting is that the longer decaying component also has its ow n com m on decay at temperatures below room temperature. This pattern is also seen in Figure 4.19a, 20 umole/gm of adsorbed pyrene. However, in Figure 4.20a, 60 umole/gm of adsorbed pyrene, the pattern begins to behave more like the excimer decays seen for the S24 samples. The point of deviation from the faster decaying component is at 40 ns for a ll of the samples at a ll the temperatures. A correlation of the spectra of S72 samples and these decays explain why the excimer emission intensity stays re la tiv e ly constant. The faster component of the excimer decays may be attributed to static dimer decay, while the longer decaying component may involve som e degree of diffusion. However, a true diffusive excimer should show a longer lifetim e with decreasing temperature which is not observed. I t may be an excimer formed with a re la tiv e ly small activation energy such as short range diffusion or just rotation for formation. 4.2.3.2 Vibrational and Mobi 1 ity Analysis of Pyrene on S72 The vibrational band intensities of the monomer emission and I e/ I m ratios are summarized in Table 4.6 for the various amounts of adsorbed pyrene on S72. A comparison of the I I I / I ratio with the standard solution values of Table 4.2 show that the S72 surface has a polarity similar to n-butyl ether. This surface seems to be s lig h tly more hydrophobic than S24 which had a polarity similar to t-amyl alcohol. 151 Table 4.6 Monomer V ib ratio n a l In te n s itie s (Pyrene on S72) Normalized Monomer Band Intensities W 1™ Sample I (374nm) II (384nm) I I I (395nm) (465nm)/(395nm) 10 umole/gm t=0, R.T. -50 C 1.00 0.79 1.31 0.87 -100 C 1.00 0.67 1.30 0.81 -196 C 1.00 0.75 1.35 0.58 150, R.T. 1.00 0.72 1.30 0.24 -50 C 1.00 0.66 1.29 0.37 -100 C 1.00 0.68 1.31 0.32 -196 C 1.00 0.63 1.27 0.25 20 umole/gm t=0, R.T. -50 C 1.00 1.00 1.56 2.47 -100 C 1.00 0.83 1.52 2.35 -196 C 1.00 0.85 1.63 1.54 150, R.T. 1.00 0.88 1.69 0.80 -50 C 1.00 0.84 1.57 0.95 -100 C 1.00 0.86 1.53 0.97 -196 C 1.00 0.73 1.44 0.92 60 umole/gm 0, R.T. -50 C -100 C A 1 1 excimer: no measurable monomer. -196 C 150, R.T. -50 C 1.00 1.40 1.90 17.05 -100 C 1.00 1.60 1.92 17.40 -196 C 1.00 1.16 2.09 7.58 152 LN I R ELA TTVE 0 >02 -0 4 -0 6 -0 8 - I -L2 -L 4 -L 6 -1 8 -2 -22 -2 .4 -2£ -22 -3 0 2 0 0 4 0 0 T IM E (N S □ R .T . + -5 0 C O 4 0 0 C a -196 0 7 7 K> Figure 4.19 Monomer Decay for 20 umole/gm Pyrene on S72. c n co LN IRELATW E 0 -02 -0 4 - 0 6 -0 8 -1 - 1 2 -L 4 -L 6 -1 8 -2 -22 -2 .4 -2 6 -2 8 -3 0 4 0 8 0 120 160 2 0 0 2 4 0 2 8 0 T IM E (N S □ R .T . + -5 0 C o -100 C A -196 0 7 7 K> Figure 4.19a Excimer Decay for 20 umole/gm Pyrene on S72. C J 1 X □ & * D a* □ A O □ + ° A 4. 0 A 0 •— □ --------- ! ----O- LNIRELATTVE £ ---------------------------------------------------------------------- -0 2 - ah -0 .4 - &<&- - 0 6 - * □ -0 8 - B < 1 - I - l> □ + O -1 2 - £ □ <*. & □ <*. -1 4 - □ □ 4 - 0 , -1 6 - eP + o -1 8 - 0 □ * 0 + 0 + 0 - 2 - -2 2 - ff □ □ > l> ♦ + 0 + -2 .4 - □ □ -2 £ - 0 A □ -2 & - □ -3 - □ -3 .2 - --------------- . ---------------- r ----- ----------- -------- 1 1 ■ .....---------- p— □ A u 200 4 0 0 T IM E (N S 0 > 50 C R .T . + 0 C 0 > 50 C 4 -9 0 C Figure 4.20 Monomer Decay for 60 umole/gm Pyrene on S72. en cn □ &>. „ ± 0 -0J5 -j □ M - $ D M * □ + ^ o □ + 0 A + 0 A D + 0 4 + O A -15 -J □ + o -2 5 - A + ° A □ + + O A A A + ° A □ + + O + 0 ^ □ O + + □ ° -3 H □ + □ □ .35 - | ----------- , ----------- ! ------------1 ------------1 ------------1 ------------1 ------------1 — --------- 1 ------------1 ----------- 1 —0 ------1 ------------1 — 0 40 80 1 2 0 1 6 0 300 240 TIME IN S ) R T . + -5 0 C o 4 0 0 C A -196 0 7 7 K ) Figure 4.20a Excimer Decay for 60 umole/gm Pyrene on S72. -i- At t=0 the I e/ I m ratios are greater than the ratios at t=150. This can be best explained by an excimer which is short lived, the static dimer. Within each time frame, however, the I e/ I m ratios stay re la tiv e ly constant; trending towards a slight decrease with decreasing temperature. This temperature independence means that this excimer emission does not result from long-range diffusion but may result from neighboring pyrenes such as those adsorbed into a pore. 4.2.3.3 Activation Energy of Excimer Formation on S72 The Arrhenius plots, Figures 4.21 and 4.22, show behavior similar to the Arrhenius plots of S24. In the early time, t=0, for 10 umole/gm, the plot is a linear negative slope indicating som e type of diffusional processes. This plot does not show a c ritic a l temperature compared with the plot at t=150. The plot at t=150 displays behavior similar to the behavior seen with the S24 plots. A small in itia l negative slope yields to a greater negative slope at -100°C, with the c ritic a l temperature at -50°C followed by a positive slope which indicates "high" temperature behavior. In Figure 4.22, 20 umole/gm, the t=0 plot shows the small negative slope yielding to a greater negative slope which then appears to yield to a smaller negative slope or i t may be near the c ritic a l temperature and attempting to turn down. At t=150, the plot shows a very small negative slope reaching a c ritic a l temperature of -50°C which then converts into a "high" temperature plot. A comparison of the slopes encountered in these plots with the slopes determined for 157 -02 -03 -05 -0 6 -07 -08 -09 1 -U -12 -13 -L4 -L5 1 3 3 U 5 7 9 l/ T X 1000 <K) □ © t-0 ® M 5 0 Figure 4.21 Arrhenius Plot for 10 umole/gm Pyrene on S72. 0 9 - O S - 07 - 06 - 05 - 03 - 02 - C U - -01 - -02 - -03 13 11 9 7 l/ T X W O O (K ) □ e t-0 + s H 5 0 Figure 4.22 Arrhenius Plot for 20 umole/gm Pyrene on S72. S24 show a remarkable sim ilarity in terms of behavior and values. The S72 plots indicate a surface activation energy to diffusion m uch less than de Mayo's 4 kcal per mole. These plots also show the c ritic a l temperature to be m uch below room temperature when compared with the solution studies of pyrene, e.g. i t is almost as if the pyrene solution study was shifted in temperature but keeping essentially the same shape. The value of the binding energy seems very low again as was seen with the S24 samples which again may re fle c t a hindered form of the pyrene excimer (26,32,37). 4.2.4.0 Variable Temperature Time Resolved Spectra of Pyrene on S63 Figures 4.23 to 4.26 are the spectra of various amounts of pyrene adsorbed on S63. Monomer and excimer emission peaks are at the sam e energies as with the S24 and S72 series. This series of spectra shows very different results than those obtained on the other surfaces. Excimer intensity again seems greater at t=0 compared with t=150 spectra. The most striking feature is that as the temperature is decreased, the re la tiv e excimer emission intensity increases dramatically with a maximum achieved at about -50°C. With further cooling, the excimer emission s lig h tly drops in intensity. This feature occurs in both time frames and with a ll loadings. O n the two lower loadings, from -100°C down to -196°C, the monomer emission becomes very well resolved. S63 has a very high surface area, 928 m^/gm, and as a result is predicted to have a large population of "reactive" silanols 160 (16,17,19,20). Furthermore, in terms of porosity, this surface has the most defined range of pores. The majority of the pore volume is found in pores of four angstroms radius by a cylindrical model or separated by four angstroms in a p a ralle l plate model. Since most of the pore volume is located in small pores, then m uch of the surface area can be expected to be represented in small pores. Because of the high surface area, comparable amounts of pyrene adsorbed onto S63 result in less excimer emission than on S24 and S72. The spectrum for 6.0 umole/gm adsorbed pyrene, Figure 4.23, show a very dominant monomer emission with a very small contribution from excimer emission at t=0, room temperature. The spectrum at t=150, Figure 4.23a, show no evidence of the excimer emission except for a small "ta il" at the end of the longer wavelength part of the monomer emission. Cooling this sample results in a dramatic increase in the excimer emission which reaches a maximum intensity at -50°C for the t=0 spectrum. At -100°C, the excimer emission intensity seems to decrease s lig h tly and does not change with further cooling to -196°C for the t=0 spectra. At t=150, cooling causes a similar dramatic increase in the excimer emission intensity which reaches a maximum at -100 C. This maximum stays constant with cooling to -125°C, but decreases at -150°C and does not change with cooling to -196°C. The monomer resolution becomes very sharp at -196°C. Winnik and Dong (30) attribute an increase in intensity of bands I and I I with decreasing temperature, which is observed in these spectra, to an increase in polarity of the environment and additional crystal forces in the frozen solvent matrix (26,29,30). Figure 4. Variable Temperature Spectra of Pyrene on S63 at t=0. 6 umole/gm 0 0 9 » O O S , OOfr U iU RELATIVE INTENSITY (ARBITRARY UNITS) Figure 4.23a Variable Temperature Spectra of Pyrene on S63 at t=150. 6 umole/gm 164 RELATIVE INTENSITY (ARBITRARY UNITS) The spectra, Figures 4.24 and 4.24a, of 11.6 umole/gm of adsorbed pyrene show the sam e trend except that the maximum intensity of the excimer emission is reached at -50°C for both time frames. Figures 4.25 and 4.25a of 49.8 umole/gm of adsorbed pyrene show equivalent amounts of monomer and excimer emission for the t=0 spectra at room temperature. At t=150, monomer emission is the dominant emission. ( It is interesting in these spectra at t=0 and at t=150 that band I I I of the monomer emission seems to be enhanced in intensity compared with the spectra of sim ilar monomer/excimer emission ratios with the samples of lower loadings; e.g., the room temperature and 0°C spectra at t=0 fo r 11.6 umole/gm. A simple interpretation might be that this enhancement is caused by contribution from the excimer but Yamazaki, Tamai and Yamazaki (33) were able to assign a specific configuration of the excimer as the cause fo r this enhancement.) Cooling to 0°C causes an increase in excimer emission intensity so that the spectrum at t=0 shows dominant excimer emission with a small contribution from monomer emission. The t=150 spectrum shows an increase in excimer emission but monmer emission is s t i l l dominating. At -50°C, the excimer emission is dominant in both time frames and at -100°C, the excimer emission intensity has s lig h tly decreased and remains constant with cooling to -196°C. Figures 4.26 and 4.26a, 60 umole/gm of adsorbed pyrene, show spectra dominated by excimer emission at a ll temperatures. The only discernible changes with decreasing temperature is an apparent increase of excimer emission intensity which reaches a maximum at -100°C, followed by a s lig h t decrease in excimer emission 166 Figure 4.24 Variable Temperature Spectra of 11.6 umole/gm Pyrene on S63 at t=0. 167 RELATIVE INTENSITY (ARBITRARY UNITS) o o Figure 4.24a V ariab le Temperature Spectra of 11.6 umole/gm Pyrene on S63 at t=150. RELATIVE INTENSITY (ARBITRARY UNITS) Figure 4.25 Variable Temperature Spectra of 49.8 umole/gm Pyrene on S63 at t=CL 171 RELATIVE INTENSITY (ARBITRARY UNITS) Figure ,25a Variable Temperature Spectra of 49.8 umole/gm Pyrene on S63 at t=150. 173J 00i7 UJU (ARBITRARY UNITS) RELATIVE INTENSITY O- Figure 4.26 V ariab le Temperature Spectra of Pyrene on S63 at t= 0. 60 umole/gm RELATIVE INTENSITY (ARBITRARY UNITS) o o- o Figure 4.26a V a ria b le Temperature Spectra of Pyrene on S63 at t=150. 60 umole/gm RELATIVE INTENSITY (ARBITRARY UNITS) O O - intensity with cooling to -196 C. The excimer emission envelope also seemed to become sharpest at -100°C. Yamazaki, Tamai and Yamazaki (33) report the picosecond time- resolved spectra of Langmuir-Blodgett film s of pyrene and analyse the emission spectra. They were able to show the evolution of the monomer/excimer emission from picoseconds to hundreds of nanoseconds and were able to separate out d istin ct emitting species which were assigned to d iffere n t configurations of pyrene dimers. They conclude the enhancement of band I I I in the monomer emission is caused by emission from a configuration of pyrene dimer which is not the normal "sandwich" geometry. The emission from the normal "sandwich" configuration excimer is reported to be e s s e n tia lly separated from the monomer. 4.2.4.1 Decay Measurements of Pyrene on S63 Table 4.7 summarizes the lifetim es of the monomer at various loadings on S63. The monomer decays are sim ilar to the monomer decays of S24 and S72. The decays are nonexponential with a fast component which is com m on to a ll temperatures followed by a longer decay component which changes to sm aller lim itin g slopes as the temperature decreases. However, the excimer decays are very d iffere n t from the S24 and S72 decays. These decays are nonexponential and composed of fast and longer decaying components. The faster decaying component shares a com m on decay at a ll temperatures. The longer decaying component is dependent on the loading in its response to decreasing temperature. 179 Table 4.7 Pyrene Monomer Lifetim es of V ariab le Concentrations at V ariab le Temperatures (S63) Sample (umole/gm) Monomer Lifetim e (ns) Temperature (C) 6 187.4 R.T. 256.9 -50 333.7 -100 408.5 -196 11.6 180.6 R.T. 262.8 -50 275.9 -100 369.5 -196 60 76.5 R.T. 126.2 -50 87.1 -100 96.6 -196 180 The trend seems to be that a lower loaded sample cooled to -50°C shows an increase in the lim itin g slope fo r the excimer decay and further cooling results in a return of "normal" behavior, e.g. the lim itin g slope becomes smaller with decreasing temperature. For a sample with greater amounts of adsorbed pyrene, the in it ia l increase in the lim itin g slope is less and, eventually, on a sample with a large amount of adsorbed pyrene, 60 umole/gm, the decay behaves "normally" even at -50°C. Figure 4.27, 6.0 umole/gm of adsorbed pyrene, shows the monomer decay at various temperatures which is nonexponential in form. The behavior of the monomer decay is sim ilar to those noted on S24 and S72. The faster decaying component is com m on to a ll temperatures and is followed by sm aller lim itin g slopes with decreasing temperature. The sm aller lim itin g slope means a longer life tim e fo r the monomer. The point of changeover from faster to slower decay occurs at 100 ns. Figure 4.27a is the excimer decay which shows nonexponential decay with fast and slow decaying components. The fa st component again has a com m on decay at a ll temperatures, as was seen with S24 and S72 excimer decays, u n til 100 ns where the slower decaying component begins. Highly unusual behavior is seen on cooling to -50 and -100°C. There seems to be a hint of a risetim e, which may be artefactual in both of these decays. The most striking feature is that the lim itin g slope increases which is contrary to what was seen on S24, S72 and in solution studies. I t is only at -196°C that the lim itin g slope becomes 181 O H *--------------------------------------------------------------------------------------------------------------------------------------- ------- -02 +A& o “04 - + An O flfi + 3 ° O - 0 6 - + & o + o -0 8 - + f , o + “ a « - 1 - * P & O ^ A i . -M “ □ + a * o - L 6 - + a 0 □ + o - U 8 - + a o _2 - Q + £° ca -23 - + o A □ -2.4 - + o a □ -2£ - 0 + + - 2 M - □ + -3 -|--------------- , --------------- , --------------- , --------------- , --------------- , --------------- , --------------- ! --------------- 0 2 0 0 4 0 0 6 0 0 8 0 0 T IM E (N S H .T . + -5 0 C O -100 C A -196 0 7 7 KJ Figure 4.27 Monomer Decay for 6 umole/gm Pyrene on S63. % + A O + □ + O *0 4 4 o A A A O C A -as - - l - -L 5 - -2 5 -3 - 4 0 0 0 R .T . T IM E (N S + -5 0 C o 4 0 0 C a 4 9 6 0 7 7 K » Figure 4.27a Excimer Decay for 6 umole/gm Pyrene on S63. sm aller. The -50°C decay never shows a return to sm aller lim itin g slope, but the -100°C decay follow s the -196°C decay and deviates from i t to show an increased lim itin g slope near 200 ns. Figure 4.28, 11.6 umole/gm of adsorbed pyrene, shows nonexponential decay with a faster decaying component which shares a com m on decay at a ll temperatures u n til s lig h tly less than 200 ns. Following the faster decaying component is a slower decaying component which begins to have a sm aller slope with decreasing temperature. This pattern has been seen on S24 and S72. Figure 4.28a is the excimer decay and i t shows nonexponential decay with a com m on unresolved fast decaying component at a ll temperatures. At about 25 ns, a longer decaying component follow s the fast component. O n cooling to -50°C, the lim itin g slope increases again as was observed fo r the lower loaded excimer decay at -50°C. The amount of increase seems to be less. At -100°C, the decay more or less p a ra lle ls the room temperature decay. The -196°C decay shows a sm aller lim itin g slope as expected. Figure 4.29, 60 umole/gm of adsorbed pyrene, shows nonexponential monomer decay at al 1 temperatures. However, the fast component no longer shares a com m on decay with decreasing temperature. Instead, the fast component begins to show greater curvature with decreasing temperature which was also seen on 91.7 umole/gm of adsorbed pyrene on S24. The lim itin g slope does seem to decrease with decreasing temperature although the monomer intensity is very low, making the determination of its life tim e d iffic u lt. Figure 4.29a is the excimer decay and i t shows non-exponential decay. With decreasing temperature, 184 L M R E LA TIV E -02 - - 0 6 - - 0 8 - - I - -L2 - -L 4 - -1 6 - -L 8 - -2 - -22 - - 2 6 - -2 6 - 1 0 08 02 0 4 0 6 R T . (Thousand*) T IM E (NSI + -5 0 C o - K » C & -196 0 7 7 K ) Figure 4.28 Monomer Decay for 11.6 umole/gm Pyrene on S63. 0 0 cn ST c*#> + □ <** + □<>* + ^ & + □ + O A + □ 0 4- 0 4 4- □ O 4- O □ 4- & o □ o £ i 0 WO 2 0 0 3 0 0 4 0 0 T IM E <NSI R .T . + -5 0 C o -100 C d -196 C i7 7 K > Figure 4.28a Excimer Decay for 11.6 umole/gm Pyrene on S63. - 0 5 - -I - -L 5 -2 - -2 5 - -3 - 0 4 0 80 1 2 0 IS O 200 2 4 0 T IM E (N S R .T . + -5 0 C o -4 0 0 C A -196 0 7 7 K ) Figure 4.29 Monomer Decay for 60 umole/gm Pyrene on S63. L Nf R E LA TIV E 0 -02 -0 8 -L2 -L 4 -L6 -L 8 -22 + o -2£ 0 4 0 1 20 200 8 0 150 T IM E (NS) D & T . + -5 0 C o 4 0 0 C A -196 0 7 7 1 0 Figure 4.29a Excimer Decay for 60 umole/gm Pyrene on S63. 0 0 00 the decays seem to share a com m on decay, sim ilar to the decays of S72. There is a fast decaying unresolved component, under 20 ns, which is followed by the longer decaying component. A correlation of the spectra of these samples with th e ir decays show why the excimer emission intensity decreases from t=0 to t=150. For example, the excimer emission fo r 11.6 umole/gm of adsorbed pyrene has a range of intensity from 14% to 27% of the original excimer emission intensity at the various temperatures. The monomer emission intensity stays constant at 47% of the original monomer emission intensity at t=150. 4 .2 .4 .2 Vibrational and M o b ility Analysis of Pyrene on S63 Table 4.8 summarizes the vibrational and I e/ I m data. The vibrational band intensities of the monomer emission show a 1.22 ra tio (band Ill/b a n d I) which is close in p o la rity to n-butyl ether when compared to the standard solution values of Table 4.2. I t is interesting that the two surfaces with more lim ited pore ranges seem somewhat more hydrophobic than the amorphous S24. M o b ility on this surface seems to be lim ited. The I e/ I m ra tio reaches a maximum at -50°C and then decreases from there. Cooling should decrease diffusion over great distances, such as interpore diffusion, but short distance diffusion or rotation, such as intrapore transport, may be a v a ila b le to the pyrene molecules. 4.2.4.3 Activation Energy of Excimer Formation on S63 Arrhenius plots of pyrene on S63 show very sim ilar behavior to 189 Table 4.8 Monomer Vibrational Intensities (Pyrene on S63) Normalized Monomer Band Intensities Sample I (374nm) I I (384nm ) I I I (395nm) (465nmV(395nm) 6 umole/gm C + II O 70 • — 1 • -50 C 1.00 0.72 1.27 0.15 -100 C 1.00 0.64 1.35 0.44 -196 C 1.00 0.63 1.22 0.27 t=150, R.T. 1.00 0.72 1.39 ------- -50 C 1.00 0.63 1.28 0.21 -100 C 1.00 0.63 1.35 0.22 -196 C 1.00 0.62 1.28 0.14 11.6 umole/gm t= 0 , R.T. -50 C 1.00 0.64 1.33 0.22: -100 C 1.00 0.69 1.28 1.00 -196 C 1.00 0.65 1.26 0.57 t=150, R.T. 1.00 0.78 1.33 0.04 -50 C 1.00 0.65 1.35 0.46 -100 C 1.00 0.67 1.39 0.34 -196 C 1.00 0.66 1.13 0.29 60 umole/gm t= 0 , R.T. -50 C -100 C A ll excimer; no measurable monomer. -196 C t=150, R.T. -50 C -100 C -196 C Arrhenius plots done for pyrene adsorbed onto S24 and S72. Figures 4.30 to 4.32 a ll show a small negative slope yieldin g to a greater negative slope, from low to high temperature, which eventually reaches a c ritic a l temperature and turns into a positive slope. These plots are very sim ilar in shape compared with solution Arrhenius plots of pyrene. The negative slope represents a diffusional process ty p ic a lly c alled "low" temperature behavior. The positive slope represents the binding energy of the excimer typical ly known as "high" temperature behavior. The greater negative slope at higher temperatures is not seen in solution studies but i t may represent som e kind of diffusional process since its shape and slope are characteristic of "low" temperature behavior (26,32). Its greater slope may indicate som e kind of diffusional b arrier which is greater than the more usual diffusional barrier. The diffusional part of the p lo t, represented by the small negative slope, is considerably less than the slope of 4 k ilo c a lo ries per mole that de Mayo (22) reports in his surface study. The binding energy of the excimer, represented by the positive slope, is a factor of 2 to 3 sm aller than the reported value of 0.37 eV (8.53 k ilo c a lo rie s ) fo r solution work (26,32). The c ritic a l temperature occurs at -50°C compared with the solution c ritic a l temperature at or s lig h tly above room temperature (26). 4.3 Discussion This study shows the importance of surface characteristics in the 191 192 -08 - 0 9 - - 1 - - U - -L 3 - - U - -IS - -L 6 - -L 7 - -1 8 - -1 9 - 3 5 7 l / T X M M O (K) □ © t-0 + e M 5 0 Figure 4.30 Arrhenius Plot for 6 umole/gm Pyrene on S63. 0 2 / -02 - - 0 6 - - 0 8 - - 1 - - 1 2 - -1 4 - -L 6 - -L 8 - -2 - -22 - -2 .4 - -2JB - - 3 - -3 2 13 3 5 7 U 9 1 /T X tO O O (K ) □ © t-0 + o t-1 5 0 Figure 4.31 Arrhenius Plot for 11.6 umole/gm Pyrene on S63. < D C O L4 - 1 2 - 08 - 06 - 02 - -04 - -0 6 - -0 8 - -12 -L4 - -1 6 - -L 8 - 1 3 U 9 7 1 /T X 1 0 0 0 (K) □ o t- 0 + o t-1 5 0 Figure 4.32 Arrhenius Plot for 60 umole/gm on S63. i — ^ VO photophysical behavior of surface adsorbed pyrene. Two apparently important properties of the s ilic a surface are the porosity and surface area. An additional important factor influencing the photophysical behavior of pyrene is the method of adsorption of the probe and the possible role of microcrystals on the surface. The surface area and porosity determine many of the properties of the emission spectra because they influence the dispersion of pyrene on the s ilic a . S24 and S72 have comparable surface areas, 415 and 398 m^/gm, respectively, compared with S63 which has a surface area of 927 m^/gm. The effect of this la tte r greater surface area is a greater dispersion of the probe. This can be seen by comparing the spectra of 14.5 and 10 umole/gm of adsorbed pyrene on S24 and S72 respectively. The spectra from these two silicas are very sim ilar and show near equivalent emission intensities from the monomer and excimer. A similar loading of 11.6 umole/gm of adsorbed pyrene on S63, however, shows the emission spectra dominated by monomer emission with a small contribution from the excimer emission. Som e of the surface area of these porous silicas is located in the pores. The amount of contribution to the surface area by pores can be correlated with the amount of pore volume associated with a specific pore size. Although S24 is described as an amorphous s ilic a , the pore distribution shows 13% of the volume located in pores less than 8 A in radius (assuming a cylindrical pore model), 38% of the pore volume located in a range of pores 12 to 32 A in radius and 48% in pores greater than 32 A radius. S72 shows 20% of the pore volume 195 located in a range of pores less than 14 A radius, 22% of the pore volume located in pores 14 to 32 A radius and 59% of the pore volume in pores greater than 32 A radius. S63 shows 83% of the pore volume in pores less than 6 A radius, 14% of the pore volume in pores 6 to 18 A radius and 3% of the pore volume in pores greater than 18 A radius. Pyrene has molecular dimensions of 7.2 A by 13.0 A. These dimensions are important to adsorption into a pore. Moldenhauer-Kopach (38) has shown the adsorption of various polyaromatics onto the sam e series of silicas depends on the dimensions of the polyaromatic re la tiv e to the size of the pore. Pores similar or larger than the dimensions of pyrene are more lik e ly to accomodate pyrene. I t is important to know where most of the surface area is represented because this knowledge w ill give a better idea of how the pyrenes are being accomodated on the surface. An important property associated with the porosity and the interaction with pyrene is the type of silanol which is present. Silanols have been shown to be sites of adsorption for aromatic hydrocarbons (16-21). Synder and Ward (39) have characterized the different types of silanols on s ilic a surfaces and one type, the "reactive" silan o l, is very important because it is reported to have enhanced reactivity and a greater adsorption potential. These "reactive" silanols result when silanols are close enough to each other to form f u ll or partial hydrogen bonds. The adsorptive strength of a surface is a function of the concentration of silanols. W hen these silanols are close to each other, the effective concentration 196 increases and the a v a ila b ility to bond to one pyrene increases. The distribution of these "reactive" silanols is not restricted to any particular domain on the surface, but the likelihood of finding these silanols in the pore is m uch greater than on the surface because of the dimensional constraints of the pore. This makes the "reactive" silanol a very important factor for adsorption into a pore, especially in pores which are about the sam e dimensions of pyrene. Further, it seems lik e ly that m obility w ill be restricted with adsorption at a "reactive" silanol because of the enhanced adsorption potential. The method of adsorption employed in this study differs from the method employed by most other studies (1-11,13-15). For example, de Mayo and co-workers use a flash adsorption method (3,8,16). A comparison study was done by Moldenhauer-Kopach (38) in which the method of flash adsorption by de Mayo (3) was contrasted with the method employed here. A scanning electromicroscopy investigation (SEM) revealed the existence of crystalloid material on surfaces after flash adsorption, especially on S63. (The resolution of the micrographs showed only "gross" features, e.g. pores were not seen for any of the surfaces.) I t was proposed that S63 showed many more crystalloids than S24 or S72 because it had many more fine pores which did not allow pyrene to enter and thus the pyrene crystallized out on the surface. Crystalloid material was found to a lesser degree on the other surfaces presumably because many of the larger pores could more easily accommodate pyrene as microcrystals. The method of adsorption used in this study involved adsorbing 197 pyrene from solution onto the silicas by stirring the s ilic a in a pyrene solution. After the solvent was removed, washings with fresh solvent were done to remove any crystals of pyrene which may have been deposited. The samples prepared in this manner showed no evidence of crystalloid material with SEM. While i t seems lik e ly that most of the conclusions reached in other studies about static dimers are correct, their observations are clouded by the possibility of microcrystal formation. This problem should be much less with the method of adsorption used in this work which is similar to the method of making Langmuir-Blodgett film s. Yamazaki, Tamai and Yamazaki (33) were interested in the picosecond time resolved spectra of pyrene films of this kind. The method involves repeated dippings into a solution of pyrene and creating very organized multilayered film s, id e a lly one monolayer on top of another. I t is thought that these monolayers are analogous to those monolayers which are seen with ca p illary condensation e.g. the BET method and porosity distributions (17,19,40). Therefore, an adsorbing pyrene from solution is assumed to adsorb into a pore and strongly interact with a silanol. The next adsorbing pyrene may adsorb in the next layer above it , creating the lik e ly configuration for static dimer formation. Washings with fresh solvent should eliminate m uch of the deposited microcrystals and any residual pyrenes which are not strongly interacting with the surface. O n a ll of the s ilic as , adsorption of pyrene led to static dimers which could be detected by the analysis of the excimer decay. I t was 198 observed that the decay was not a simple exponential; an unresolved fast decay was seen followed by a longer decay component. I f excimers were a ll d iffu s iv e ly formed, a risetime should have been observed as other studies have shown (26-28). No such risetime is seen and the decay seems consistent with the existence of static dimers as o rig in a lly suggested by Singer, Francis and Lin (28). Diffusive excimers should show longer lifetimes with decreasing temperatures as the monomer lifetim es are increased and diffusion is slowed. The emission also seem to show evidence for static dimer contributing to the emission. O n som e samples, 20 umole/gm and 49.8 umole/gm of adsorbed pyrene on S72 and S63 respectively, which showed predominantly excimer emission at t=0 but showed monomer emission dominating at t=150, band I I I of the monomer emission seems to have been more enhanced than expected. Yamazaki, Tamai and Yamazaki (33) report an excimer band showing major contributions at this energy which may enhance band I I I intensity. This band has been assigned to static dimer. They also report another excimer band which has its maximum emission at this energy. This band is proposed to be dimers of "unknown configuration(s)". The emission from adsorbed pyrene is a mixture of monomer and static/diffu sive excimers on the surface as well as in the pores. Cooling leads to an increase of excimer emission intensity which reaches a maximum in the range of -50 to -100°C. Further cooling, in general, leads to a decrease in the excimer emission intensity. The effect of cooling seems to be a decrease in the number of diffusive 199 encounters; what might be termed a "diffusive encounter volume". At room temperature, this volume is very large and the diffusion of pyrene is very rapid which has been documented (8). Diffusion on the surface may be of two varieties, long range and short range. Long range diffusion refers to the usual, random collisions observed at room temperature; best demonstrated by solution studies. Short range diffusion refers to re la tiv e ly selective collisions occuring from nearest neighbors. This kind of diffusion is best demonstrated from studies involving solvents of different viscosities or low temperature solution studies. Both kinds of diffusion are present at room temperature with long range diffusion seeming more dominant. Cooling seems to remove long range diffusion effects and shorter range diffusion becomes more important. Long range diffusion seems to be selective because encounters w ill tend to favor formation of excimers if the colliding pyrene molecules are of the "correct" orientation. Short range diffusion leading to excimer formation seems less selective because the effective encounter volume on cooling seems to be increasing. Since long range diffusion is slowed down, the pyrene molecule on the surface may be presented with more nearby neighbors which are not in the "correct" orientation for excimer formation. Short range diffusion either by direct translational motion or rotation to a proper orientation for excimer formation seems to be the important processes with cooling. This may be why the excimer emission intensity increases with the in itia l cooling. An alternative explanation as to why the excimer emission increases with cooling 200 comes from Birks and Kazzaz (41) in their work with pyrene crystals at different temperatures. They reported that the excimer emission intensity increases with decreasing temperature, however, they also report that the emission envelope also sharpens dramatically and is blue shifted. These were not observed in this series of experiments which further lends evidence against emission from microcrystal s. The effect of further cooling is to decrease the encounter volume and eventually short range diffusion and rotation are stopped. This may be the cause for the eventual decrease of the excimer emission intensity with cooling. As diffusive excimers are gradual ly stopped in formation, the contribution from static dimers to the excimer emission become dominant. Intrapore effects w ill become more important than surface effects if the pyrene molecule is in a pore when cooling begins, e.g. size constraints w ill lim it motions. The decay analyses show nonexponential behavior for both the monomer and excimer emissions. This has been seen by others (1-15,23,28,32). The monomer decays a ll behave sim ilarly with a fast decay component that is attributed to excimer emission and a longer decaying component which is attributed to monomer emission at long times according to the scheme by Singer, Francis and Lin (28). The excimer decays show similar behavior but each surface has characteristic decay patterns at lower excimer loadings. (This is observed when monomer emission intensity is comparable to excimer emission intensity.) At greater excimer loadings, where no monomer emission is seen, a ll excimer decays on a ll surfaces behave sim ilarly. 201 The spectral differences seen for spectra recorded at t=0 and t=150 can be correlated with the excimer decay patterns for each of the samples. Since a ll variables are equal, except for the porosity and surface area, on a ll these surfaces, i t is reasonable to suggest porosity and surface area are responsible for the unique excimer decays. S24 and S72 share sim ilar surface areas and comparable amounts of adsorbed pyrene w ill produce comparable spectra between the two surfaces. However, the excimer decays are completely different between the two surfaces. The only real difference between these surfaces is the porosity. In pores less than 32 A radius, both S24 and S72 have similar proportions, 51% and 44% respectively. More importantly, there is a greater number of very small pores (less than 32 A) on S72 than S24. The S24 excimer decays seem more indicative of diffusive excimers because their behavior p a ra lle ls solution behavior with the exception of the faster decaying component. The S72 excimer decays behave as i f static dimers are the dominant excimer as soon as the temperature decreases. The pyrenes seem to be more localized on S72 and i t seems possible that the smaller pores are populated with static dimer or near static dimer forms. Surface and porosity differences may also be responsible for pyrene's behavior on S63. Because of its great surface area, 925 m^/gm, comparable amounts of pyrene on this surface result in more monomer emission than excimer emission re la tiv e to the other surfaces. 202 Monomer decays on this surface behave similar to the monomer decays on the other surfaces. I t is the excimer decays at the lower loadings, where som e monomer emission is observed along with som e excimer emission, that show very unusual behavior. The excimer decays are nonexponential and there is a fast component which can be correlated with static dimer, but the longer decaying component shows an increase in the lim iting slope with cooling at -50° and -100°C. I t is only at -196°C that the lim iting slope decreases according to the behavior seen on the other surfaces and in solution studies. This increased lim iting slope means that the "excimer" has a shortened 1ifetime. I t is possible that decreasing the temperature has slowed down the formation of som e "form" of the pyrene dimer. This form of the pyrene dimer may be a rotational form of the dimer or a dimer where the pyrene molecules are not exactly parallel but at som e small angle. This explanation is plausible since Yamazaki, Tamai and Yamazaki were able to observe various forms of the pyrene excimer (33). Cooling at -196°C seems to stop this rotation or small motion to excimer formation. The large portion of small pores seem to be inaccessible to pyrene adsorption because of the dimensional constraints. I t is not clear that som e excimers are from strained dimers because other excimers ,static and diffusive, are dominant forms which tend to mask these lesser forms. These strained dimers would be blue shifted compared with a "normal" dimer and the enhancement of band I I I of the monomer emission may be indicative of this strained dimer as the Yamazaki report suggests (33). (Again, an alternative interpretation for the enhancement of band I I I may be contribution from an overall dominating excimer emission rather than specifically from a "kind" of excimer.) O n S63, many of the adsorbing pyrene molecules may reside in pores greater than 6 A which is only 17% of the surface. Many of these sites may have m ultiple occupancy leading to static dimer formation. I t is interesting that at very large amounts of adsorbed pyrene, the excimer decay behaves more 1 ike the excimer decays on S24, or the higher loaded S72 excimer decay. The vibrational analyses of the monomer emission indicates that the surfaces have a polarity similar to n-butyl ether for S72 and S63 while S24 has a polarity sim ilar to t-amyl alcohol. The Arrhenius plots show a ll the surfaces to have considerably smaller activation energies for pyrene diffusion than the 4 kilocalories (0.17 eV) de Mayo has reported (22). The plots also show a binding energy which is considerably less than the solution value of 0.37 eV (8.5 kilocalories) (26,32). I t has been shown in solution studies (26,32,37) that smaller values of the binding energy are seen for structurally hindered systems. This may be the situation for pyrene on these silicas where surface structure constraints may prevent formation of the ideal excimer geometry. The c ritic a l temperature for pyrene excimer in solution studies is at or s lig h tly above room temperature, but the c ritic a l temperature on these native surfaces is at -50°C. An additional feature of these 204 Arrhenius plots is the presence of the greater negative slope which may represent som e kind of diffusional barrier for excimer formation such as entering or leaving a pore. This negative slope can be characterized as a diffusional barrier because diffusional processes show greater I e/ I m ratios with increasing temperatures as reported by Birks (26). 4.4 Experimental 4.4.1 M aterial s Three silicas, Syloids 24, 72 and 63 were obtained from Grace- Davidson Chemical Division. Their physical properties have been discussed here and in Chapters 2 and 3. Hexane solvent (Mai 1 inkrodt) was used in the adsorption procedures after i t was stirred and le ft standing over magnesium sulfate. Pyrene (Aldrich) was chromatographed on alumina and s ilic a twice. I t was then sublimed twice and kept in a foiled vial prior to use. Polydimethyl siloxane (Petrarch, PS-040, viscosity = 50 centipoise) was used as received in the UV-visible assays of the loaded silicas. 4.4.2 Equipment UV-visible absorption spectra were recorded on a Beckman Acta VI spectrophotometer using a matched pair of one centimeter pathlength quartz cel 1 s. Steady-state emission spectra were recorded on an American Instrument Co. spectrof1uorimeter with a Hamamatsu R446 205 photomultiplier tube. Spectra were not corrected for instrument response. Time-resolved emission spectra were recorded on a pulsed nitrogen laser (337 nm, 10 ns, peak power lOOkW) in conjunction with a Princeton Applied Research boxcar integrator (Model 160) as previously described (25). A Hewlett-Packard 183A oscilloscope (50 ohm termina tion) was used to monitor and photographically record the fluorescence decay of the various samples. Analyses of the data were processed by an IBM PC/PC clone using the Lotus spreadsheet as a template. 4.4.3 Methods 4.4.3.1 Adsorption of Pyrene on the Various Si 1icas A stock solution of 5 X 10“^M pyrene in hexane was prepared. From this stock solution, volumetric aliquots were taken to prepare sets of s e ria lly diluted pyrene solutions. About 0.5 gram samples of s ilic a , oven dried at about 100 C, were placed into small beakers, wetted with 1 ml of hexane and introduced to the specific diluted pyrene solutions. These samples were capped, parafilmed and le ft stirring for variable periods (from hours to overnight) to effect variable adsorption. The silicas were centrifuged, decanted and washed 2-3 times with fresh hexane. The washings consisted of stirring the s ilic a in the fresh hexane for a few minutes and then centrifuging and decanting. The silicas were placed into v ia ls , covered with fo il and placed under a vacuum overnight to remove solvent. Samples were always kept in a foiled vial until they were used for spectroscopic measurements. 206 4.4.3.2 UV-visible Absorption Assays After samples were prepared, UV-visible assays were done by suspending a small quantity of s ilic a with adsorbed pyrene into the polydimethyl siloxane. Suspension was accomplished by sonic mixing until the sample was homogenous. Good quality transmission spectra could be obtain by this method because of the close matching of the refractive indices between s ilic a and po1 ydimethlysiloxane. 4.4.3.3 Fluorescence Spectra and Decay Measurements Fluorescence spectra were taken after samples had been degassed, sealed and labeled in pyrex cuvettes. Decays were a ll photographically recorded off the Hewlett-Packard 183A oscilloscope. Low temperature studies were done by placing the degassed, sealed samples into a sealed dewar which was modified to allow an optical pathway, quartz windows, and temperature modification and monitoring. Temperature modification was achieved by passing nitrogen gas through a coil in another dewar f il l e d with liquid nitrogen. The resultant cooled gas was passed into the dewar with the sample. Temperatures could be changed by varying the flow rate of the gas. Samples were allowed to equilibrate for at least 25 minutes before measurements were started. Decays were photographically enlarged and analyzed. 207 REFERENCES 1) Weis, L.D., Evans, T.R., Leermakers, P.A., JACS, 90(22), (1968)6109 8 10 11 12 de Boer, J.H., Z. Phys. Chem., B14, (1931)163 Hara, K., de Mayo, P., Ware, W., Weedon, A., Wong, G., Wu, K., Chem. Phys. L e tt., 69(1), (1980)105 Oelkrug, D., Radjaipour, M., Z. Phys. Chem., NF88, (1974)23 Ishida, H., Takahashi, H., Tsubomura, K., Bui 1. Chem. Soc. Jpn., 43, (1970)3130 Oelkrug, D., Radjaipour, M., Z. fu r Phys. Chem. Neue Folge Bd, 123S, (1980)163 de Mayo, P., Bauer, R., Ware, W., J. Phys. Chem., 86, (1982)3781 de Mayo, P., Bauer, R., Ware, W., Natajaran, L., JACS, 104, (1982)4635 Thomas, J.K., Chandrasekaran, K., J. of Colloid & Interfacial Sci., 100, (1984)116 Lochmuller, C.H., Colborn, A.S., Hunnicutt, M.L., Harris, J.M., JACS, 106, (1984)4077 Bauer, R., de Mayo, P., Ware, W., Natarajan, L., Can. J. Chem., 62, (1984)1279 Singer, L., Francis, C., Lin, J., Chem. Phys. L e tt., 94, (1983)162-167 13) Avnir, D., Daniel, H., Grauer, Z., J. Colloid & Interfacial Sci., 1983 14) Turro, N., Cheng, C.C., JACS, 106, (1984)5022 15) Drake, J.M., K1 a fte r , J., 0. of Lumin. 31-32, (1984) 642 16) Synder, L.R., Separation Science, JL{2 & 2I» (1966)191-218 17) Adamson, A.W., Phys. Chem. of Surfaces, 4th ed„, Wi 1 ey and Sons, New York, (1982) 18) Boehm, H.P., Knozinger, S., Catalysis-Sci. and Tech., Vol.4, M. Boudart, J.R. Anderson eds., Springer-Veriag, New York, 1982 19) Gregg, S.J., Sing, K.S.W., Adsorption, Surface Area and Porosity, 2nd ed., Academic Press, New York, (1982) 20) Okkrese, C., Phys. and Chem . Asp, of Adsorbants and Catalysts, B.G. Linsen ed., Academic Press, New York, (1970) 21) Nishioka, G.M., Schranke, J.A., J. Colloid & Interfacial Sci. 105(1), (1985) 102 22) de Mayo, P., Organic Phototransformations in Nonhomogenous Media, ACS Symp. Ser. #278, M.A. Fox ed., (1984)1-19 23) Avnir, D., Ottolenghi, M., Wellner, E., Langmuir, 2, (1986)616 24) A v ir, D., in press, (to be in Sept. Langmuir) 25) Brown, R.E., Legg, K., Wolf, M., Singer, L.A., Parks, J.H., Anal. Chem., 46, (1974)1690 26) Birks, J. B., Photophysics of Aromatic Molecules, Wiley- Interscience, New York, 1970 and references therein 209 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 Singer, L.A., A tik , S., Nam, M., Chem. Phys. L e tt., 67, (1979)75 Singer, L.A., Francis, C., Lin, J., Chem. Phys. L e tt., 94, (1983)162-167 Thomas, J.K., Chem. Rev iews, 80(4), (1980)283 Dong, D.C., Winnik, M.A., Photochem. and Photobio., 35, (1982)17-21 Nakajima, A., Bui 1. Chem. Soc. Jpn., 44, (1971)3272 Aoudia, M., Rodgers, M., Wade, W.H., J. Phys. Chem. 88, (1984) 5008 Yamazaki, T., Tamai, N., Yamazaki, I., Chem. Phys. L e tt., 124(4), (1986)326 Bauer, R., de Mayo, P., Ware, W., Natarajan, L., Can. J. Chem., 62, (1984)1279 de Mayo, P., A b del-M alik, M.M., Can. J. Chem., 62, (1984)1275 Bauer, R., de Mayo, P., Ware, W., Natarajan, L., Chem. Phys. L e tt., 107(2), (1984)187 Birks, J.B., Braga, C.L., Lumb, M.D., Proc. Royal Soc. London Ser. A, 283, (1965)83 Moldenhauer-Kopach, K., Dissertation, University of Southern California, (1987) Synder, L.R., Ward, J.W., J. Phys. Chem., 70(12), (1966)3941 The reader is directed to Chapters 1-3 in this work. Birks, J.B., Kazzaz, A., Proc. Royal Soc. London, Ser. A, 304, (1968)291 210 CHAPTER 5 PHOTOPHYSICAL STUDY O F PYRENE ADSO RBED O N TO SILYLATED SILICA 5.1 Introduction Surface modification of s ilic a greatly influences the surface properties such as surface area and porosity (1-8). Som e modifications w ill generate new domains on the surface while others w ill tend to homogenize the surface such as with alcohol modification that was studied by Bauer and de Mayo (5). Another method, s ily la tio n , should also theoretically homogenize the surface (2). S ilylatio n was discussed in Section 1.5. The modification is the reaction of a s ily la tin g agent, hexamethyldisilazane used in this work, with hydroxyls found on the surface and in pores which substitutes a bulky trimethyl si ly l group for a hydrogen. There are many other methods of s ily la tio n (2) and theoretically the sam e results should occur. However, depending on the type of s ily la tio n reaction used, s lig h tly different properties can result as was reported by Okkrese (9). S ilylatio n w ill tend to block pores near their opening; Nishioka and Schramke (10) have reported that deep silanols, ty p ic a lly "reactive" silanols, are not accessible to s ily la tio n . This is an important finding because any adsorption would have to be related to these deep silanols since it is known that surface adsorption of hydrocarbons is via the interaction between the silanol and the hydrocarbon (1,4,9,15,18). Surface silanols w ill have 211 been silylated and are not available as an adsorption site. Homogenization of the surface should increase the dispersion of pyrene as Singer, Francis and Lin (6) have shown. 5.2 Si 1y 1ated Si 1ica Results 5.2.1.0 Variable Temperature Time Resolved Spectra of Pyrene on SS24 S24, S72 and S63 were silylated according to the method outlined in the Experimental section of Chapter 2. The resulting silicas are designated as SS24, SS72 and SS63. Pyrene was adsorbed from concentrated solutions of pyrene as described in this Experimental section. In general, i t was more d iffic u lt to adsorb pyrene onto the silylated silicas. This observation was also noted by Moldenhauer- Kopach (11) who studied the adsorption isotherms of arenes on various silicas. The attempt to have three loadings of pyrene; "low", "medium" and "high" was unsuccessful. (These terms are only q u alitative and used to describe the intensity of the excimer emission compared with the monomer emission intensity; e.g. these terms describe the probe dispersion on the surface. Exact concentrations are discussed in the text.) Only small amounts of pyrene were adsorbed leading to problems. For example, the emission spectra showed mostly excimer emission and since only small amounts of pyrene were adsorbed, i t was d iffic u lt to analyze the concentration of adsorbed pyrene by u lta v io le t-v is ib le spectroscopy. Figures 5.1 and 5.2 are the variable temperature time resolved spectra of SS24. The most striking feature of these spectra is that 212 Figure 5.1 V ariab le Temperature Spectra of 1.78 umole/gm Pyrene on SS24 at t=0. RELATIVE INTENSITY (ARBITRARY UNITS) O o O- r v > . o Figure 5.1a Variable Temperature Spectra of 1.78 umole/gm Pyrene on SS24 at t=150. 215 RELATIVE INTENSITY (ARBITRARY UNITS) O- O- low amounts of pyrene were adsorbed, but very large amounts of excimer emission are present. This is contrary to early results by Singer, Francis and Lin (6) and suggests that the primary effect of s ily la tio n is not dispersion, but rather aggregation of the pyrene molecules. Figure 5.1, 1.78 umole/gm of adsorbed pyrene, shows very dominating excimer emission with a small contribution from monomer emission. These spectra taken at t=0 show a greater re la tiv e excimer emission than spectra taken at t=150. A decrease in temperature causes the monomer emission intensity to increase until i t reaches near or greater intensity than the excimer emission intensity at -196°C. The increase in monomer emission intensity is not unusual but the dramatic change in monomer-excimer emission intensities for this temperature range compared with samples with similar amounts of in itia l excimer emission intensity on native silicas is a dramatic difference, e.g. compare these spectra with the spectra of 36.5 umole/gm of adsorbed pyrene on S24. Additionally, there is no increase in excimer emission intensity with in itia l cooling as was observed for the native samples. This sample shows a consistent decrease in excimer emission intensity which is similar to pyrene in solution, e.g. compare these spectra with the spectra of 6.5 X 10"^ M pyrene in hexane. The monomer emission, when observed at -196°C, shows an extremely well resolved emission pattern which is more similar to monomer emission of pyrene in hexane solutions than for any of the monomer emission of adsorbed pyrene on the native silicas. The monomer emission peaks are at the sam e wavelengths as the monomer emission of 217 Figure 5.2 Variable Temperature Spectra of 13.8 umole/gm Pyrene on SS24 at t=0. 218 n m 460 500 RELATIVE INTENSITY (ARBITRARY UNITS) O ” I I Figure 5.2a V aria b le Temperature Spectra of 13.8 umole/gm Pyrene on SS24 at t=150. RELATIVE INTENSITY (ARBITRARY UNITS) pyrene in solution and monomer emission of adsorbed pyrene on native s ilic a . The excimer emission in it i a l ly is at 474 n m at room temperature but gradually shifts its maximum emission intensity out to 480 n m at -196°C. This observation has been seen for pyrene crystals in the temperature range of -118 to -196°C (12,13) and suggests that som e microcrystals may be present on the surface. I t was proposed that the red shift is a result of the pyrene crystal undergoing a phase change at -148°C from a dimeric form to an unknown high density form (12). The evidence for microcrystal deposition versus true adsorption is not easily demonstrated. The methods employed in this work have tried to minimize the 1 ikelihood that microcrystals are being deposited. Moldenhauer-Kopach (11) completed a scanning electromicroscopy (SEM) study which did not show any evidence of crystalloid structures employing the sam e adsorption procedures used in this work. She was able to show evidence for the presence of crystalloid structures using a flash deposition method employed made popular by de Mayo (14). Furthermore, her work in the adsorption isotherms of various arenes are consistent with typical adsorption isotherms. Other procedures used to minimize microcrystal deposition involved using fresh hexane to wash the sample 2-3 times in order to avoid such a possibility and the loadings are at such a low concentration that microcrystal formation is not re a lly conceiveable. The question of microcrystal deposition versus true adsorption has been minimized to a great extent by the methods employed here and the 222 bulk of the evidence suggests that more or less true adsorption has been demonstrated in this work. However, microcrystal deposition can not be exclusively ruled out. (This is especially true because there is no definition as to what is "microcrystal deposition".) Figure 5.2, 13.8 umole/gm of adsorbed pyrene, shows the same kind of behavior with the exception that more excimer emission is present because of the greater amount of pyrene adsorbed. 5.2.1.1 Decay and Monomer Emission Vibrational Analysis of Pyrene on S524 Tables 5.1 and 5.2 summarize the data on the monomer lifetimes and vibrational band intensities. The monomer lifetimes are are longer than the native lifetimes and are more consistent with solution lifetimes. Figures 5.3 and 5.4, 1.78 and 13.8 umole/gm of adsorbed pyrene, respectively show the variable temperature monomer emission decays. The decays are nonexponential and show a faster decaying component followed by a longer decaying component as was seen with the monomer emission decays of pyrene adsorbed on native s ilic a . Figures 5.3a and 5.4a are the corresponding excimer emission decays. The decays for both loadings show nonexponential decays. The decays do not show any evidence for a risetime for diffusive excimers but are more consistent with contributions from a static dimer population as Singer, Francis and Lin (6) have ascertained. At 1.78 umole/gm of adsorbed pyrene, the excimer decays show an apparent com m on fast decay at a ll temperatures followed by a shift to smaller 223 Table 5.1 Pyrene Monomer Lifetim es of V ariab le Concentrations at Variable Temperatures (SS24) Sample (umole/gm) Monomer Lifetim e (ns) Temperature (C) 1.78 311.1 R.T. 431.6 -50 340.9 -100 212.5 -196 13.8 312.2 R.T. 321.2 -50 310.6 -100 417.9 -196 224 L N 1 R E LA TIV E 0 -02 -0 .4 -06 -0 8 -1 - 1 2 -L 4 -1 6 -L8 -2 -22 -2 .4 - 2 6 -2 6 -3 □ R .T . Lo □ a o a A * 0 A + □ A o A A 0 -p & 4 - 0 □ + o 4 - 0 + □ 02 + -5 0 C — i ---------------1 ---------------- r— 0 4 0 6 (T h o u sa n d *) T IM E (N S o -100 C ro ro cn Figure 5.3 Monomer Decay for 1.78 umole/gm Pyrene on SS24. 08 1 - 1 9 6 077 1 0 LMRJELATTVE -O l5 - -1 - L 5 - -2 - - 3 - 0 4 0 1 2 0 280 200 240 1 6 0 T IM E (NS) □ R T . + -S O C O -100 C a -196 0 7 7 1 0 ro ro C T l Figure 5.3a Excimer Decay for 1.78 umole/gm Pyrene on SS24. LZZ - S aT. 02 04 0j6 (T houaan dc) T IM E (NS) o -400 C 08 + -SDC O -400 C A - 196 077 K ) Figure 5.4 Monomer Decay for 13.8 umole/gm Pyrene on SS24. LNtRELATTVE 0 -02 -04 -0 6 -08 -1 -L 2 -L4 -16 -18 -2 -22 -2.4 -26 -26 -3 0 4 0 BO Q O 160 2 0 0 2 4 0 2 8 0 T IM E (NS) □ R T . + -5 0 C O 4 0 0 C 4 -196 0 7 7 K ) Figure 5.4a Excimer Decay for 13.8 umole/gm Pyrene on SS24. ro 00 \ % lim itin g slopes with decreasing temperature. This decay is sim ilar to the excimer decays of S24. At 13.8 umole/gm of adsorbed pyrene, the excimer decays show an apparent com m on decay at a l 1 temperatures except a s lig h t deviation to a sm aller lim itin g slope at -196°C. This decay is somewhat sim ilar to the S72 excimer decays. The monomer emission vibrational data is d if f ic u lt to interpret because of the large contribution from excimer emission which distorts the true emission intensities of the longer wavelengths of the monomer emission. At best, for the 1.78 umole/gm sample at -196°C, the monomer emission 111/I ra tio indicates an environment sim ilar to n-butyl ether. Singer, Francis and Lin (6) reported an environment sim ilar to a siloxane. 5.2.1.2 Mobi1ity and Activation Energy of Excimer Formation on SS24 Table 5.2 contains the data of the I e/ I m ratios. The greatest decrease in this ra tio is seen between room temperature and -50°C. Thereafter, the decrease is s lig h t. This suggests that any m obility is rapid at room temperature but becomes lim ited to a sm aller area which is r e la tiv e ly constant as the temperature is decreased. Figures 5.5 and 5.6 are the Arrhenius plots for SS24. The plots show a bimodal curve where no "high" temperature behavior is seen implying that the c ritic a l temperature is greater than room temperature. A steep negative slope develops at both loadings. These plots are d iffere n t from the Arrhenius plots of the native s ilic a s , but are somewhat sim ilar to the Arrhenius p lo t of pyrene in hexane. 229 Table 5.2 Monomer V ib ratio n a l In te n s itie s (Pyrene on SS24) Normalized Monomer Band Intensities Ip /I_ Sample I (374nm) I I (384nm) I I I (395nm) (465nm)/(395nm) 1.78 umole/gm t= 0 , R.T. -50 C 1.00 1.41 4.74 -100 C 1.00 1.03 1.50 2.22 -196 C 1.00 0.74 1.56 1.16 150, R.T. 1.00 ------- 1.50 1.33 -50 C 1.00 0.67 1.33 0.75 -100 C 1.00 0.80 1.20 0.57 -196 C 1.00 0.75 1.18 0.51 13.8 umole/gm t= 0 , R.T. -50 C 1.00 1 1.70 7.59 -100 C 1.00. 0.97 1.33 5.03 -196 C 1.00 0.61 1.27 3.83 150, R.T. 1.00 ------- 1.57 4.43 -50 C 1.00 0.87 1.39 2.08 -100 C 1.00 0.70 1.21 1.93 -196 C 1.00 0.67 1.07 1.90 L4 - 12 - 0 8 - 06 - 02 - -0 6 - 13 7 9 I/TXIOOO (K ! □ t t - 0 + • ® t-l5 0 Figure 5.5 Arrhenius Plot for 1.78 umole/gm Pyrene on SS24. L9 - L 7 - L6 - L 4 - 13 - L2 - U - 09 - 08 - 0.7 - 06 U 5 9 7 1/TX1000 (K) 0 W *0 + ■ e t-1 5 0 Figure 5.6 Arrhenius Plot for 13.8 umole/gm Pyrene on SS24. ro c * > ro The difference between these Arrhenius plots and the pyrene in hexane Arrhenius p lo t is the small negative slope yielding to a greater negative slope. O n the s ily la te d surface, diffusion occurs but i t seems to require two processes, normal surface diffusion and som e other kind of diffusion with a greater barrier. These plots show diffusional barriers which are considerably less than the 4 kilo c a lo rie s de Mayo reported (15). A d ditio nally, the plots seem to suggest that an effect of s ily la tio n is to s h ift the c ritic a l temperature back above room temperature as i t is in solution studies. 5.2.2.0 Variable Temperature Time Resolved Spectra of Pyrene on SS72 Figures 5.7 to 5.9 are the spectra fo r 1.75, 26.8 and 28.8 umole/gm of adsorbed pyrene on SS72, respectively. Again, the most strikin g feature of these spectra is the low loading with a high excimer emission intensity content. Excimer emission intensity at t=150 is less than at t=0 again. The monomer and excimer emission peaks are at the same energies as the monomer and excimer emission peaks seen on the SS24. The intensities also show a large decrease in excimer/monomer emission intensity on going from room temperature to -50°C but a r e la tiv e ly smaller decrease from -50°C to -196°C. The only sample showing any appreciable amount of monomer at -196°C is the 1.78 umole/gm sample. There is greater aggregation on this surface than on SS24 for nearly identical loadings, e.g. compare the -196°C spectra of 1.78 umole/gm of adsorbed pyrene on SS24 to the -196°C spectra of 1.75 umole/gm of adsorbed pyrene on SS72. This effect can be attributed to the sm aller surface area of SS72. 233 Figure 5.7 V ariab le Temperature Spectra of 1.75 umole/gm Pyrene on SS72 at t=0. RELATIVE INTENSITY (ARBITRARY UNITS) o- Figure 5.7a Variable Temperature Spectra of 1.75 umole/gm Pyrene on SS72 at t=150. 236 RELATIVE INTENSITY (ARBITRARY UNITS) O O I I Figure 5.8 Variable Temperature Spectra of 26.8 umole/gm Pyrene on SS72 at t=0. 238 RELATIVE INTENSITY (ARBITRARY UNITS) O cr o Figure 5.8a Variable Temperature Spectra of 26.8 umole/gm Pyrene on SS72 at t=150. 240] nm4 0 0 5 0 0 600 : : : 24-1 RELATIVE INTENSITY (ARBITRARY UNITS) Figure 5.9 Variable Temperature Spectra of 28.8 umole/gm Pyrene on SS72 at t=0. 242 RELATIVE INTENSITY (ARBITRARY UNITS) O - O r O r ro co Figure 5.9a Variable Temperature Spectra of 28.8 umole/gm Pyrene on SS72 at t=150. 244 RELATIVE INTENSITY (ARBITRARY UNITS) o- o - ro 4 s . in The 1.75 umole/gm of adsorbed pyrene spectra show a very dominating excimer emission intensity with a very minor contribution from monomer emission. With cooling, the monomer emission intensity increases while the excimer emission intensity decreases. At -196°C, the monomer emission is very w ell resolved and the excimer emission envelope has sharpened. The other two higher loaded samples show predominantly excimer emission at a ll temperatures with an almost n eg lig ib le amount of monomer emission. 5.2.2.1 Decay and Monomer Emission Vibrational Analysis of Pyrene on SS72 The monomer life tim e and vibrational data are found in Tables 5.3 and 5.4. The life tim e of the monomer is comparable to the SS24 monomer lifetim es. Figures 5.10 and 5.11 are the monomer emission decays and they show the decay to be nonexponential, a fast decay followed by the r e la tiv e ly long monomer decay. Decreasing the temperature has an effect on both the fast component and the longer decaying component. The fast component shows a greater curvature and the longer decaying component shows a sm aller lim itin g slope with decreasing temperatures. Figures 5.10a and 5.11a, the excimer decays, show an unresolved fast component, s tatic dimer, followed by the longer d iffu sive excimer decay. The decays show a small contribution from the fast decay which is a com m on decay at a ll temperatures. The longer decay component has a smaller lim itin g slope with decreasing temperatures which is expected. A comparison of the emission spectra and these decay plots 246 Table 5.3 Pyrene Monomer Lifetim es of V ariab le Concentrations at Variable Temperatures (SS72) Sample (umole/gm) 1.75 26.8 Monomer Lifetim e ( 307.8 472.2 468.9 549.2 235.4 441.0 496.9 538.4 ) Temperature (C) R.T. -50 -100 -196 R.T. -50 -100 -196 247 Table 5.4 Monomer Vibrational Intensities (Pyrene on SS72) Sample Normalized I (374nm) Monomer Band I I (384nm ) Intensities I I I (395nm) (AeSnmj/^OSnm) 1.75 umole/gm t= 0 , R.T. -50 C 1.00 1.10 1.70 7.38 -100 C 1.00 0.97 1.39 5.62 -196 C 1.00 1.15 1.70 2.39 t=150, R.T. 1.00 1.17 1.79 2.16 -50 C 1.00 0.96 1.63 1.32 -100 C 1.00 0.94 1.51 1.32 -196 C 1.00 0.92 1.39 1.36 26.8 umole/gm t=0, R.T. A ll excimer; no measurable monomer -50 C 1.00 ------- 1.78 10.12 -100 C 1.00 0.77 1.36 9.47 -196 C 1.00 1.02 1.31 6.41 t=150, R.T. 1.00 1.62 2.27 5.08 -50 C 1.00 1.13 2.40 4.90 -100 C 1.00 1.03 1.83 5.18 -196 C 1.00 0.73 1.16 5.20 248 LNt R ELA TIVE □ KJ. -5 0 C 0 4 0 j 5 (T h o tts c n d t) T IM E INS) O -400 C -196 0 7 7 K ) ro - p > Figure 5,10 Monomer Decay for 1.75 umole/gm Pyrene on SS72. L M R E LA TIV E > A -0 6 - 1 2 -L 4 -L 6 -Lfl -22 -24 - 2£ 2 8 0 0 200 2 4 0 60 1 2 0 £0 4 0 T IM E <NS □ R .T . + -5 0 C o -1 0 0 C a -196 0 7 7 K> n o cn o Figure 5.10a Excimer Decay for 1.75 umole/gm Pyrene on SS72. LNWELATTVE 0 -02 -04 -0 6 -08 -1 -L2 -1 4 -L 6 -18 -2 -22 -2.4 - 2£ - 2& -3 □ R .T . 9 b r o 0 4D+ O a + o °6 - 4 □ i □ □ a □ A A + ■ ■ b +b A + A 4 □ □ 0 -I------------- B 200 4 0 0 4- -5 0 C T IM E <NS) o 4 0 0 C a -196 0 7 7 K ) Figure 5 .I I Monomer Decay for 26.8 umole/gm Pyrene on SS72. A 600 ro cn LNCRELATIVE -0.5 -IS - -2 - -2 .5 - -3 - -3 .5 □ at. m no m no T IM E (N S + *5 0 C c -100 C ^ -196 C (77 K> Figure 5.11a Excimer Decay for 26.8 umole/gm Pyrene on SS72. show that excimer emission intensity decreases by over 90% in going from t=0 to t=150 ns. In comparison, the monomer emission intensity decreases by about 60%. Monomer vibrational analysis is almost impossible to do with the very large excimer emission intensities that are present. The best approximation of the surface p o la rity , at -196°C for 1.75 umole/gm of adsorbed pyrene, shows a p o la rity intermediate between n-butyl ether and polydimethyl siloxane when comparing the standard solution values of p o la rity found in Table 4.2. 5.2.2.2 Mobi1ity and Activation Energy of Excimer Formation on SS72 The m obility of pyrene on this surface, assuming I e/ I m values are indicative of m obility, is very lim ited. Maximum m obility is at room temperature with the greatest observed decrease of I e/ I m between room temperature and -50°C. There is only a s lig h t decrease of I G / I m with cooling to -196°C. This suggests that short range diffusion becomes the more important process at temperatures below room temperature compared with long range diffusion. The a b ility to move on this surface seems more complex than for SS24. Figure 5.12 is the Arrhenius plo t of 1.75 umole/gm of adsorbed pyrene and i t shows a bimodal curve as was observed fo r SS24. The plot does not show any "high" temperature behavior suggesting that the c ritic a l temperature must be greater than room temperature as was also observed with SS24. However, there are som e differences between this p lo t and the SS24 plots. 253 { JUMJ 19 - 18 - L 7 - 16 - L5 - M - L 3 - 12 - U - 0 .9 - 0 8 - 0 7 - 06 - 05 - 04 - 03 - 02 - 13 U 9 7 l/TXUJOO (K ) □ o t-0 -t- ® M 5 0 ro cn - P > Figure 5.12 Arrhenius Plot for 1.75 umole/gm Pyrene on SS72. The early time plot, at t=0, shows the characteristic small negative slope yielding to the steeper negative slope, but never reaching a c ritic a l temperature. The later time plot, at t=150, shows a slight positive slope leading to a steep negative slope. This kind of behavior has been seen in solution studies of structurally hindered benzene excimers by Birks (16,17). Because the slope is so slight and pattern is not consistent with a ll of the Arrhenius plots done for a ll the silylated surfaces, there may have been error in measuring the exact excimer/monomer ratio because of the large amount of contributing excimer emission. It is possible that the large excimer emission intensity could enhance band I I I intensity of the monomer so that the apparent I e/ I m ratio is smaller than it should be. At a higher concentration, Figure 5.13, at 26.8 umole/gm, the Arrhenius plot shows a linear plot at t=0 but a bimodal curve at t=150. The linear plot is very suggestive of purely surface diffusional processes at work at early times. I f s ily la tio n had completely homogenized the surface and m ade it more solution-like, then this kind of Arrhenius plot would have been expected. However, the later plot at t=150 shows a bimodal pattern similar to the later time plot at 1.75 umole/gm. The barriers to diffusion are also considerably less than the reported 4 kilocalories of de Mayo (15). Furthermore, as observed for SS24, the c ritic a l temperature has shifted above room temperature and the surface seems to be expressing som e solution-like properties. 255 L 7 1 - 169 - L66 - L65 - 162 - L 5 9 - ©t-0 l/TX I000 + - (K ) •4 -1 5 0 Figure 5.13 Arrhenius Plot for 26.8 umole/gm Pyrene on SS72. 5.2.3.0 Variable Temperature Time Resolved Spectra of Pyrene on SS63 Figures 5.14 and 5.15 are the spectra of 2.70 umole/gm and 14.6 ! umole/gm of adsorbed pyrene on SS63 respectively. O n this surface, the jpyrenes are more dispersed than on any of the silylated silicas. This result is surprising since the surface area is comparable with the SS24 surface area. S t i l l , the emission for 1.78 umole/gm of adsorbed pyrene on SS24 shows m uch more excimer emission than the emission spectra for 2.70 umole/gm of adsorbed pyrene on SS63. The excimer emission is again greater at t=0 than at t=150. Upon cooling, the excimer emission decreases until at -196°C, there is v irtu a lly no excimer le ft, Figure 5.14. S t i l l , the most striking feature is the amount of excimer emission present for such a low loading compared with the loadings and spectra for the native silicas. Even more interesting is the monomer emission at -196°C. The emission is extremely well defined and it is surprising that there should be so m uch monomer on this type of surface considering aggregation seems to be a dominant process. Emission peaks are the same as before except the excimer emission peak does not red shift to 480 nm as in the other silylated samples but stays at 474 nm . Figure 5.14, 2.70 umole/gm of adsorbed pyrene, shows almost equivalent emission intensities from monomer and excimer at t=0. At t=150, the excimer emission has decreased. With decreasing temperatures, the excimer emission intensity decreases for both time frames until at -196°C, only a small amount of excimer emission intensity is observed. The monomer emission at -196°C is extremely 257 Figure 5.14 Variable Temperature Spectra of Pyrene on SS63 at t=0. 2.7 umole/gm 258 n m 460 1 5 0 0 600 RELATIVE INTENSITY (ARBITRARY UNITS) r\3 U D I Figure 5.14a Variable Temperature Spectra of Pyrene on SS63 at t=150. 2.7 umole/gm 260 RELATIVE INTENSITY ^ARBITRARY UNITS) o- Figure 5.15a Variable Temperature Spectra of 14.6 umole/gm Pyrene on SS63 at t=150. I 262 RELATIVE INTENSITY (ARBITRARY UNITS) £ V V Figure 5.15 Variable Temperature Spectra of 14.6 umole/gm Pyrene on SS63 at t=0. 264 RELATIVE INTENSITY (ARBITRARY UNITS) well resolved. Figure 5.15, 14.5 umole/gm of adsorbed pyrene, shows spectra dominated by excimer emission with a smaller contribution from monomer emission. The t=150 spectra again show a decrease in excimer emission intensity while there is an increase in monomer emission intensity. Cooling causes the excimer emission intensity to decrease and the monomer emission to increase for both time frames. 5.2.3.1 Decay and Monomer Vibrational Analysis of Pyrene on 5S63 Tables 5.5 and 5.6 contain the data of the monomer lifetimes and monomer emission vibrational band intensities for the variably loaded SS63. The monomer lifetimes are comparable with the lifetimes seen on SS24 and SS72. Figures 5.16 and 5.17 are the monomer emission decay plots and they show decays which are nonexponential. There appears to be a very small contribution from a fast decaying component which is followed by a longer decay component. The 2.70 umole/gm monomer decay shows the longer decay components at different temperatures share a com m on decay which has a s lig h tly smaller lim iting slope than the room temperature monomer decay. At 14.6 umole/gm of adsorbed pyrene, the longer decay components are more spread out; e.g. as the temperature decreases, the lim iting slope also decreases. These smaller limiting slopes indicate the monomer lifetim e is increasing. Figures 5.16a and 5.17a are the excimer decays and they are nonexponential. The excimer decays are very nearly exponential but there is som e unresolved fast component present in the decays. At 2.70 umole/gm, the excimer decay seems to have a com m on decay at a ll Table 5.5 Pyrene Monomer Lifetim es of V ariab le Concentrations at Variable Temperatures (SS63) Sample (umole/gm) Monomer Lifetime (ns) Temperature (C) 2.7 352.0 R.T. 449.9 -50 475.5 -100 485.7 -196 14.6 338.9 R.T. 469.5 -50 488.6 -100 586.6 -196 267 i Table 5.6 Monomer V ib ratio n al In te n s itie s (Pyrene on SS63) I ] Normalized Monomer Band Intensities W 1™ i Sample I (374nm) I I (384nm ) I I I (395nm) (465nm)/(395nm) I 2.70 umole/gm ’ t=0, R.T. -50 C 1.00 1.18 1.53 0.67 -100 C 1.00 1.01 1.31 0.49 -196 C 1.00 0.72 1.03 0.21 t=150, R.T. 1.00 1.40 1.85 0.20 -50 C 1.00 1.05 1.29 0.19 -100 C 1.00 0.91 1.21 0.14 -196 C 1.00 0.79 1.08 0.08 14.6 umole/gm t=0, R.T. -50 C 1.00 1.67 1.80 3.56 -100 C 1.00 1.26 1.91 2.19 -196 C 1.00 1.11 1.31 0.66 t=150, R.T. 1.00 1.64 1.89 0.88 -50 C 1.00 1.25 1.68 0.78 -100 C 1.00 1.08 1.33 0.58 -196 C 1.00 1.08 1.44 0.30 i i LNfRELATIVE 0 -02 -0 .4 - 0 6 -0 8 -1 - 12 -L 4 -L 6 -12 -2 -22 -2 .4 -22 -22 -3 0 0 2 0 4 0 2 0 8 1 (T h o u sa n d *) T IM E (NS) □ R .T . + - 5 0 C < 0 4 0 0 C a -196 0 7 7 K ) Figure 5.16 Monomer Decay for 2.7 umole/gm Pyrene on SS63. X a % -A O 4.^ O L + i + o □ r s 3 c n * j O LNtRELATfVE -02 - -0 6 - -08 - 4 - -12 - d- * -L4 - -L 6 - □ + -18 - -2 - -22 - -2.4 - -26 - 2 8 0 0 240 40 80 1 2 0 160 200 R .T . -5 0 C T IM E (NSI o -100 C -196 0 7 7 K ) ro -s i o Figure 5.16a Excimer Decay for 2.7 umole/gm Pyrene on SS63. L N ! RELATIVE -02 -06 -08 0 L - 1 2 -U -L8 -22 -2 .4 - 2 6 L2 0 02 06 08 1 R .T . r v > '-a (T h o u sa n d *) T IM E (N S + -5 0 C o -100 C a -196 0 7 7 K ) Figure 5.17 Monomer Decay for 14.6 umoTe/gm Pyrene on SS63. L M RELATIVE \ 196 0 7 7 K> -5 0 C T IM E (N S -100 C IN 3 ro Figure 5.17a Excimer Decay for 14.6. umole/gm Pyrene on SS63. temperatures except a deviation to a smaller limiting slope at -196°C. The 14.6 umole/gm excimer decay shows a com m on fast decay component at a ll temperatures which is followed by longer decaying components which have a smaller lim iting slope as the temperature decreases. Comparing the emission spectra with these decays shows that the excimer emission intensity decrease at t=150 can be correlated with the decay of the excimer emission. At t=150, the excimer emission intensity is about 14% of the original intensity whereas the monomer emission intensity has only decayed to about 40% of its original intensity. The monomer vibrational band analysis is the best am ong a ll silylated surfaces because the -196°C spectra of 2.70 umole/gm loading shows almost no excimer. The polarity sensed by the pyrene corresponds to a polarity similar to a t-amyl alcohol. This means the environment sensed by pyrene is a re la tiv e ly hydrophilic surface, comparable to the native silic a. 5.2.3.2 Mobi1ity and Activation Energy of Excimer Formation on SS63 Table 5.6 show I e/ I m ratios which decrease with decreasing temperatures. The greatest decrease is from room temperature to -50°C with a gradual decrease from -50°C to -196°C. This is indicative of the decreasing importance of diffusion. The Arrhenius plots, Figures 5.18 to 5.19, a ll show similar patterns. At early time, the small negative slope yields to a greater negative slope which does not show any trend towards bending down so there is no apparent c ritic a l temperature. However, data collected at t=150 ns shows behavior ______________________________________________________________________________2ia_ 274 -02 -04 - -0 6 - -08 - 4 _ -L2 -L4 -L 6 -L8 -2 H -22 -2.4 H - 2£ T * 7 U 13 l/T X 1 0 0 0 ( K ) + «t-l50 □ Ot-0 Figure 5.18 Arrhenius Plot for 2.7 umole/gm Pyrene on SS63. szz L r t E x c n n e r / M o o t n n e d t o sz -s fD c n g o 3> -5 ~ S =r n > in ~ o o r+ -+ i O -s i — » -P» cr > o _i fD C Q 3 ~ 0 < < -i fD 3 a > o 3 G O G O c r > G O □ H X ft 8 u tn sr G O suggestive of a c ritic a l temperature. These plots show diffusional | | barriers which are much smaller than the reported 4 kilocalories (15).j | These plots seem to indicate the operation of two different | I diffusional processes. i ! • 5.3 Discussion j 1 ! | The s ily latio n procedure has demonstrated that the individual j 1 i I f | properties of the native silicas have been homogenized in general. Thej ) \ | importance of surface area and porosity are further exemplified by j i . J this investigation. I t seems that the primary effect of s ily la tio n is j not the dispersion of pyrene, which is contrary to the early results j i of Singer, Francis and Lin (6), but rather the greater aggregation of j i the probe molecules. ! j | ! The contradictory results may be explained by comparing the j silylated silicas. The silylated s ilic a used in this work was made using the s ily latio n procedure twice. Also a different reaction was used that generated ammonia as a side product. The e a rlie r silylated j s ilic a , by Singer, Francis and Lin (6), was silylated only once and j used a reaction which generated hydrogen chloride. I t has been j documented that the method of preparation can affect the properties of I i the silylated surface (1,2,9). j The surface area of each of the silicas was m uch lower than for the native s ilic a with the exception of SS24 which shows an increase i ! in surface area. The surface areas for SS24, SS72 and SS63 are 453, | ! 9 258 and 462 rrr/gm respectively. I t seems that "new" surfaces were made | on SS24 but most of the pores were blocked on the other two as was discussed in Section 3.2. The porosity was also decreased on a ll surfaces. O n SS24, 25% of the pores are about 18 A and less in radius, assuming a cylindrical pore model, 28% are pores between 18 A and 40 A in radius and 47% of the pores are greater than 40 A in radius. O n SS72, 16% of the pores are about 5 A and less in radius, 35% of the pores are between 5 A and 20 A in radius and 49% of the pores are 20 A and greater in radius. O n SS63, 49% of the pores are 7 A and less in radius and 45% of the remaining pores are between 7 and 30 A in radius and 6% of the pores are greater than 30 A in radius. The adsorption of pyrene onto s ilic a has been attributed to the interaction between silanols and pyrene (1,2,9,10,18,19). An important observation by Nishioka and Schramke (10) was that s ily latio n seemed not to prevent adsorption of water into pores but slowed the rate of adsorption in their studies on the thermal desorption of water from porous and non-porous materials. Their findings suggest that silanols are s t i l l responsible for adsorption of pyrene but that this process would take a longer time (slower rate of adsorption) especially since pyrene is m uch bigger than water. This would correlate with the study of the adsorption of arenes on s ilic a by Moldenhauer-Kopach (11) who observed that adsorption of pyrene onto silylated s ilic a was very limited. She was also able to show that the geometry of the adsorbing arene also dictated whether or not it would be adsorbed. Sim ilarly, the adsorption of pyrene must also be restricted to those pores which have the sam e size as the pyrene. Adsorption on these silylated surfaces is complicated by the j presence of the bulky trimethyl s ily l groups which have reacted with f ■ exposed silanols. There should be few or no surface silanols because I ! the surfaces were silylated twice. The surface area and porosity j | distribution did not change after the second s ily latio n . Silanols ! i | located deep in a pore may not be accessible by sily latio n and this | observation has been documented (10). Som e of these deep silanols are j i ! I the "reactive" silanols which have enhanced reactivity and adsorption j | j potential. ! l | The Polyani adsorption theory describes isopotential surfaces of i I ] adsorption (18,19). The potential is strongest near the surface and j f\ I would diminish as 1/r . Since nearly a ll surface silanols are j s ilylated , then the adsorption onto silylated silicas must be a pore j i i adsorption process. In contrast, adsorption onto the native s ilic a is j both a surface and a pore adsorption process because of the presence j I of both surface and pore silanols. The largest barrier to adsorption j on the silylated surface may be the presence of the bulky s j trimethyl s ily l groups which block pores. j The large amount of excimer emission that is seen on the j silylated surface may be partly caused by the decreased surface area j but as w e ll, the aggregation of pyrene molecules in or on top of a j pore with a "reactive" silanol group. At room temperature, the j 1 trimethyl si ly l groups blocking a pore may have a rotational frequency [ which makes the effective pore radius seem smaller and makes j adsorption more d iffic u lt. Cooling, may slow the trimethyl si ly l groups! t enough to increase the effective pore radius and increase the probability of adsorption into the pore, e.g. the rotational frequency of the trimethyl s ily l group must be matched with the encounter frequency of the pyrene molecule in order for adsorption into or out of a pore. Once the pyrene molecule is in the pore, "intrapore dynamics" then become most important for pyrene excimer formation. I f the pyrene is in a pore and its encounter frequency does not match with the rotational frequency of the trimethyl s ily l group then i t is trapped within a pore either as an isolated pyrene or perhaps as a static dimer or som e distorted dimer form. A lternatively, a simple interpretation would have the pyrenes being blocked from entering a pore by the bulky trim eth lysilyl groups. In this model, pyrenes would predominantly populate the surface and have a greater probability to aggregate and form dimers. The silylated surface may be envisioned as a homogeneous plane with areas of adsorption. This differs from the native surface which can be envisioned as a heterogeneous surface with many points of adsorption. The solution-like characteristics of the silylated surface has been documented with other procedures which have homogenize the surface (1,2,5,8,9,15). Others have noted that the monomer lifetimes increased with surface modification as was an observation in this study (5,15,17). I t is interesting that s ilylatio n makes the surface have solution-1 ike properties. The Arrhenius plots may be showing the effect of these trimethyl s ily l groups on adsorption or diffusion. The native Arrhenius plots 279 a ll show a c ritic a l temperature which was lower than the c ritic a l temperature of solution studies. However, the silylated Arrhenius plots did not show any tendency towards changing from "low" temperature behavior to "high" temperature behavior. A ll Arrhenius plots showed a small negative slope yielding to a greater negative slope which may be the barrier to diffusion in/out of a pore. There are no known studies to correlate this effect. I t was interesting that SS63 appears to be the most hydrophilic of a ll surfaces. This may be the result from the greater tendency to block the smaller pores upon s ilylatio n so that the "reactive" silanols deep in the pore are inaccessible (10). Adsorption into the pores of SS63 would present the pyrene with more silanols. Adsorption into small pores which are less than either dimension of the pyrene molecule, 7.2 A by 13.0 A, can be ruled out since Moldenhauer-Kopach has documented that arene adsorption depends on a match between the probe and pore geometries. However, i t is not clear how pyrene senses a silanol if it is contained in a pore. Most d iffic u lt to interpret and understand are the excimer decays. The monomer decays are re la tiv e ly straightforward, but each native surface has shown different excimer decay behavior at lower loadings and the silylated surface has shown different excimer decay patterns from the native excimer decays but at least are more consistent with each other. This points out the importance of comparing the surface features with the probe behavior. Three important observations can be made: the loading w ill dictate the 280 excimer decay; there is a fast component which can be attributed to | static dimer which shares a com m on decay at a ll temperatures; the j i j surface features w ill create special environments. At greater loadingsj t of pyrene, the excimer decays a ll behaved in the same manner, but i t ! was at the lower loadings, in which the emission spectra showed | comparable amounts of excimer and monomer emission intensities, which j i had the most unusual excimer decays. J I 5.4 Experimental 5.4.1 Materials (20) | i 5.4.2 Equipment (20) j 5.4.3 Methods (20) i - i ! i I | j | i j I i i i i i i j 1) 2) 3) 4) 5) 6) 7) 8) 9) 10) 11) 12) REFERENCES Synder, L.R., Separation Science, ^£2 & 3|, (1966)191-218 Synder, L.R., Ward, J.W., J. Rhys. Chem., 70(12), (1966)3941 Thomas, J.K., Chandrasekaran, K., J. of Colloid and Interface Sci., 100, (1984)116 Lochmuller, C.H., Colborn, A.S., Hunnicutt, M.L., Harris, J.M., JACS, 106, (1984)4077 Bauer, R., de Mayo, P., Ware, W., Natarajan, L., Can. J. Chem., 62, (1984)1279 Singer, L., Francis, C., Lin, J., Chem. Phys. L e tt., 94, (1983)162-167 Bauer, R., de Mayo, P., Ware, W., Natarajan, L., Chem. Phys. L e tt., 107(2), (1984)187 Kitahara, S., Takada, K., Sakata, T., Muraishi, H., J. of Col loid Interface Sci., 84(2), (1981)519 Okkrese, C., Phys. and Chem . Asp, of Adsorbants and Catalysts, B.G. Linsen ed., Academic Press, New York, (1970) Nishioka, G.M., Schranke, J.A., J. Colloid & Interfacial Sci. 105(1), (1985) 102 Moldenhauer-Kopach, K., Dissertation, University of Southern California, (1987) Uchida, K., Takahashi, Y., In t. J. of Quant. Chem., 18, (1980)310-305 282 13) Fischer, D., Narmdorf, G., V o lp e rt, A., JACS, 107, (1985)3368 14) Hara, K., de Mayo, P., Ware, W., Weedon, A., Wong, G., Wu, K., Chem. Phys. L e tt., 69(1), (1980)105 15) de Mayo, P., Organic Phototransformations in Nonhomogenous Media, ACS Symp. Ser. #278, M.A. Fox ed., (1984)1-19 16) Birks, J. B., Photophysics of Aromatic Molecules, Wiley- Interscience, N ew York, 1970 and references therein 17) Birks, J.B., Braga, C.L., Lumb, M.D., Proc. Royal Soc. London, Ser. A, 283, (1965)83 18) Gregg, S.J., Sing, K.S.W., Adsorption, Surface Area and Porosity, 2nd ed., Academic Press, New York, (1982) 19) Adamson, A.W., Phys. Chem. of Surfaces, 4th ed., W i ley and Sons, N ew York, (1982) 20) The materials, equipment and methods are as described in the Experimental section of Chapter 4 with the only substitution of the silylated s ilic a for the native s ilic a 283 Selected Bibliography Adamson, A.W., Phys. Chem. of Surfaces, 4th ecL, W i 1 ey and Sons, N ew York, (1982) Aoudia, M., Rodgers, M., Wade, W.H., J. Phys. Chem. 88, (1984) 5008 Bauer, R., de Mayo, P., Ware, W., Natarajan, L., Chem. Phys. L e tt., 107(2), (1984)187 Birks, J. B., Photophysics of Aromatic Molecules, Wiley- Interscience, N ew York, 1970 and references therein Brunauer, S., Emmett, P.H., T e lle r , E., J ACS, 60, (1938)309 Brunauer, S., Mikhail, R.S.H., Bodor, E., J. Colloid & Interfacial Sci. 24, (1967) 45131) de Mayo, P., Organic Phototransformations in Nonhomogenous Media, ACS Symp. Ser. #278, M.A. Fox ed., (1984)1^T9 Dong, D.C., Winnik, M.A., Photochem. and Photobio., 35, (1982)17-21 Gregg, S.J., Sing, K.S.W., Adsorption, Surface Area and Porosity, 2nd ed., Academic Press, New York, (1982) Hara, K., de Mayo, P., Ware, W., Weedon, A., Wong, G., Wu, K., Chem. Phys. L e tt., 69(1), (1980)105 Moldenhauer-Kopach, K., Dissertation, University of Southern California, (1987) Nishioka, G.M., Schranke, J.A., J. Colloid & Interfacial Sci. 105(1), (1985) 102 Okkr^se, C., Phys. and Chem . Asp, of Adsorbants and Catalysts, v B.G. Linsen ed., Academic Press, New York, (19701 Singer, L., Francis, C., Lin, J., Chem. Phys. L e tt., 94, (1983)162-167 Suib, S.L., Kostapapas, A., JACS, 106, (1984)7705 Synder, L.R., Separation Science, 1(2 & 3) (1966)191-218 284 Synder, L.R., Ward, J.W., J. Phys. Chem., 70(12), (1966)3941 Thomas, O.K., Chem. Reviews, 80(4), (1980)283 Turro, N., Modern Molecular Photochemistry, The Benjamin/Cummings Co., Menlo Park, 1978 Yamazaki, T., Tamai, N., Yamazaki, I., Chem. Phys. L e tt., 124(4), (1986)326 285

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Chemiluminescence of 9,10-dihydro-10-methylacridylic imidazolide and 9-alkylidene-10-methylacridans

PDF

Application of nonlinear spectroscopic techniques to the studies of molecular electronic states and reaction dynamics.

PDF

A highly efficient and general method for the chemical synthesis of beta,gamma-fluoro- and difluoromethylene analogs of purine and pyrimidine 5'-ribo- and deoxyribonucleotides

PDF

Adsorption of arenes and reactivity of pyrene and diphenylanthracene with singlet oxygen on different silicas

PDF

A study of the double sulphates of some rare-earth elements with sodium, potassium, ammonium and thallium

PDF

Computer simulation of chemical reactions in aqueous solutions and biological systems

PDF

A study of the kinetics of the reaction of sulfenyl halides with olefins

PDF

Crystallographic studies on new Pt(IV) anti-cancer agents and their reaction products

PDF

A preliminary investigation of the formation of molecular addition compounds by phosphorus chloronitride

PDF

A synthetic model approach to H-bonded oxyhemoglobin

PDF

Crystallographic stydies on Zn(2), Mn(2) and Pt(2) complexes of phosphonates

PDF

A.The anions of C(60) and its pyrrolidine derivatives B.The search for hydridocobaloximes

PDF

Basic studies in the analytical chemistry of beryllium

PDF

An investigation of the effects of output variability and output bandwidth on user performance in an interactive computer system

PDF

A study in hygrometry

PDF

Chemically modified electrodes and microelectrodes

PDF

Computer simulation of polar solvents with the SCAAS model

PDF

A comparative study of the chemistry and biology of Heparin

PDF

A classical mechanical method of analysis and periodic orbit search: Applications to HCN/HNC, C1HC1, FH2, and their deuterates

PDF

A study of phonetic (sound) reinforcement and generalization learning.

Asset Metadata

Creator Yee, Herman T (author)

Core Title A photophysical study of pyrene adsorbed onto silicas of variable surface area and porosity

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Chemistry

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag chemistry, organic,OAI-PMH Harvest

Language English

Advisor Singer, Lawrence (committee chair), [illegible] (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-652825

Unique identifier UC11342717

Identifier DP21981.pdf (filename),usctheses-c17-652825 (legacy record id)

Legacy Identifier DP21981.pdf

Dmrecord 652825

Document Type Dissertation

Rights Yee, Herman T.

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA