SURFACE MODIFICATION of SILICON THROUGH THERMAL ANNEALING and RINSING of SOLVENT CAST POLYSTYRENE FILMS a Thesis Presented to Th

SURFACE MODIFICATION OF SILICON THROUGH THERMAL ANNEALING AND RINSING OF

SOLVENT CAST POLYSTYRENE FILMS

A Thesis

Presented to

The Graduate Faculty of The University of Akron

In Partial Fulfillment

of the Requirements for the Degree

Master of Polymer Engineering

Steven V. Kalan

August, 2011

SURFACE MODIFICATION OF SILICON THROUGH THERMAL ANNEALING AND RINSING OF

SOLVENT CAST POLYSTYRENE FILMS

Steven V. Kalan

Thesis

Approved: Accepted:

Co-Advisor Department Chair Dr. Kevin Cavicchi Dr. Sadhan Jana, Ph.D

Co-Advisor Dean of the College Dr. Alamgir Karim Dr. Stephen Z.D. Cheng, Ph.D.

Faculty Reader Dean of the Graduate School Dr. Mark Soucek Dr. George R. Newkome, Ph.D.

Date

ABSTRACT

Surface modification is an important process for many applications. It has been found that an ultrathin polymer film can be generated on silicon substrates with a native oxide layer by spin-coating, thermally annealing, and washing films of thiol terminated polystyrene (PS-SH) or anionically polymerized polystyrene (PS). This is a useful modification technique as it requires little pretreatment of the substrate or specialized polymerization chemistry for the case of PS films. These ultrathin films were characterized using contact angle (CA), x-ray reflection (XR), and x-ray photoelectron spectroscopy (XPS).

Experimental results indicated that a thin polymer film formed characteristic of the grafted polymer for both PS-SH and PS due to the physical adsorption of the PS chains. It is suggested that the benzene rings of polystyrene all share a bond with the silicon surface creating a relatively overall strong bond from the polymer to the silicon wafer. Differences in the PS-SH and PS films are consistent with additional end- tethering of the PS-SH chains. Spin coated films underwent different levels of rinsing with multiple solvents. After all levels of rinsing, the polymer was still present on the surface, indicating a relatively strong physical bond between the polymer and silicon oxide. Control experiments using non-annealed versus annealed polystyrene polymers showed that grafting depends on the mobility of the chains during thermal annealing of

ii the polystyrene films. Two other samples, polytert-butyl styrene (PtBS) and block copolymer poly (styrene-b-dimethylsiloxane) (PS-b-PDMS), were also used and compared with PS to show the effects of thermal annealing of PS to a silicon surface. XR results show an increase in residual layer thickness with molecular weight (Mw).

Analysis of water contact angle of PS residual layers as a function of annealing time and annealing temperature illustrates directly proportional relationships. Further surface modification analysis by ultra violet light and ozone (UVO) treatment demonstrates that the surface energy can be roughly tuned for a small range of UVO treatment times.

iii

ACKNOWLEDGEMENTS

I would like to thank first and foremost my advisors Dr. Cavicchi and Dr. Karim for their continued support and guidance. I also would like to thank Dr. Christopher

Stafford for his role at NIST in collecting valuable data for XR and XPS analysis. Mr. Jon

Page for allowing me the time to use the Polymer Science contact angle equipment. My fellow group members for their admirable kindness and time they gave to demonstrate or help with aspects of my research, especially Yuqing Liu whom would drop what he was doing at any time to help another. To these people I express my deepest gratitude, for I would not have accomplished this without them.

TABLE OF CONTENTS

Page

LIST OF TABLES ...... vii LIST OF FIGURES ...... viii CHAPTER I-INTRODUCTION ...... 1 1.1 Covalently Bonded Polymer Brushes and Mats ...... 7 1.1.1 Grafting-To Approach ...... 11 1.1.2 Grafting-From Approach ...... 15 1.2 Non-Covalent Bonding of Polymer Brushes and Mats ...... 21 1.2.1 Formation of Polymer Mats by Non-Covalent Bonding from Solution ...... 21 1.2.2 Guiselin Brush ...... 22 1.2.3 Polymer Brush Through Non-Covalent Bonding ...... 24 1.2.4 Polymer Adsorption Through Thermal Annealing ...... 26 II-EXPERIMENTAL SECTION ...... 28 2.1 Materials ...... 28 2.1.1 Anionically Polymerized Polystyrene ...... 28 2.1.2 RAFT Polymerized Polystyrene ...... 29 2.1.2.1 RAFT Agent ...... 29 2.1.2.2 Styrene Monomer/n-Butylamine ...... 30 2.1.3 PS-b-PDMS ...... 30 2.1.4 PtBS ...... 30 2.1.5 Silicon Substrate ...... 30 2.2 Experimental Process ...... 31 2.2.1 RAFT Polymerization of Polystyrene ...... 31 2.2.2 Aminolysis of RAFT Polymerized Polystyrene ...... 32 2.2.3 Preparation of samples ...... 33

2.2.3.1 Sample Preparation ...... 33 2.2.3.2 Levels of Rinsing ...... 34 2.3 Equipment/Measurement Methods ...... 34 2.3.1 Gel Permeation Chromatography (GPC) ...... 34 2.3.2 Contact Angle (CA) ...... 35 2.3.3 Optical Microscopy (OM) ...... 36 2.3.4 X-ray Photon Spectroscopy (XPS) & X-ray Reflection (XR) ...... 36 III-SURFACE MODIFICATION OF SILICON OXIDE VIA THIOL TERMINATED PS (COVALENT) ...... 37 3.1 Introduction ...... 37 3.2 Materials and Water Contact Angle ...... 38 3.3 Effect of Different Levels of Rinsing ...... 42 3.4 Effect of Different Solvents ...... 44 3.5 Topological Changes (OM) ...... 46 3.6 Dewetting (OM) ...... 50 3.7 Control Experiments (PS-SH vs PS_RAFT vs PS) ...... 54 3.8 Summary ...... 56 IV-SURFACE MODIFICATION OF SILICONO OXIDE VIA THERMALLY ANNEALED PS (NON-COVALENT) ...... 58 4.1 Introduction ...... 58 4.2 Effect of Different Solvents and Different Levels of Rinsing ...... 59 4.3 Adsorption of PS vs PtBS vs PS-b-PDMS ...... 64 4.4 X-Ray Reflectivity (XR) and X-ray photon spectroscopy (XPS) ...... 65 4.5 Effect of Annealing Time and Annealing Temperature ...... 69 4.6 UVO Surface Energy Modification...... 71 4.7 Summary ...... 73 V-CONCLUSIONS ...... 75 VI-REFERENCES ...... 77

LIST OF TABLES

Table Page

Table I. 1: Types of polymer layers bound to surface or interface ...... 6

Table III. 1: RAFT polymerized (PS_RAFT) and thiol terminated PS (PS-SH) material data ...... 40

Table III. 2: Water contact angle measurements of PS versus PS-SH annealed and rinsed in cyclohexane and dimethylformamide (DMF) ...... 55

Table III. 3: Water contact angle measurement of PS (9k vs 123k) versus PS-SH annealed and rinsed in different solvents. Bare wafer values are water contact angle due to any solvent adsorption...... 55

Table III. 4: Water contact angle measurement of 9k pure PS refluxed in different solvents at different temperatures. 37˚C is the theta temp (Tθ) for diethyl malonate (DM)...... 56

Table IV. 1: Water contact angle of as-cast films rinsed in selected solvents at three different

Mw ...... 60

Table IV. 2: Water contact angle of annealed films rinsed in selected solvents at four different

Mw ...... 61

Table IV. 3: Water contact angle of 123k PS annealed films before (top) and after (bottom) rinsing with Toluene. Standard error is less than 0.5 for all cases...... 63

vii

LIST OF FIGURES

Figure Page

Figure I. 1: Examples of polymer brushes in three different polymer systems ...... 4

Figure I. 2: Illustration of chemical reaction from thermal annealing of a (a) thiol terminated polymer and (b) hydroxyl terminated polymer on a silicon surface with native oxide layer ...... 7

Figure I. 3: Schematic of structures formed by tethered polymer chains on substrate at low grafting density in (a) good solvent and (b) poor solvent...... 9

Figure I. 4: End tethered homopolymer chains in (a) good solvent and (b) poor solvent, for high grafting density ...... 9

Figure I. 5: Polymer brushes not consisting of just single homopolymer: (a) random copolymer brush, (b) mixed homopolymer brush, and (c) block copolymer brush ...... 11

Figure I. 6: Illustration of grafting-to approach where ‘R’ represents the end-function of the terminated polymer...... 12

Figure I. 7: Illustration of grafting-from method where dots are representative of monomer units ...... 16

Figure I. 8: Schematic representing effect of solvent on mixed homopolymer brush  (a) non- selective brush, (b) selective for polymer A, (c) selective for polymer B ...... 19

Figure I. 9: Schematic of polymer brush of block copolymer by physisorption ...... 25

Figure II. 1: RAFT agent DBTC used for RAFT polymerization ...... 29

viii

Figure II. 2: Cartoon of silicon wafer ...... 31

Figure II. 3: RAFT polymerized PS using DBTC as RAFT agent ...... 32

Figure II. 4: Thiol terminated PS ...... 33

Figure III. 1: GPC trendlines for three different Mw of RAFT polymerized PS (PS_RAFT) and thiol terminated PS (PS-SH) ...... 40

Figure III. 2: Water contact angle images. Solvent used is toluene. Silicon wafer has no surface changes or modifications...... 41

Figure III. 3: Effect of molecular weight on water contact angle of annealed films rinsed in toluene overnight. Standard error bars based on 6 measurements...... 42

Figure III. 4: Effect different levels of rinsing with toluene on annealed PS-SH films: submerged in toluene for 15 minutes (F), submerged in toluene overnight (O), sonicated for 1 hour (S), and soxhlet extraction in toluene overnight (Sox). Error bars are standard error based on 6 measurements...... 43

Figure III. 5: Effect of solvents on water contact angle using toluene (Tol.), tetrahydrofuran (THF), acetone, and chloroform...... 45

Figure III. 6: Effect of solvent chosen to spin cast films from and subsequently rinsed with ...... 46

Figure III. 7: Cylindrical forming block copolymer oriented (a) perpendicular and (b) parallel to the substrate ...... 47

Figure III. 8: Optical Microscope images at 500x for PS-b-PMMA (30k:30k) films thermally annealed (170˚C) overnight on (a) bare wafer, t = 47nm and (b) 2k PS-SH residual layer, t = 46 nm ...... 49

Figure III. 9: Domain spacing for PS-b-PMMA films (PS-black, PMMA-white) with smooth topology on (a) silicon oxide and (b) PS brush modified silicon substrate ...... 50

Figure III. 10: Optical Micrscope images at 100x for (top) 5.3k PS_RAFT, t = 77nm and (bottom) 3.3k PS-SH, t = 49 nm after (a) as-cast and (b) thermally annealed at 170˚C for 24 hours ...... 51

Figure III. 11: Water contact angle of PS-SH () and PS_RAFT (,) after rinsing with toluene. After overnight in toluene PS_RAFT sonicated in toluene for 1 hour ...... 53

Figure III. 12: Water contact angle of PS-SH () and anionically polymerized PS () after rinsing with toluene overnight...... 53

Figure IV. 1: Water contact angle vs Mw for annealed (solid) and as-cast (open) films after rinsing with selected solvents overnight. Solvent: Toluene (), THF (), Cyclohexane (),

CHCl3 (). Error bars are standard error using six measurements...... 61

Figure IV. 2: Water contact angle of PS annealed films rinsed at toluene at different levels ...... 63

Figure IV. 3: Bar plot of water contact angle for materials PS, PtBS, PS-PDMS. Scale bars are standard error using six measurements. Film refers to annealed films. Rinsed refers to annealed films that underwent Toluene Wash overnight...... 65

Figure IV. 4: Thickness vs Mw of anionically polymerized PS via XR...... 66

Figure IV. 5: Thickness versus Mw of pure PS and PS-SH ...... 67

Figure IV. 6: XPS data for Carbon, Oxygen, and Nitrogen percent in the PS layer at different Mw. Standard error bars are present for three measurements but cannot be seen due to the small values ...... 68

Figure IV. 7: Water contact angle vs annealing time for 200k pure PS after rinsed in toluene overnight. Annealed at 170˚C...... 70

Figure IV. 8: Water contact angle vs annealing temperature for 200k pure PS after rinsed in toluene overnight. Annealed for 18 hours...... 71

Figure IV. 9: UVO effects on PS film and PS residual layers. Closed symbols are water contact angles before any UVO exposure. Blue squares (,) are PS residual layers from a toluene rinsed PS film...... 73

CHAPTER I

INTRODUCTION

Surface modification and functionalization is a useful tool in controlling the surface properties of substrates that has broad applications in adhesion, lubrication, wetting, and colloidal stabilization [1-14]. Polymer coatings are a useful way to achieve this type of modification.

Silicon surfaces have attracted a great amount of interest in the field of photonics and electronics due to its semiconducting behavior [15]. Integrated circuits, microprocessors, and photovoltaic sheets are a few examples that utilize silicon surfaces

[16,17]. Most electronic devices today likely consist of silicon parts. Silicon is also widely used for the spin coating of polymer films, a process that is used in the fabrication of some of these devices.

Similarly, block copolymers, or layered polymer films, are becoming popular in the field of polymer light emitting diodes, using layers as electron donors and electron acceptors [1]. Nanotechnology and self assembly interest in orienting nanostructure block copolymers can be used with block lithography in patterning templates [1]. These block copolymers at the nano-level offer the possibility of nano-structured polymer

1 membranes [20]. Polymer coatings could help with surface passivation as well as changing the surface energy of the silicon substrate for layer deposition or second coatings.

Layers can be either strongly or weakly attached to a surface or interface.

Within the literature the former is often termed chemisorption, whereas the latter is termed physisorption. Chemisorption is described as a strong bond between two species often through covalent bonding or ionic bonding [21]. Physisorption is considered reversible having an unstable bond, usually through Van der Waals forces, that can be easily broken [22]. Physisorption can also be viewed as an adsorption between two species. Since the focus of this work is on these two aspects, but more importantly the bonding nature, we instead focus on the polymer layers that are covalently or non-covalently bonded to the substrate. Two ways to covalently bond a polymer to a substrate is end tethering to form a polymer brush or bonding along the polymer backbone to form a polymer mat.

A polymer brush is by definition an assembly of polymer chains, which are tethered by one end to a surface or interface [11]. A more detailed definition of a polymer brush is one that consists of end-tethered polymer chains stretched away from the substrate so that in the given solvent the brush height is an appreciable fraction larger compared to the end-to-end distance of the same non-grafted chains dissolved in the same solvent [12]. While one of these definitions goes into greater detail about the stretching of chains, both imply that the chains must somehow be end tethered to a

2 surface or interface. Examples of polymer brushes within polymer systems can include

(but are not limited to) polymer micelles, end-grafted polymers, and adsorbed diblock copolymers. The latter is a polymer brush formed through non-covalent bonding.

Figure I.1 is a schematic of these three types of polymer brushes that can exist.

All three representations have a chain configuration which shows some sort of deformation. In polymer brush systems the presence of solvent can either be included or absent. Not only will the presence of a solvent effect the polymer brush configuration but also the type of solvent used will have an impact.

Polymer micelles are an example of a polymer brush in which the polymer is not tethered to a solid surface but rather at the interface between the two polymer chains.

In a selective solvent, which dissolves one block of a copolymer but not the other, micelles will form. Micelles do not always have to be spherical in shape as they can become long, tube like, or bilayer flat structures that can curve around to form vesicles

[23]. Polymer micelles can be used to transport insoluble molecules through a solvent.

Figure I. 1: Examples of polymer brushes in three different polymer systems

Covalent tethering of polymer chains is often done through two types of processes known as ‘grafting-to’ and ‘grafting-from’. Both approaches, especially the grafting-from approach, are extensively discussed within the literature for different materials, substrates and methods as a means of modifying the substrate [14, 23-30].

While many of these approaches and methods are sufficient in creating a covalently bonded polymer to a certain substrate, they are often done through chemical modification of one or more parts involved in the process, whether it is the polymer itself or the substrate that the polymer is attached to.

As mentioned earlier, a polymer can also be attached to a substrate by non- covalent bonding. This can be accomplished through dip coating, spin coating, or adsorption of polymer solutions/melts and appropriate annealing and washing of the substrate. From Figure I. 1, an example of a polymer brush fabricated using this non- covalent adsorbing behavior is depicted. By having a block copolymer with one strongly

4 adsorbing block, it is possible to attach the adsorbing block while the other block stretches away from the surface.

By either a covalent or non-covalent bonding approach, it is possible to force polymer chains to attach at one end of the chain to the surface of the substrate or interface. When the density of strongly tethered chains is high enough, the chains reduce overlapping and the free volume by stretching away from the surface to form what appears to look like a brush on the surface [13,14,31-33]. While both methods will generate what is known as a polymer brush, they individually have different advantages and disadvantages consequently causing the brushes to take on slightly different physical properties. However, polymer brushes are not the only type of adsorbed layer.

It is possible to have chains covalently bonded or non-covalently bonded along the chains of the polymer. For the purpose of this work, chains bonded along the polymer chain will be referred to as polymer mats. Table I. 1 illustrates the different types of coatings for a surface or interface.

In this approach, covalent and non-covalent bonding of polystyrene (PS) homopolymer is investigated on silicon substrates with native oxide layers. A covalently bonded homopolymer brush is the first target of this research. It falls into category (1) of Table I. 1. Thiol terminated polystyrene is investigated as the brush forming polymer by a grafted to reaction at elevated temperature under vacuum. While thiol terminated

PS has been shown to attach to gold surfaces, there has been no known studies using a silicon oxide surface [34-36]. The hypothesis is the thiol bond and the hydroxyl units of

5 the native oxide layer on silicon will react to bond sulfur to the silicon oxide surface, similar to the attachment of a hydroxyl end-functionalized chain illustrated in Figure I. 2.

RAFT polymerization is used to generate PS that is then cleaved in half with aminolysis of the chain to give thiol terminated PS. Since RAFT polymerization is a simple process which offers a variety of polymers that can be fabricated, it is plausible to attach other polymer chains using the sulfur to silicon bond.

Covalent Non-Covalent Block copolymer End-functional Polymer Brushes (1) brushes (2)

Side-functional Polymer Mats (3) Polymer Mats (4)

Table I. 1: Types of polymer layers bound to surface or interface

It will be shown in Chapter III, that while the data shows support of an end tethered polymer brush, additional non-covalent polymer adsorption along the chain also occurs. This type of physically adsorbed layer, category (4) of Table I. 1, has been seen, however there has been very little study into the topic. This type of phenomena may lead to an alternate and easier modification route. Therefore a second study was carried out to investigate the non-covalent adsorption of anionically polymerized PS in

Chapter IV to form a polymer mat. Additional experimental procedures using UVO to tune the surface energy of the polymer layer and therefore the surface energy of the substrate are investigated in Chapter IV as well.

Figure I. 2: Illustration of chemical reaction from thermal annealing of a (a) thiol terminated polymer and (b) hydroxyl terminated polymer on a silicon surface with native oxide layer

The rest of this thesis is organized as follows. The remaining portion of the introduction provides background information and a review of the relevant literature.

Chapter II will explain the materials and experimental procedures used in this work.

Chapter III will display and discuss the results for grafted to thiol terminated PS for surface modification. Chapter IV will display and discuss results for PS adsorption for surface modification. Chapter V provides a summary and conclusions and suggests future possibilities and thoughts for experiments.

1.1 Covalently Bonded Polymer Brushes and Mats

Chemisorption, as mentioned earlier, refers to the strong adsorption of one species on another through valence forces such, as covalent bonding or ionic bonding [37-40].

Due to the nature of this type of adsorption it is much stronger and more stable than typical physisorption. Often times the chemisorption process is irreversible due to the bonding behavior since desorption may be hard to accomplish. End-tethered polymer

7 brushes are most often found with chemisorption. This review will focus on covalently bonded polymers.

While there are two different methods that can be utilized in generating these polymer brushes there are other aspects of a polymer brush that can be observed besides the method used. For example, the type of solvent that a polymer brush is formed in, or exposed to, will affect the resulting structure of the polymer chains

[11,21]. In a good solvent, the polymer chain will extend out further due to the attractive forces of the solvent and polymer chain. Since the polymer chains try to avoid contact with one another to have more contact with the solvent, the resulting structure is more stretched and appears as a mushroom shape. In a poor solvent, the polymer chain will be a flattened layer near the substrate if the forces are stronger between the polymer and substrate versus the polymer and the solvent. This structure will appear as a pancake structure. These two types of structures can be seen in Figure I. 3. While the solvent has an impact on the form of the polymer chains at the substrate there are other factors leading to the final structure. Brush properties can be affected by the grafting density, chain length, and chemical composition of the chains as well [21]. For example, when the grafting density increases for mushroom structure the polymer chains are crowded and extend away from the substrate to form an end-grafted brush like structure seen in Figure I. 4 (a). On the other hand, when the grafting density increases for the pancake like structure the chains stretch only so much resulting in a cluster like formation which can be seen in Figure I. 4 (b).

Figure I. 3: Schematic of structures formed by tethered polymer chains on substrate at low grafting density in (a) good solvent and (b) poor solvent

Figure I. 4: End tethered homopolymer chains in (a) good solvent and (b) poor solvent, for high grafting density

So far the polymer brushes mentioned have been composed of just a single homopolymer, however, random copolymer, mixed homopolymer, and block copolymer brushes exist as well. In fact, as will be seen later, block copolymers can also be used in physisorption to form a polymer brush. A random copolymer brush is when the tethered polymer chains are made up of two different repeat units randomly spread along the polymer chain [41]. A mixed homopolymer brush is when two different

9 homopolymer chains are end tethered along the substrate surface [42]. A block copolymer brush consists of a block copolymer where one end is end tethered to the substrate while the other end is exposed to the surface [43]. These types of brushes are depicted in Figure I. 5. The solvent effects are also present within these copolymer systems as well. With the use of these systems, different surface energies can be obtained, as well as neutral surfaces which will be talked about further into this work.

For example, if polymer A is difficult to tether to a certain surface but can be tethered to polymer B, which can be tethered to that surface, then a modified surface with polymer

A exposed to the air interface is still accomplished.

In the upcoming sections there will be a focus on the two approaches of binding polymer chain ends to a solid surface or substrate. The grafting-to approach and the grafting-from approach are explained and compared here.

Figure I. 5: Polymer brushes not consisting of just single homopolymer: (a) random copolymer brush, (b) mixed homopolymer brush, and (c) block copolymer brush

1.1.1 Grafting-To Approach

The grafting-to approach involves tethering the polymer chain to an already modified substrate or functional chain end by exposing the surface to the polymer solution or melt [13,14,44-47]. This is one method to form a covalent bond between the surface and the polymer chain end which can be resistant to chemical and/or environmental conditions. An illustration of this method is depicted in Figure I. 6.

In order to accomplish the grafting-to method, several different mechanisms for modifying the polymer and/or substrate have been utilized. Self-assembled monolayers

(SAMs) are an effective way of modifying the substrate that act as a coupling agent for the surface and chain to increase reactivity, thus increasing the strength of the chemical attachment [14,48,49]. Alternate chain ends are continually being explored as ways of coupling the chain end to an accepting surface.

Huang and Penn [44] list some of the ways chain ends can encourage bonding to a surface including polar groups [50,51], charged groups [52,53], chemically reactive groups[46,54,55], and blocks of adsorbing segments different from the rest of the chain

[56,57]. Since the polymers are already synthesized prior to the formation of the polymer brush, advantages of this method include accurate control of the molecular weight and polydispersity as well as the ability to run this approach in ambient conditions [44] and ability to tether mixed brushes through the surface [58]. The largest disadvantage that is observed in the grafting-to approach is the lack of density of the polymer chains that are grafted.

Figure I. 6: Illustration of grafting-to approach where ‘R’ represents the end-function of the terminated polymer.

Lemieux et al [14] state that steric constraints and kinetic factors [46] as well as limitations of the number of reactive sites due to space [46,59,60] are reasons that the density brush is restricted. The advantages and disadvantages are investigated extensively as well as ways to fix the disadvantage of low density within the grafting-to approach. Others look at switching properties of mixed brushes through good vs. bad

12 solvents. Huang et al. for instance looked at the concentration of a polymer solution when using the grafting-to approach and found that there was a maximum concentration of solution for which there is no additional grafting density onto the substrate [44]. Taylor et al. goes into a greater detail about some of the disadvantages involving the procedures of the grafting-to approach including a melt approach or solvent approach [61]. The melt approach involves heating a thin grafted layer past its melt transition point which can sometimes be aided by the presence of a solvent to lower the melt transition temperature and the solvent approach is one in which the polymer chain is grafted, or ‘grown’, in solvent at the theta (θ) temperature so the volume interactions are suppressed. Poor solvents have been chosen as well to encourage grafting to the substrate when in solution. Taylor et al. goes on to explain that these procedures have disadvantages described by a need of high temperatures or control of vapor pressure or use of poor solvents and solvents at the θ temperature. His work employed a method utilizing thiol end grafted poly ethylene oxide (PEO-SH) that can be grafted on a gold surface in an aqueous solution without the use of any elevated temperature with grafting density control up to 0.3 chains/nm2.

Yu et al. used a grafting-to approach to form a patterned self adaptive poly

(styrene - 2-vinyl pyridine), (PS-P2VP) which had switching properties of the mixed polymer brush depending on solvent exposure [62]. In this study, the compound 3- glycidoxypropyl trimethoxysilane (GPS) was chemisorbed on the surface of a silicon wafer that the carboxyl-terminated PS and P2VP (PS-COOH, P2VP-COOH) could graft to.

Here exists one example of how the surface is modified as well as the polymer chain

13 ends. By exposing the samples to preferential solvents one could obtain switching properties of the polymer brush. If the system is exposed to a solvent such as toluene, which is selective for PS, the PS chains will extend outward while the P2VP chains will collapse leaving what appears to be a PS surface. By having the ability to switch the brush properties of the surface, Yu et al. was able to obtain water contact angles of 64˚ to 85˚ (P2VP to PS).

As mentioned earlier, the control of molecular weight and molecular weight distribution (polydispersity, PDI) is a major advantage of the grafting-to approach. The types of polymerizations that can be utilized to achieve this advantage with an end- functional polymer chain include living anionic, cationic, radical, group transfer, and reversible addition-fragmentation transfer (RAFT) [11]. One study used living radical polymerization to synthesize hydroxyl terminated random poly (styrene-r-methyl methacrylate) (PS-r-PMMA) that could be grafted on a silicon wafer that had been modified with silanol groups [41]. Due to the nature of the random copolymer brush structure, the surface was found to be neutral towards a block copolymer film of PS and

PMMA (PS-b-PMMA) being useful in controlling the orientation of the block copolymer.

Bergbreiter et al. used living radical polymerization as well to make diaminopoly (tert- butyl acrylate) which was then acid terminated and tethered onto an oxidized polyethylene (PE) surface [63]. Ebata et al. employed anionic polymerization to make polysilanes which were end terminated with a butyl lithium anion and subsequently tethered to a quartz substrate with reactive sites along the surface [64]. A slightly different approach to covalently grafting the polymer ends to the surface involved the

14 use of UV light utilized by Prucker et al. [65]. Within their work, PS or poly

(ethyloxazoline) was spun cast on a silicon wafer modified by attaching 4-(3’- chlrodimethylsilyl) propyloxybenzophenone which then bonded the chains to the surface through photochemical attachment using UV light.

Examples listed previously are only a few among many that are found within the literature that utilizes the grafting-to approach. From the examples given it can be seen that either the polymer chain itself must contain a specific terminal functional group or a particular surface or surface modification must be used to obtain complementary functional groups for bonding. In doing so, the polymer chains are able to covalently bond to the surface. While polymerization methods are given for the grafting-to approach, that does not exclude them being useful for the grafting-from approach. As will be discussed in the next section, the grafting-from approach will improve upon the disadvantages of the grafting-to approach. However, we will see that neither technique is perfect and that the grafting-from method will also have its own disadvantages.

1.1.2 Grafting-From Approach

The grafting-from approach improves upon the disadvantage of the grafting-to approach but also has its own disadvantages to be overcome. Unlike the grafting-to method where the polymer is already synthesized before the brush is formed, the grafting-from approach ‘grows’ the polymer from initiating sites located at the surface.

An illustration of this method is depicted in Figure I. 7. By exposing the surface which

15 has immobilized initiators on it to a monomer solution, it is able to polymerize from the surface to form polymer chains which are end tethered to the surface [13,66,67]. By applying a self assembled monolayer (SAM) to the substrate (or other treated surface/monolayer of initiators), the monomer can then be polymerized from the surface by (while not limited to) living free radical polymerization such as, atom-transfer radical polymerization (ATRP), and/or RAFT polymerization.

Figure I. 7: Illustration of grafting-from method where dots are representative of monomer units

Husseman et al used silicon wafers with alkoxyamine initiators and α- bromoester initiators to polymerize PS and PMMA by means of living free radical polymerization [29]. The procedure was shown to have a stronger control over molecular weight as well as the PDI but there is complex chemistry involved and high temperatures for polymerization. A similar strategy by Devaux et al used a Langmuir-

Blodgett deposition technique [68], rather than the typical SAMs, to graft initiators on the surface of a silicon wafer which then PS was polymerized from [69]. The polymerization was shown to graft high densities and also show better control over the molecular weight and PDI, however, many steps and chemical processes are involved

16 while a necessary environment is needed, such as argon for polymerization within this system.

Jeyaprakash et al polymerized polymer brushes using the grafting-from approach using the atom transfer radical polymerization (ATRP) technique as a way to have better control over the PDI of the polymer brush [70]. In their work (3-chlorodimethylsilyl)allyl-

2-bromopropionate was immobilized on the surface of a silicon wafer (SiO2) from which

PS was then grafted from, all of this taking place through a series of steps and chemistry, to form robust thick polymer brushes. Chen et al also used ATRP to prepare densely grafted PS brushes on silicon wafers after treating the surface with isotropic oxygen plasma (IOPT) to induce hydroxyl layers that a 2-bromo-2mthylproplonyl bromide layer could then be immobilized on [71]. Chen goes on to show that introduction of water or toluene will change the behavior of the polymer chains, changing the surface structure and morphology, exhibiting switching properties through solvent responsive behavior.

Lemieux et al shows similar results of solvent responsive behavior and reversible switching properties of a mixed homopolymer brush consisting of poly (methyl acrylate)

(PMA) and poly (styrene-co-2,3,4,5,6-pentrafluorosytrene) (PSF) through the use of selective solvents [13,14]. Since PMA is more a rubbery polymer where PSF is a glassy polymer, the switching behavior gives considerably different properties when introduced to different solvents selective for each polymer. Sidorenko et al describes how the grafting ratio of a mixed homopolymer brush consisting of PS and poly (2-vinyl pyridine) (P2VP) can be varied by controlling the time of grafting reaction for each monomer since first PS is grown from the surface and subsequently the P2VP is grown

17 afterwards [72]. Figure I. 8 illustrates this responsive behavior for a mixed homopolymer brush. It can be seen that the solvent choice will greatly affect what type of properties the surface will undertake. While there are other responsive behaviors involved with other systems they will not be investigated here. Minko further details the responsive behavior of mixed homopolymer and copolymer systems [12].

Reversible addition fragmentation chain transfer (RAFT) polymerization has also been executed successfully in the formation of polymer brushes. Tsujii et al showed that it was possible to graft PS from silica particles [73]. Baum and Brittain successfully polymerized PS, PMMA, poly (N,N-dimethylacrylamide) (PDMA) as well as block copolymers PS-b-PDMA and PDMA-b-PMMA by grafting from a silica surface with immobilized initiators to produce densely grafted surfaces [74]. Both Tsujii and Baum show that the use of RAFT polymerization has shown improvement and promise in controlling the molecular weight of the polymer chains while maintaining a fairly narrow

PDI.

A common theme seen from the literature is that the density of polymer brushes can be greatly increased using the grafting-from approach. Due to the increase in initiator sites at the surface for polymerization to take place, the density of the polymer brushes can increase and the thickness of the brush can be maximized as well. Other than using a SAM, or other monolayer, it has been shown that plasma and/or glow discharge treatment can be used to implement initiator groups onto the surface of the substrate. Ito et al used glow discharge to induce peroxide initiators on a fluoropolymer

18 membrane that they could graft poly [3-carbamoyl-1-(p-vinylbenzyl) pyridium chloride] on to [75]. Iwata et al also introduced peroxide initiators onto a Nuclepore membrane for poly (acrylic acid) (PAA) but did so through plasma treatment [76].

Figure I. 8: Schematic representing effect of solvent on mixed homopolymer brush  (a) non- selective brush, (b) selective for polymer A, (c) selective for polymer B

Densities of 2-10 mg/m2 with a thickness around 10 nm [14,32,77] has been reported for grafting-to methods while for grafting-from methods densities can range from 15-100 mg/m2 with thicknesses up to 100 nm [78-80]. The density and thickness plays an important role in the structure of the brushes created, as described earlier in this chapter. Lemieux et al [13] explains how at low concentrations, or density, the chains will form a pancake like structure or mushroom like structure, depending on the solvent (see Figure I. 3), where the chains don’t extend away from the surface. As the density of the polymer brush is increased the chains will alleviate stress and overlapping between neighboring chains to cause them to stretch in a way that causes a brush-like structure.

However the grafting-from method is not without disadvantages as well. While the density is a large advantage, and even with the introduction of polymerization techniques such as living radical, RAFT, or ATRP to improve upon controllable synthesis, this method fails to have the control that the grafting-to approach possesses.

Furthermore, the synthesis is complicated with several steps and needs a specific environment; or rather it cannot be done in ambient conditions. The uncontrollable molecular weight and polydispersity, where the grafting -to approach has an advantage, has been noted in the literature [81,82].

While these two methods of chemisorption are beneficial in the creation of polymer brushes on different substrates, each has their own advantages and disadvantages and consequently their own uses. However, in order for these polymer approaches to work, some type of chemistry, sometimes complex chemistry, is needed whether it is on the surface of the material or the polymer itself. Chemisorption needs the chemical modification where physisorption does not. However, chemisorption, due to the nature of the bonding, is stronger than physisorption. In physisorption, the chains are not end tethered to the surface but rather the chain itself. That is not to say that physisorption cannot be utilized to form a polymer brush and that it is the only way to accomplish a side chain grafted polymer. Furthermore in the following section, physisorption will be detailed with an attempt to publicize and spread an interest in new research recently found and investigated within the literature that may show a stronger adhesion to the surface than previously expected.

1.2 Non-Covalent Bonding of Polymer Brushes and Mats

Physisorption is described as a reversible process that is achieved by the self assembly of polymer surfactants or end-functional polymers on a solid surface [83]. The process is reversible due to the weak bonding strength of the adsorption that takes place. Van der Waals forces, deriving from the dipoles of molecules, are often used as an example of non-covalent bonding. The best known non-covalent attachment is adsorption which occurs because even if a small segment has a preferential adsorption then the chain will show large adsorption [87-89]. Selective solvents, selective surfaces, and polymer chains all have an impact on the adhesion properties [11]. Polymers have been known to adsorb through the interaction of a polymer melt and/or semi dilute solution.

1.2.1 Formation of Polymer Mats by Non-Covalent Bonding from Solution

Polymers such as PMMA, PVP, and PS have been shown to adsorb to multiple surfaces when the surface of the material is brought into contact with solutions of these polymers in different solvents. Esumi et al adsorbed PVP to silica particles from a sodium chloride (NaCl) solution at a constant temperature of 25˚C [84], while also adsorbing PVP to alumina particles in an aqueous solution with the use of surfactants

[85]. Peyser and Ullman adsorbed ionized PVP on glass particles from solutions in chloroform, methanol, and water, however, the adsorption was greatest for chloroform solutions for which they explained that adsorption will be greatest with the poorest

21 solvent [86]. They go on to explain that as the ionization increases the adsorption of

PVP decreases and as the ionic strength increases the adsorption increases.

Frantz and Granick were able to adsorb protio PS (PS-h) to silicon wafers by immersing them into solutions of cyclohexane at the θ temperature [87,88].

Subsequently, they observe desorption kinetics of the PS-h by adsorbing deuterio PS

(PS-d) on silicon wafers with already adsorbed PS-h. After 6-10 hours it was seen that the PS-h started to desorb and the PS-d started to adsorb where the sites had once been occupied by the PS-h showing stronger adsorption properties for the PS-d and silicon.

Frantz and Granick also state that hydroxyl groups, often found at the surface of a silicon oxide substrate, or even grown using procedures such as UVO, or piranha solution, will have a strong influence on the adsorption properties.

Johnson and Granick show similar results through the use of PMMA-h and

PMMA-d in solutions of carbon tetrachloride (CCl4) at the θ temperature; however, in this case the PMMA-d is displaced by the PMMA-h [89]. Other studies have shown that

PMMA and PS systems will exhibit similar tendencies where PMMA will desorb an already adsorbed PS layer from a silicon oxide surface [90] and in fact have shown that in a 50:50 mixture PMMA will adsorb and PS will be excluded [91].

1.2.2 Guiselin Brush

The studies previously mentioned often look at the kinetics of the desorption/adsorption process. Furthermore, these studies have also been in solvents

22 at the theta temperature to mimic a polymer melt. Guiselin [92] points out that when a polymer melt or semi-dilute solution is brought into contact with an attractive surface that the polymer chains will instantaneously and irreversibly adsorb to the surface.

Guiselin goes on to describe this interfacial layer, known as a Guiselin brush, as a loop and tail structure on the surface, much like a Gaussian model, where the loops are adsorbed at the surface. Guiselin explains that even after washing in solvent the initially adsorbed layer is still attached. Other theoretical analysis has been performed to not only observe the Guiselin brush but to analyze chain conformation of melt adsorbed surface layers and solvent swelling effects on those layers [93].

In et al [94] accomplished a similar model in a slightly different way, falling into category (3) of Table I. 1. By modifying PS-r-PMMA with added random traces of HEMA into the chain, hydroxyl groups from the HEMA could chemically attach to a silicon surface to formulate a neutral surface that was bound along the chain much like an adsorbed layer.

Raviv et al [95] not only used polyethylene oxide (PEO) on a mica surface but the adsorption process is done in a solution of toluene. In addition, the polymer chains are confined between two layers to form a modified version of the Guiselin brush by limiting the relaxation of polymer chains.

1.2.3 Polymer Brush Through Non-Covalent Bonding

Even though polymer chain adsorption is not considered a polymer brush since the chains don’t extend away from the surface, there is still a way of generating one by utilizing block copolymers (Category (2) of Table I. 1). With block copolymer systems it is possible to form polymer brushes through the use of one block preferentially adsorbing to the surface while the other protrudes away from it. An illustration of this can be seen in Figure I. 9.

The non-covalent polymer brush structure is highly impacted by the nature of the copolymers, architecture of the copolymers, length of each block and the interactions between the blocks and surface [11]. For the system a selective solvent must be chosen such that one block adsorbs to the substrate surface forming a ‘stuck’ layer for which the other block is allowed to from a brush away from the surface.

Various copolymers with a PS block have been used within the literature as a means to produce polymer brushes through non-covalent adsorption. Studies by Kelly et al [96] and Parsonage et al [97] found PVP to adsorb to a mica surface while PS formed a brush when used in a toluene solution of PS-b-PVP.

Fytas et al also used toluene with the copolymer PS-b-PEO to form polymer brushes on a glass surface [98]. Similar to the PS-b-PVP system, the PS formed the polymer brush since the PEO showed a stronger attraction to the surface forming the adsorbed layer. Motschmann et al used PS-b-PEO from a toluene solution as well but were able to adsorb the polymer brush onto a silicon wafer [99].

Figure I. 9: Schematic of polymer brush of block copolymer by physisorption

A useful advantage of polymer brushes through non-covalent bound layers is that it is not a difficult procedure to execute, especially compared to the chemistry, control, and steps needed for covalently bound layers. However, as mentioned earlier in this chapter, non-covalent bond strength is inferior to that of covalent bonds and is consequently susceptible to desorption due to good solvents, temperature, or dewetting due to the weak interactions between the chains and substrate surface. The next section will further examine non-covalent bound layers from the standpoint of polymer brush fabrication and not from a solvent solution. Taking a look into the non- covalent nature in more detail introduces stronger adhesion than once believed.

1.2.4 Polymer Adsorption Through Thermal Annealing

A new observation within the literature is the change of the glass transition temperature, Tg, of nano thin films [100-108]. While at first it may appear irrelevant to a polymer brush or layer on a surface there has been reports of interfacial layers when collecting data for Tg changes with thickness of the film. What’s more interesting is that the change in Tg has been reported different with decreasing thickness from different authors. Reports have been found in literature that as the thickness of the film gets as low as 10 nm that the Tg will increase by as much as 50˚C [104,105] or decrease by as much as 20˚C [106,107].

Fujii et al [108] tries to explain this phenomenon by examining what is on the surface of silicon when PS is cast on it. In trying to determine whether or not oxidized silicon vs hydrogenated silicon may cause a difference in thin film Tg through thermal annealing Fujii observed that a strongly adsorbed layer will stick to the surface of the silicon, in either case, even after rinsing with solvent. However, a thicker interfacial layer was observed with the hydrogen passivated silicon. Fujii further noticed a direct correlation with the thickness of the residual layer when the molecular weight was increased, however, the initial thickness of the polymer had little to no effect on the residual thickness that was left after thermal annealing and subsequent solvent rinsing in toluene.

Napolitano and Wubbenhorst [109] also try to explain the phenomena of thin films with change of Tg. After annealing the films at different times for different

26 molecular weights and film thicknesses, they observed that increased annealing time and molecular weight showed an increase in the Tg. However, there PS films spun cast on aluminum also showed a residual layer even after rinsed with toluene which they then compared to the Tg of the films the residual layers had been adsorbed with.

Furthermore, they observed that the residual PS layer thickness would increase as the annealing time increased before leveling out after a certain length of time.

CHAPTER II

EXPERIMENTAL SECTION

In this section the materials and experimental procedure will be detailed and elaborated. Furthermore, the characterization methods and equipment will be described as well.

2.1 Materials

Materials used to obtain experimental data were either synthesized in the lab with existing materials or purchased from outside suppliers. The following will describe the process or acknowledge the supplier.

2.1.1 Anionically Polymerized Polystyrene

Anionically polymerized polystyrene (PS) was purchased through Alfa Aesar. The

PS standards were purchased at four different molecular weights all having a polydispersity of 1.06. The four molecular weights included 1.3k, 50k, 123k, and 200k.

2.1.2 RAFT Polymerized Polystyrene

RAFT polymerization is employed to generate PS samples for use in fabricating thiol terminated PS. RAFT involves reversible chain transfer. Therefore, the chain transfer agent is an important aspect of the process. The distribution and molecular weight of these chains are dictated by the RAFT agent and initiator combination.

Whereas other methods need control of the number of initiators, the number of chains is determined by combination of the RAFT agent and initiators. This is an effective way to control the molecular weight and distribution of the polymer.

2.1.2.1 RAFT Agent

The RAFT agent used in the polymerization of RAFT polymerized PS is Dibenzyl

Trithiocarbonate (DBTC). DBTC synthesis is described elsewhere [110] and illustrated in

Figure II. 1.

S S n

Figure II. 1: RAFT agent DBTC used for RAFT polymerization

2.1.2.2 Styrene Monomer/n-Butylamine

Styrene Monomer was purchased through Alfa Aesar and purified through a column. The n-Butylamine was purchased through TCI America and used for the aminolysis procedure of the RAFT polymerized PS.

2.1.3 PS-b-PDMS

Block copolymer PS-b-PDMS was synthesized by Maurice Wadley at the

University of Akron at different volume ratios keeping the PDMS block Mw at 10k. The two volume ratios used are 3:7 and 7:3.

2.1.4 PtBS

Poly (tert-buytlstyrene) was synthesized by Yuqing Liu by RAFT polymerization to achieve a polymer with Mw of 21k and PDI of 1.5.

2.1.5 Silicon Substrate

Silicon (100) wafers purchased through Silicon Quest Int’l are UVO exposed for

20 minutes. UVO cleans the surface of organic material as well as induces hydroxyl functionality at the surface illustrated in Figure II. 2.

Figure II. 2: Cartoon of silicon wafer

2.2 Experimental Process

This section can be split up into three different, but important, steps that will detail the process used to obtain materials for data collection. The first section will go over how to synthesize the polystyrene (PS) polymer via RAFT polymerization. The second section will focus on the aminolysis of the RAFT polymerized PS used to try and create a polymer brush. The third section will elaborate on the process used to prepare samples for data collection.

2.2.1 RAFT Polymerization of Polystyrene

RAFT polymerization of PS is done by adding the RAFT agent to an amount of styrene monomer. The following equation can be used to calculate the amount of RAFT agent needed based on a target Mn.

The above equation describes how to get the weight, in grams, of the RAFT agent based on the Mw of the RAFT agent, the desired target Mn and the weight, in

31 grams, of the monomer used which is simply the volume, in ml, multiplied by the density of the monomer (.909g/ml for styrene).

After measuring the desired amount of RAFT agent and monomer, it is placed in a 10ml or 15ml vial with stir bar and sealed. The system is then flushed with nitrogen gas for 15-20 minutes and polymerized at 130˚C while stirring for 6 hrs. The material is then precipitated in methanol solvent, filtered, and vacuum dried overnight to eliminate any presence of solvent. Figure II. 3 is an illustration of the RAFT polymerized PS.

S S n n Figure II. 3: RAFT polymerized PS using DBTC as RAFT agent

2.2.2 Aminolysis of RAFT Polymerized Polystyrene

From RAFT polymerized PS, aminolysis is performed in order to generate thiol terminated. The RAFT polymerized PS is dissolved in toluene (Tol) at a 1:4 ratio. A stir bar is added to the liquid in a 10 or 15 ml round flask and nitrogen flushed for 15-20 minutes. A butylamine bottle is nitrogen flushed as well as the needle used to inject the butylamine into the system. The amount of butylamine needed is based on a 1:2 ratio with PS. Once the butylamine is added to the nitrogen

32 flushed system it is to be stirred for 30 minutes. The material is precipitated in methanol, filtered, and vacuum dried overnight. Figure II. 4 is an illustration of the thiol terminated PS.

SH n Figure II. 4: Thiol terminated PS

2.2.3 Preparation of samples

The following section will detail how to prepare the samples for data collection including definitions used in this thesis.

2.2.3.1 Sample Preparation

Approximately 1-2% PS solutions in toluene are spun cast on a silicon wafer

UVO treated for 20 minutes. Spin coating parameters are 2000 rpm for 30 seconds.

Films that are annealed are done so under vacuum conditions at 170˚C overnight

(16-24 hrs). After cooled the films are rinsed in solvent to wash away the film. The

33 substrates are then analyzed using contact angle, x-ray photon spectroscopy, x-ray reflectivity, and optical microscopy. These will be detailed in later sections.

2.2.3.2 Levels of Rinsing

Six levels of rinsing are used within this work including 15 minute solvent rinse, 1 day solvent rinse, 7 day solvent rinse, sonication, soxhlet extraction, and reflux. The 15 minute, 1 day, and 7 day solvent rinse both have the samples submerged in solvent for the labeled times. Sonication places samples in a sealed

20ml vial filled with solvent and sonicated in water for 1 hr. Soxhlet extraction and reflux both consist of samples submerged in heated solvent.

2.3 Equipment/Measurement Methods

Multiple methods were used in order to analyze and characterize the substrate properties and were mentioned earlier. This section will go over what types of methods were used and, when applicable, will go over the steps in using them.

2.3.1 Gel Permeation Chromatography (GPC)

Ways of collecting information about the molecular weight information as well as the molecular weight distribution include size exclusion chromatography (SEC), osmotic pressure, light scattering, sedimentation equilibrium, and intrinsic viscosity measurements [21]. GPC is a type of size exclusion chromatography (SEC) that

34 separates molecules based on their size, or in this case the radius of gyration. The separation occurs with the use of an appropriate column packed with porous particles that molecules of different size can penetrate. Since the largest molecules will have the hardest time to fill in these porous spots they will be excluded from the column first whereas the smallest molecules will have the ability to fully permeate the pores taking more time to escape the column. Given the time it takes these molecules to leave the column, the GPC can analyze how big the molecules are from the largest to smallest, which can then in turn give information about the distribution of the molecules in the solution. Hiemenz and Lodge point out that the separation is based on size rather than the molecular weight, which can be described as the hydrodynamic volume, Vh [23].

GPC measurements were carried out by dissolving approximately 1-2 mg of material in THF solvent and injected through a filter into the GPC machine at 0.5ml/min.

The equilibration is done with PS standard. After a 30 minute run time the data is collected and analyzed.

2.3.2 Contact Angle (CA)

Contact Angle measurements were carried out at the Polymer Science facility at

The University of Akron. The Rame-Hart contact angle utilizes a light and camera connected to a computer with software to measure contact angle. Water is chosen as the primary liquid of choice. Drops are placed on the substrate at constant volumes.

After the drop is placed on the surface, a picture can be taken and the contact angle can

35 be measured using the installed Drop Image Advanced software. For this work, at least three CA measurements were taken when the sample size permitted three spots.

2.3.3 Optical Microscopy (OM)

Optical microscopy is used to analyze the topological characteristics of the films.

Two magnifications of 500x and 100x are utilized in looking at the surface by using reflection at the surface for analysis.

2.3.4 X-ray Photon Spectroscopy (XPS) & X-ray Reflection (XR)

Both XPS and XR were carried out in the facilities at NIST by Christopher M.

Stafford.

CHAPTER III

SURFACE MODIFICATION OF SILICON OXIDE VIA THIOL TERMINATED PS

(COVALENT)

3.1 Introduction

Initial studies of thermally annealed films of thiol terminated PS are investigated.

The aim is for the thiol group at the end of the PS chain to react with a silicon oxide surface, with hydroxyl functionality, to form a sulfur-silicon bond leaving an end tethered PS brush on silicon. Studies have been done in the literature to show that PS-

SH can easily adsorb to a gold surface [34,35] as well as form a polymer brush [35].

Taylor and Jones showed that when polyethylene oxide (PEO) had a thiol terminated group they were unable to remove the layer from the gold surface [61].

When the PEO had a hydroxyl terminated group however, they were able to remove the layer through sonication with chloroform. Koutsos et al used a similar method by grafting thiol terminated polystyrene (PS) onto gold from a toluene solution [111]. They go on to show that as the molecular weight is increased that the coverage of the grafted brush starts to decrease for constant exposure times.

Stuen et al have shown that it is possible to covalently bond hydroxyl terminated

PS (PS-OH) to silicon oxide surfaces [112]. By spin casting PS-OH and annealing for 48 hours, the films were chemisorbed at the substrate surface.

It has been shown that sulfur has an affinity for gold surfaces while PS-OH can bind to a silicon surface. So the question arises whether or not it is possible to observe similar results on a silicon oxide surface. In other words, can a thiol terminated PS be tethered to a silicon surface via the sulfur atom.

Coulter et al performed tests to study the tethering behavior of sulfur atoms to a pure silicon surface (no oxide layer) through contact with benzenethiol (C6H5SH) [113].

In their work, they utilize FTIR and XPS to show that while there can be some adsorption though the aromatic ring, most of the layer is bonded to the silicon surface through the sulfur atom. By cleaving the thiol group the sulfur can then bond to silicon. Since their work shows the tethering of benzene through the use of a thiol functional end group, it is fair to assume that it is possible to graft thiol terminated PS to a silicon oxide wafer by the formation of Si-S-R bonds.

3.2 Materials and Water Contact Angle

To study this hypothesis, reversible addition fragmentation transfer (RAFT) polymerization of styrene monomer with the RAFT agent dibenzyltrithiocarbonate was performed. This polymer was then treated with a primary amine to form thiol terminated PS (PS-SH).

After annealing spun cast films of PS-SH, they are subsequently rinsed in solvent and water contact angle measurements are taken and analyzed. A GPC plot for the three materials used is given in Figure III. 1 and summarized in Table III. 1. Initial results with 7k molecular weight PS-SH can be seen from Figure III. 2.

By examining the images in Figure III. 2, a couple things of interest can be noted.

First, the annealing process is a vital step in getting the PS-SH film to leave a residual layer, even after rinsing. When the as cast film was rinsed the contact angle decreases, no longer having characteristics of a PS film. The annealing step raises the temperature above the Tg of the polymer, allowing mobile chains to orient in a way that the thiol group reacts with the hydroxyl functional surface to form sulfur to silicon bonds leaving water as a byproduct. The second curious aspect is that when the film is not annealed and rinsed, it does not show contact angles similar to values comparable with a bare silicon wafer surface. This can be explained by the results of Mirji et al who observed that toluene will adsorb to the surface of silicon and give water contact angles in the range of 52-69˚, depending on how long they allowed the toluene to adsorb *114].

Furthermore, they attempted to lift the solvent off by thermally heating the silicon samples, however, from around 200-300˚C the contact angle only went down to about

45˚ and only approached 20˚ once the temperature reached 500˚C. It can be concluded that the toluene adsorption is fairly strong and gives contact angle measurements higher than a clean silicon wafer. For future reference, the PS water contact angle is taken as 85˚-90˚.

Figure III. 1: GPC trendlines for three different Mw of RAFT polymerized PS (PS_RAFT) and thiol terminated PS (PS-SH)

Table III. 1: RAFT polymerized (PS_RAFT) and thiol terminated PS (PS-SH) material data

Figure III. 2: Water contact angle images. Solvent used is toluene. Silicon wafer has no surface changes or modifications.

An interest in how the molecular weight affects the water contact angle of the surface is investigated. Three molecular weight PS-SH films were annealed and rinsed in toluene overnight. Their contact angles were measured and are plotted in Figure III. 3.

It can be seen from the plot that all three data points are within the range of 87˚ to 89˚ which is comparable to a PS film. While it may seem that there is some variation, the changes are so small that they can be assumed negligible for this range of molecular weight. A broader range will be investigated in a later section.

Figure III. 3: Effect of molecular weight on water contact angle of annealed films rinsed in toluene overnight. Standard error bars based on 6 measurements.

3.3 Effect of Different Levels of Rinsing

One mode of rinsing has been investigated using toluene solvent. To further explore the residual layer binding strength, three additional levels are performed.

Water contact angle is measured and analyzed for an insight to the bound PS-SH layer after rinsing with four different levels. Different levels of rinsing include submerged for

15 minutes, submerged overnight, sonicated in solvent for 1 hour, and soxhlet extraction overnight.

A bar graph of the four levels and their affect on the resulting water contact angle for annealed films and as-cast films can be seen in Figure III. 4. Once again the

42 thermal annealing step is important in binding the PS-SH to the surface. However, the main purpose of this graph shows that the four levels of rinsing have very little effect on what the resulting water contact angle is. In the case of the annealed films, the rinsed films contact angle all range from 86˚ to 89˚. In the case of the as-cast film, longer period of exposure to the surface causes the water contact angle to rise. However, the contact angle never rises above 65˚. These angles are characteristic of the toluene adsorption that takes place when exposed to the silicon surface.

3.4 Effect of Different Solvents

After rinsing at different levels, the overnight method was chosen for future experiments for simplicity. However, since other solvents are known to dissolve PS, three other solvents were chosen to observe their affects on the resulting water contact angle. Results, in Figure III. 5, show that none of the four solvents affect the final water contact angle result. The range of contact angles is once again from 86˚ to 89˚ still confirming a residual layer of PS bound at the surface. While the contact angle does change slightly from solvent to solvent, it is not nearly enough to conclude that the layer has been pulled off the surface completely, or at all. In fact, the difference in the contact angle can arise from the solvents themselves and how the polymer chains behave when exposed to them causing different chain morphologies at the surface.

Even though different solvents do not succeed in pulling off the polymer film, it does show that the solvents may have some affect on the morphology of the surface of the chains and consequently the resulting surface energy.

Figure III. 5: Effect of solvents on water contact angle using toluene (Tol.), tetrahydrofuran (THF), acetone, and chloroform.

It was hypothesized that there may be a possibility that the spun cast solution affect the resulting contact angle results. Therefore, the bar plot in Figure III. 6 investigates the difference of using THF versus toluene as the spin casting solution. The bar plot shows a film that is cast in THF, thermally annealed, and subsequently rinsed overnight in THF. Similar results can be seen for THF results in Figure III. 5. The THF shows a slightly lower contact angle most likely due to the solvent interactions with the chains as compared to the interactions that toluene may have. It can be conclusively stated then that the solvent chosen to spin cast from has little affect on the resulting contact angle of the annealed films rinsed in solvent. Once the annealing process takes

45 place the type of solvent used in spin casting and the affect on the chains movement along the substrate will have been nullified.

Figure III. 6: Effect of solvent chosen to spin cast films from and subsequently rinsed with

3.5 Topological Changes (OM)

Topological behavior of block copolymer films cast on different surfaces are investigated. With a residual layer of PS-SH at the surface, an investigation into block copolymer morphology changing due to this layer is performed. Block copolymers have the ability to orient themselves in parallel or perpendicular fashion based on the volume fraction and surface energy of the blocks as well as the thickness of the film. Parallel and perpendicular orientation of cylindrical block copolymers is displayed in Figure III. 7.

Figure III. 7: Cylindrical forming block copolymer oriented (a) perpendicular and (b) parallel to the substrate

The theory is that when casting a block copolymer film on a bare silicon wafer, the resulting morphology will differ from the same block copolymer cast at the same thickness but over the residual PS-SH layer that exists at the substrate of the silicon. In order to test this theory, a lamellar forming poly (styrene-b-methyl methacrylate) (PS-b-

PMMA) with a molecular weight ratio of 30k-b-30k is spun cast onto a silicon wafer with the PS-SH residual layer at the substrate surface as well as a bare silicon oxide wafer.

Optical microscopy at a magnification of 500x was used to examine the surface of the films after going through thermal annealing to investigate. Comparisons can be located in Figure III. 8. The image shows two PS-b-PMMA films of equal thickness annealed at 170˚C overnight. Since the polymer film thicknesses are almost identical it is expected that the block copolymer morphology at the surface will not change, however it is clear that when there is a presence of the PS-SH residual layer under the film that the film surface goes from a spotty looking surface to a smooth surface. Since

PMMA is known to preferentially wet a silicon oxide surface, it is believed that the PS

47 phase will be attracted to the surface of the wafer with the PS-SH layer which affects the resulting morphology. The domain spacing of the block copolymer is affected by this layer eliminating the island (darker/brown region) and hole (lighter/white region) type morphology at the surface in Figure III. 8 (a). This can be explained by Smith et al who perform a gradient thickness study of lamellar forming PS-b-PMMA on silicon oxide

[116]. They notice that at different thicknesses, as well as different block Mw, different arrangements of islands and holes, and what they term a spinodal pattern, form at the film surface due to domain spacing versus thickness of the block copolymers. When PS- b-PMMA film thickness is on the order of (m + ½)L0, where m is an integer and L0 is the domain spacing of the copolymer, the film surface will be smooth with PMMA at the substrate and PS at the surface. When the thickness is in between (m + ½)L0, then different islands, holes, and spinodal formations are observed. Furthermore, when their data for domain spacing is plotted against the block copolymer molecular weight the

-.6229 power law equation L0 = 2.4269·Mw is obtained. Using this equation, it is estimated that for a 30k-b-30k lamellar block of PS-b-PMMA the domain spacing is 20.2nm. This value combined with the information obtained from Smith et al gives valuable insight into what is occurring in our system.

On a silicon surface PMMA preferentially wets the substrate surface while PS wants to go to the air interface. This is the system described by Smith et al. However with a PS brush on the substrate, the PS block will preferentially wet the substrate surface and still want to wet the air interface. In order for the PS at the air interface to

48 remain smooth, the thickness must be an integer of the domain spacing (t = m·L0).

These two systems (for smooth surface topology) are illustrated in Figure III. 9.

When dividing the thickness of the copolymer film on silicon oxide (47nm) by the calculated domain spacing (20.2nm), a value of 2.3L0 is obtained. Since this value is less than a 2.5L0 value needed to obtain a smooth surface morphology, holes will be observed which can be illustrated in Figure III. 8 (a). When dividing the thickness of the copolymer film on a PS brush it is important to also take into account the thickness of the brush. To take into account the thickness, it is taken as approximately 3nm and subtracted from the total thickness giving a value of 43nm. This thickness divided by

20.2nm gives a value of 2.1L0. This value is slightly above the 2L0 value needed for a smooth surface topology and may show signs of islands forming. From Figure III. 8 (b), the topology is clearly smooth with some small dark spots where islands are forming, proof that there is a PS layer at the surface of the silicon substrate.

Figure III. 8: Optical Microscope images at 500x for PS-b-PMMA (30k:30k) films thermally annealed (170˚C) overnight on (a) bare wafer, t = 47nm and (b) 2k PS-SH residual layer, t = 46 nm

Figure III. 9: Domain spacing for PS-b-PMMA films (PS-black, PMMA-white) with smooth topology on (a) silicon oxide and (b) PS brush modified silicon substrate

3.6 Dewetting (OM)

PS has been known to dewet on silicon especially for thin thicknesses. However, it has been observed that the PS-SH films show very little dewetting. Studies were done to show that the thiol group has an effect on this behavior. A comparison of thermally annealed PS-SH films and RAFT polymerized PS is shown in Figure III. 10. The PS_RAFT

50 image (top) had an as-cast thickness almost double that of the of the PS-SH as-cast film

(bottom) but when annealed the PS-SH film showed almost no dewetting while the PS-

RAFT film dewet completely showing PS droplets at the surface. Based on these images, the thiol group of the PS-SH has a large impact on the binding behavior of the PS chain to the surface which reduces the effects of dewetting.

Figure III. 10: Optical Microscope images at 100x for (top) 5.3k PS_RAFT, t = 77nm and (bottom) 3.3k PS-SH, t = 49 nm after (a) as-cast and (b) thermally annealed at 170˚C for 24 hours

After confirming that there is a PS-SH layer at the surface of the silicon substrate, it is important to also confirm that the thiol group is the primary and sole cause of the

PS binding to the surface. By running the same experiments with the RAFT polymerized

PS and also anionically polymerized PS, a control experiment will confirm that PS without the thiol end group can be rinsed away from the silicon surface. Results and comparisons of these two control experiments are shown in Figure III. 11 and Figure III.

12, where an unexpected result can be observed. Regardless of the type of PS chain that is annealed to the surface, a residual layer still remains at the substrate after rinsing with toluene.

Figure III. 11: Water contact angle of PS-SH () and PS_RAFT (,) after rinsing with toluene. After overnight in toluene PS_RAFT sonicated in toluene for 1 hour

Figure III. 12: Water contact angle of PS-SH () and anionically polymerized PS () after rinsing with toluene overnight.

3.7 Control Experiments (PS-SH vs PS_RAFT vs PS)

Further analysis was performed on the anionically polymerized PS, which will be labeled in this chapter from here on as ‘pure PS’ or ‘PS’. Additional levels of rinsing were looked at as well as different solvents, and a broader range in molecular weight.

Two molecular weights (9k, 123k; PDI: 1.06) are chosen for pure PS and are compared to values of PS-SH at a molecular weight of 7k. Quantitative results are shown in Table III.

2 and Table III. 3. Similar to previous data, the solvent choice has very little effect on the resulting contact angle of the residual layer and does not pull off the PS film completely. Cyclohexane and DMF were chosen as additional solvents to try, however, neither lead to a difference in the water contact angle. Even when the molecular weight is increased from 9k to 123k there is almost no variation at all between the measured contact angles.

Reflux experiments were performed with toluene and THF to see if heated solvent has a greater effect in pulling off the film completely from the substrate.

Furthermore, Hinkley et al did a study of adsorption of PS onto silicon wafers from solvent and noticed that when diethyl malonate was used at the theta temperature he observed no PS adsorption to the silicon surface [115]. Therefore, along with the reflux experiments, diethyl malonate was heated to the theta temperature in an effort to pull off the film by desorbing the PS layer and adsorbing the diethyl malonate instead. The results of the reflux data can be seen in Table III. 4. Still the residual layer of PS is left at the surface of the substrate for all three solvents signifying that the PS chain is adsorbing and binding strongly to the silicon.

Cyclohexane DMF

PS (9k) PS-SH (7k) PS (9k) PS_SH (7k)

88.7˚ 88.5˚ 88.4˚ 85.8˚

Table III. 2: Water contact angle measurements of PS versus PS-SH annealed and rinsed in cyclohexane and dimethylformamide (DMF)

7k PS-SH 9k PS (pure) 123k PS (pure) Bare Wafer

Non- Toluene 88˚ 88˚ 89˚ 64˚ Polar Chloroform 90˚ 88˚ 87˚ 35˚

Acetone 91˚ 90˚ 88˚ 54˚ Polar THF 87˚ 89˚ 88˚ 31˚

Table III. 3: Water contact angle measurement of PS (9k vs 123k) versus PS-SH annealed and rinsed in different solvents. Bare wafer values are water contact angle due to any solvent adsorption.

Reflux Data for 9k pure PS

Toluene @ 70˚C 89.3˚

THF @ 100˚C 86.3˚ DM @ 37˚C 87.5˚ Table III. 4: Water contact angle measurement of 9k pure PS refluxed in different solvents at different temperatures. 37˚C is the theta temp (Tθ) for diethyl malonate (DM).

3.8 Summary

Initial theory and hypothesis stated that a thiol terminated PS-SH chain would react with the surface of a silicon oxide substrate. Water contact angle results and block copolymer orientation studies prove the presence of a residual PS-SH layer. Solvent type, solvent rinsing level, and Mw all have little effect on the resulting water contact angle of the PS-SH film. Dewetting behavior shows strong evidence of an end tethered

PS-SH chain through silicon to sulfur bonds formed after cleavage of the thiol group.

However, after control experiments it appears that PS adsorbs to the surface of silicon oxide through thermal annealing. This suggests that there is binding of the PS-SH chain through covalent attachment (thiol group) and non-covalent attachment (adsorption along the chain).

The adsorption of the PS chains is an unexpected result. There has been little work done in the literature concerning PS bound layers through thermal annealing.

Furthermore, the phenomenon appears to be only briefly mentioned along with little

56 concern. However, the residual layer of PS not only affects data for experiments that may involve re-use of silicon surfaces after solvent cleansing but offers an alternative to modifying the surface energy of a substrate. The next chapter will investigate this adsorption process in more detail.

CHAPTER IV

SURFACE MODIFICATION OF SILICONO OXIDE VIA THERMALLY ANNEALED PS

(NON-COVALENT)

4.1 Introduction

In the previous section, it was seen that no matter what PS type chain was used that when annealed and subsequently washed in solvent that a PS residual layer was still present at the surface of the silicon wafer samples. In an effort to examine the strength and behavior of these adsorbed layers similar studies will be performed as in the previous chapter.

Due to small molecules preferential adsorption and the knowledge that toluene adsorbs to the surface of silicon oxide, it is believed that a PS chain will show preferential adsorption as well. While the structure of toluene is somewhat similar to a

PS monomer, it can be reasoned that the monomer will also adsorb. With a polymerized chain, each individual segment will have preferential adsorption adding up to a strongly adsorbed layer of PS. Through thermal annealing the chain mobility will increase allowing movement for the chains to orient in such a way for the maximum

58 amount of contact with the surface resulting in an attached layer with more stability than expected.

Contact angle measurements were taken before and after annealing as well as before film removal with selected solvents. As the surface energy increases the water contact angle decreases, giving a rough estimate of how the surface energy changes as the contact angle changes. Once again it is important to note that the water contact angle of bare silicon can range from less than 10˚ to about 20˚ based on UVO exposure and cleanliness of silicon substrate; furthermore, the water contact angle of a PS film can range from 85˚ to 90˚ regardless of as cast or annealed films. These values will be compared to the following contact angle results.

4.2 Effect of Different Solvents and Different Levels of Rinsing

It was originally observed that when annealed films of PS were rinsed with toluene that something foreign is still present on the silicon oxide surface. To further observe this, three other solvents were selected that dissolve PS strongly and used for multiple molecular weight samples. Figure IV. 1 shows the results of this while Table IV.

1 and Table IV. 2 are the raw data for Figure IV. 1. The four solvents selected were

Toluene (Tol), Tetrahydrofuran (THF), Cyclohexane (Cyclo), and Chloroform (CHCl3).

Results from Figure IV. 1 prove that there is a clear distinction between the annealed samples and the as-cast samples after rinsing with the chosen solvents. After annealing

PS films overnight at 170˚C any attempt to wash away all presence of the PS film is not

59 possible for these four solvents, which have been known to be good solvents for PS.

While only small variations exist for an increase in Mw, the trend is curious since it seems to first decrease at 50k Mw and then rise after 50k. To explain this, it is suggested that for smaller molecular weights there is a larger impact from the polymer chain ends from the anionic polymerization. At smaller molecular weights the functional ends will show a higher percentage within the layer, where ends can stick up towards the surface, causing a slightly higher contact angle. Inversely, when the molecular weight increases the roughness of the layer is expected to also increase. This increase in roughness will cause the water contact angle to rise. Therefore, even though the smallest molecular weight can be assumed to have the smoothest surface the effect of the chain ends will be more prevalent and cause a larger change in water contact angle than at higher molecular weights. As in the case for toluene, other solvents may show adsorption tendencies which explains the difference in contact angle from an unmodified bare silicon oxide wafer (≈10˚-20˚) to values of as-cast wafers washed in solvent.

As-cast water contact angle

Solvent Mw (kg/mol) 1.3 50 200 THF 51.6˚ 54.3˚ 51.7˚ Toluene 49.5˚ 53.0˚ 51.3˚ Cyclohexane 56.5˚ 57.3˚ 47.4˚ CHCl3 49.4˚ 54.9˚ 65.0˚

Table IV. 1: Water contact angle of as-cast films rinsed in selected solvents at three different Mw

Annealed water contact angle

Solvent Mw (kg/mol) 1.3 50 100 200 THF 88.7˚ 85.1˚ 86.8˚ 90.4˚ Toluene 85.5˚ 82.9˚ 86.6˚ 88.0˚ Cyclohexane 88.2˚ 84.2˚ 87.2˚ 89.7˚ CHCl3 84.3˚ 80.8˚ 88.3˚ 97.1˚ Table IV. 2: Water contact angle of annealed films rinsed in selected solvents at four different

Figure IV. 1: Water contact angle vs Mw for annealed (solid) and as-cast (open) films after rinsing with selected solvents overnight. Solvent: Toluene (), THF (), Cyclohexane (),

CHCl3 (). Error bars are standard error using six measurements.

Not only were different solvents investigated but also different types, or levels, of rinsing were used to observe the strength of the adsorption. Water contact angle measurements for 123k PS annealed films rinsed in toluene at the four different levels are shown in Table IV. 3 and Figure IV. 2

Four different levels were chosen with the solvent toluene. The types of rinsing are described under section 2.2.3.2 Levels of Rinsing. What is important to note from this data is that there is no level of rinsing that will completely wash away the PS film completely. In all cases, water contact angle measurements show a presence of a residual layer left behind. Due to this observation, future experiments are performed using the 1 day solvent rinse method for simplicity.

Contact Angle (degrees) Annealed Films

Reflux Mw Toluene Wash Toluene Bath Sonicated (1hr) (70˚C) Before Rinsing 123k 91.2˚ 91.7˚ 92.2˚ 92.3˚ After Rinsing 123k 89.1˚ 90.1˚ 88.8˚ 96.7˚ Table IV. 3: Water contact angle of 123k PS annealed films before (top) and after (bottom) rinsing with Toluene. Standard error is less than 0.5 for all cases.

Figure IV. 2: Water contact angle of PS annealed films rinsed with toluene at different levels

4.3 Adsorption of PS vs PtBS vs PS-b-PDMS

Adsorption studies of different PS chains were investigated on silicon oxide with the use of two different PS species. Block copolymer poly (styrene-b-dimethylsiloxane)

(PS-b-PDMS) as well as poly (tert butylstyrene) (PtBS) and the behavior of these materials are investigated using the conditions previously for thermal annealed films.

Figure IV. 3 is a plot of the water contact angles of these spun cast annealed films before and after rinsing with toluene. It can be seen that before attempting to wash away the films that the water contact angles are different between the three polymer films.

While the difference is not a large one, it is still present after the films are rinsed in toluene. A PS-PDMS film has the largest water contact angle while PtBS is the next highest and PS is the lowest. The water contact angle slightly decreases for each material after rinsing, but the change is consistent among all three polymers and the difference in contact angle is still observed with PS-PDMS having the largest contact angle and PS having the lowest. This confirms that the residual polymer layer present is truly that of the materials being washed and that the polymers are all adsorbing to the silicon surface. Although possibilities of adsorption through different groups, such as the PDMS block, it is believed that the PS chain is adsorbing to the surface allowing the

PDMS to sit closer to the air interface generating larger water contact angle measurements.

4.4 X-Ray Reflectivity (XR) and X-ray photon spectroscopy (XPS)

To measure the thickness of the PS adsorbed layers x-ray reflectivity was utilized.

Figure IV. 4 is a plot of the measured thicknesses for different molecular weights of annealed PS films rinsed in toluene. This plot illustrates a direct correlation between the range of Mw and the measured thickness of the PS layer, where as the molecular weight increases the thickness of the adsorbed layer also increases. While there are only three points in the plot, the increase in thickness looks to be linear for this range of molecular weights.

Figure IV. 4: Thickness vs Mw of anionically polymerized PS via XR

This is expected since with increasing molecular weight the chains will be longer as well as more entanglements will be present with chains only having portions of it sticking to the surface. During thermal annealing, the lower Mw will possess more mobility than the larger Mw to move about, allowing more of the chain length to adsorb with less chain sticking away from the surface. The length of the chains due to molecular weight will cause more robust layers.

By using XR to obtain thickness values for PS-SH residual layers from the previous chapter, a comparison between the pure PS and PS-SH is plotted in Figure IV. 5 and thickness variations are observed. The increase in thickness with Mw is faster with the

PS-SH polymer. This further supports the theory of a thiol end group having an effect at the silicon oxide surface. Since the sulfur is bonding to the silicon substrate, the tethered PS is allowed to extend away from the substrate causing more of a brush-like layer. Without the thiol group, the PS is left to bind along the chain reducing the ability to stretch away from the chain and therefore making the thickness of the layer lower.

Figure IV. 5: Thickness versus Mw of pure PS and PS-SH

XPS measurements are taken on PS residual layers to determine element percentage content at the surface. Figure IV. 6 represents the percentage amount of carbon, oxygen, and nitrogen detected by x-ray photon spectroscopy in the residual layer. The small amount of nitrogen detected is unexpected and can arise from the lab

67 environment or vacuum oven conditions. The oxygen content comes from the silicon oxide surface since the penetration depth of the XPS is approximately 7nm, which is equal or greater to the largest thickness layer we observed from XR. Carbon is from the polymer chain, or residual layer, itself. The XPS plot matches up well with the XR plot in that as the molecular weight increases there is more polymer on the layer which in turn creates a higher carbon content. Since there is more carbon content the percentage of carbon compared to nitrogen and oxygen increases. For the lowest molecular weight, oxygen even outweighs the percentage of carbon on the substrate; however, for the highest molecular weight, the percentage of carbon has increased so large that the oxygen percentage has fallen to a small trace.

Figure IV. 6: XPS data for Carbon, Oxygen, and Nitrogen percent in the PS layer at different Mw. Standard error bars are present for three measurements but cannot be seen due to the small values

4.5 Effect of Annealing Time and Annealing Temperature

Two variables for thermal annealing, annealing time and the annealing temperature, are varied and investigated with water contact angle measurements. The first plot, Figure IV. 7, represents how the annealing time effects the water contact angle of the rinsed film. As the annealing time is increased the water contact angle also rises. With longer periods of annealing time the polymer chains are able to move around and bind to the surface; however, at shorter time periods the chains don’t have enough time to attach to the maximum amount of locations. Consequently, when the film is rinsed in toluene there is less coverage of the PS at lower annealing times and the contact angle decreases. It is interesting to note that from a time period of 3 hours to

12 hours, the water contact angle shows little change. It is believed that entanglements in the chain resist movement and require longer periods of time to overcome this restriction.

The second plot, Figure IV. 8, represents the annealing temperature effects on PS adsorption using water contact angle measurements. Similar to the previous plot, when the annealing temperature increases the water contact angle increases. Unlike the previous plot however, this one shows a linear progression. However, any value under

100˚C would show water contact angles around 60˚ since under the Tg of PS there would be no movement of the chains and consequently the film would show behavior similar to an as-cast film. It is believed that for both variables there is a limit to how much time or how high the temperature is before the water contact angle starts to level off.

Based on these results it can be concluded that annealing at higher temperatures for longer annealing times will let the PS film have the best chance of binding to the surface at all sites possible, since there will be better chain mobility at longer periods of time allowing optimum chain movement.

Figure IV. 7: Water contact angle vs annealing time for 200k pure PS after rinsed in toluene overnight. Annealed at 170˚C.

Figure IV. 8: Water contact angle vs annealing temperature for 200k pure PS after rinsed in toluene overnight. Annealed for 18 hours.

4.6 UVO Surface Energy Modification

Ultraviolet light and ozone (UVO) have been used, not only as a means of ridding surfaces of organic species for cleaning purposes, but also as a means to control the surface properties such as the surface energy. For example, it is well known that UVO can be used to adjust the properties of PDMS from a hydrophobic surface to a hydrophilic surface. Efimenko et al showed that with UVO you can change the water contact angle of poly (vinylmethyl siloxane)

(PVMS) [117]. As the UVO exposure time was increased, the water contact angle of the PVMS surface would decrease.

Figure IV. 9 displays UVO results for PS. As the UVO exposure time increases, the water contact angle decreases for both a pure PS film before rinsing in toluene and pure PS after rinsing (residual layer) in toluene. Differences are observed in the rate that the contact angle measurements decrease. With a PS film, the layer is being stripped away but also oxidized at the surface. When UVO exposure is increased to a time of 30 minutes, the PS film is stripped away completely and the exposed silicon is oxidized giving water contact angles of 20˚ or less.

The PS residual layer is much more sensitive to the UVO exposure time, dropping the water contact angle to 20˚ or less in approximately 2 minutes time. Less than 2 minutes however, displays surface energy, or water contact angle, tunability. Water contact angles can range anywhere from 30˚ to 90˚ depending on the exposure time. Since UVO is an easy process and the residual layers are simple to fabricate, this could be a novel way in modifying the surface energy of a silicon substrate and possibly any other materials PS can bind to through thermal annealing.

4.7 Summary

Initial theory and hypothesis was that thermal annealing will encourage PS adsorption along the chain to a silicon oxide surface. Water contact angle measurements prove PS chains are strongly attaching to the surface of silicon oxide.

Even after different levels of rinsing and the use of different solvents, the residual layer

73 is unable to detach. Annealing time and annealing temperature both display a direct correlation of water contact angle measurements. As time or temperature is increased the contact angle increases, displaying a linear relationship with respect to temperature.

XPS and XR measurements not only confirm the presence of the residual layer but also display a direct correlation between Mw and the resulting thickness of the layer. Lastly, different PS chains can adsorb to the surface showing flexibility for polymer adsorption to the surface.

CHAPTER V

CONCLUSIONS

Thiol terminated and anionically polymerized PS has been proven to bind to the surface of silicon oxide through a thermal annealing process. Anionically polymerized PS is strongly adsorbed along the chain to the substrate surface while the thiol terminated

PS residual layer has end functionality that can graft to the surface at the chain end.

The attachment of PS layers through thermal annealing is a simple way of modifying the surface energy of a substrate. Furthermore, UVO exposure offers additional modification of the surface energy. This simple attachment method and surface energy modification scheme presents future possibilities into control of oriented block copolymer films without the use of complex chemistry.

The ease of RAFT polymerization and subsequent aminolysis to produce thiol terminated polymers offers flexibility for future attachment of polymer chains onto silicon oxide surfaces through sulfur and silicon interactions. Further analysis of sulfur to silicon bond needs to be investigated in an attempt to confirm and detail the nature of this behavior. In addition, the possibilities of PS adsorption to other surfaces can be

75 investigated as well since other polymers may strongly adsorb to silicon surfaces through thermal annealing.

CHAPTER VI

REFERENCES

1. Wool, R.P. (1995). Polymer Interfaces: Structure and Strength. New York: Hanser/Gardner Press

2. Koberstein J.T., Ed. (1996) Polymer Surfaces and Interfaces. MRS Bull.

3. Israelachvili J. (1992) Intermolecular and surface forces. Academic, New York

4. Napper D.H. (1983) Polymeric stabilization of colloidal dispersions. Academic, London

5. Ratner B. (1993) Jour. Biomed. Mater. Res., 27, 837

6. Mayes A. M., Kumar S. K. (1997) MRS Bull., 22, 43

7. Luzinov I., Minko S., Tsukruk, V.V. (2004) Prog. Poly. Sci., 29, 635

8. Yu K., Cong Y., Fu J., et al. (2004) Surface Science, 572, 490

9. Bates C.B., Strahan J.R., Santos L.J., et al. (2010) Lanmuir, 27(5), 2000

10. Ji S., Liu G., Zheng F., et al. (2008) Adv. Mater., 20, 3054

11. Zhao B., Brittain W.J. (2000) Prog. Poly. Sci., 25, 677

12. Minko S. (2006) Jour. of Macromol. Sci. Part C: Poly. Rev., 46, 397

13. Lemieux M., Usov D., Minko S., et al. (2003) Macromolecules, 36, 7244

14. Lemieux M., Minko S., Usov D., et al. (2003) Langmuir, 19, 6126

15. Lipson M. (2005) J. Lightw. Tech., 23(12), 4222

16. Irom F., Farmanesh F.H., Swift G.M., et al. (2003) IEEE, pp. 1

17. Muller A., Ghosh M., Sonnenschein R., et al. (2006) Mat. Sci. Eng. B, 134, 257

18. Sommer M., Huettner S., Thelakkat M. (2010) J. Mater. Chem., 20, 10788

19. Tang C., Lennon E.M., Fredrickson G.H., et al. (2008) Science, 322, 429

20. Wilson E.K. (2010) C&EN, 88(15), 29

21. Minko S., Stamm, M., Ed. (2008) Polymer Surfaces and Interfaces, 215

22. Bug A.L.R., Cates M.E., Safran S.A., et al. (1987) J. Chem. Phys., 87, 1824

23. Hiemenz P.C., Lodge T.P. (2007) Polymer Chemistry: second edition

24. Jordan R., Ulman A., Kang J.F., et al. (1999) Jour. Am. Chem. Soc. 121, 1016

25. Zhao B., Brittain W.J., (2000) Macromolecules, 33, 342

26. Prucker O., Ruhe J., (1998) Langmuir, 14, 6893

27. Luzinov I., Minko S., Senkovsky V., et al. (1998) Macromolecules, 31, 3945

28. Matyjaszewski K., Miller P.J., Shukla N., et al. (1999) Macromolecules, 32, 8716

29. Husseman M., Malmstrom E.E., McNamara M., et al. (1999) Macromolecules, 32, 1424

30. Jordan R., West N., Ulman A., et al. (2001) Macromolecules, 34, 1606

31. Alexander S.J. (1997) J. Phys., 38, 977

32. de Gennes P.G. (1980) Macromolecules, 13, 1069

33. Milner, S.T. (1991) Science, 251, 905

34. Mounir El Sayed A. (2002) Jour. Appl. Poly. Sci., 86, 1248

35. Schlenoff J.B., Dharia J.R., Xu H., et al. (1995) Macromolecules, 28, 4290

36. Koutsos V., Van der Vegte E.M., Pelletier E., et al. (1997) Macromolecules, 30, 4719

37. Rowe J.E., Margaritondo G., Christman S.B. (1977) Physical Rev. B, 16(4), 1581

38. Schonhammer K., Gunnarsson O. (1983) Physical Rev. B, 27(8), 5113

39. Narayana D., Lal J., Kasavulu V. (1968) J. Phys. Chem., 74(23), 1970

40. Cini M. (1975) Surface Science, 52, 75

41. Mansky P., Liu Y., Huang E., et al. (1997) Science, 275, 1458

42. Soga K., Zuckermann M.J., Guo H. (1996) Macromolecules, 29, 1998

43. Zhao B., Brittain W.J. (1999) Jour. Am. Chem. Soc., 121, 3557

44. Huang H., Penn L.S. (2005) Macromolecules, 38, 4837

45. Minko S., Patil S., Datsyuk V., et al. (2002) Langmuir, 18, 289

46. Luzinov I., Julthongpiput D., Malz H., et al. (2000) Macromolecules, 22, 1043

47. Luzinov I., Julthongpiput D., Tsukruk V.V., et al. (2000) Macromolecules, 33, 7629

48. Ulman A., (1996) Chem. Rev., 96, 1533

49. Tsukruk V.V., Luzinov I., Julthongpiput D., (1999) Langmuir, 15, 3029

50. Mansfield T.L., Iyengar D.R., Beaucage G. et al. (1995) Macromolecules, 28, 492

51. Koutsos V., van der Vegte E.W., Hadziioannou G., (1999) Macromolecules, 32, 1233

52. Kumacheva E., Klein J., Pincus P, et al. (1993) Macromolecules, 26, 6477

53. Anastassopoulos D.L., Vradis A.A., Toprakcioglu C., et al. (1998) Macromolecules, 31, 9369

54. Karim A., Tsukruk V.V., Douglas J.F., et al (1995) J. Phys. II, 5, 1441

55. Sirard S.M., Gupta R.R., Russell T.P., et al (2003) Macromolecules, 36, 3365

56. Taunton H., Toprakcioglu C., Fetters L.J., et al. (1990) Macromolecules, 23, 571

57. Schorr P.A., Kwan T.C., Kilbey S.M., et al. (2003) Macromolecules, 36, 389

58. Huang H., Penn L.S., Quirk R.P., et al. (2004) Macromolecules, 37, 5807

59. Luzinov I., Julthongpiput D., Liebmann-Vinson A., et al. (2000) Langmuir, 16, 504

60. Prucker O., Naumann C.A., Ruhe J. (1999) Journ. Amer. Chem. Soc., 121, 8766

61. Taylor W., Jones R.A.L. (2010) Langmuir, 26, 13954

62. Yu K., Cong Y., Fu J., et al. (2004) Surface Science, 572, 490

63. Bergbreiter D.E., Franchina J.G., Kabza K. (1999) Macromolecules, 32, 1233

64. Ebata K., Furukawa K., Matsumoto N. (1998) J. Am. Chem. Soc., 120, 7367

65. Prucker O., Naumann C.A., Ruhe J., et al. (1999) J. Am. Chem. Soc., 121, 8766

66. Feng J., Haasch R.R., Dyer D.J. (2004) Macromolecules, 37, 9525

67. Zhao B., Brittain W.J. (2000) Macromolecules, 33, 8813

68. Fang J., Knobler M. (1995) J. Phys. Chem., 99, 10425

69. Devaux C., Chapel J.P., Beyou E., et al. (2002) Eur. Phys. J.E., 7, 345

70. Jeyaprakash J.D., Samuel S., Dhamodharan R., et al. (2002) Macromol. Rapid Commun., 23, 277

71. Chen J-K., Zhuang A-L. (2010) J. Phys. Chem., 114, 11801

72. Sidorenko A., Minko S., Schenk-Meuser K., et al. (1999) Langmuir, 15, 8349

73. Tsujii Y., Ejaz M., Sato K., et al. (2001) Macromolecules, 34(26), 8872

74. Baum M., Brittain W.J. (2002) Macromolecules, 35(3), 610

75. Ito Y., Nishi S.W., Park Y.S., et al. (1997) Macromolecules, 30, 5856

76. Iwata H., Hirata I., Ikada Y. (1998) Macromolecules, 31(11), 3671

77. Halpern A., Tirrell M., Lodge T.P. (1992) Adv. Poly. Sci., 100, 31

78. Luzinov I., Voronov A., Minko S., et al. (1996) Macromolecules, 61, 1101

79. Boven G., Oosterling M.L.C.M., Chall G., et al. (1990) J. Polymer, 31, 2377

80. Prucker O., Ruhe J. (1998) Macromolecules, 31, 592

81. Biesalski M., Ruhe J., Johannsmann D.J. (1999) J. Chem. Phys, 111, 7029

82. Habicht J., Schmidt M., Ruhe J., et al. (1999) Langmuir, 15, 2460

83. Bug A.L.R., Cates M.E., Safran S.A., et al. (1987) J. Chem. Phys., 87, 1824

84. Esumi K., Oyama M. (1993) Langmuir, 9, 2020

85. Otsuka H., Esumi K. (1993) Langmuir, 10, 45

86. Peyser P., Ullman R. (1965) Journ. Poly. Sci.: Part A, 3, 3165

87. Frantz P., Granick S. (1991) Physical Review Letters, 66 (7), 899

88. Frantz P., Granick S. (1992) Langmuir, 8, 1176

89. Johnson H.E., Granick S. (1990) Macromolecules, 23 (13), 3367

90. Enriquez E.P., Schneider H.M., Granick S. (1995) Jour. Poly. Sci.: Part B: Poly. Phys., 33, 2429

91. Johnson H.E., Douglas J.F., Granick S. (1993) Physical Review Letters, 70 (21), 3267

92. Guiselin O. (1992) Europhys. Lett., 17 (3), 225

93. O’Shaughnessy B., Vavylonis D. (2003) Europhys. Lett. (preprint)

94. In I., La Y-H., Park S-M., et al. (2006) Langmuir, 22, 7855

95. Raviv U., Klein J., Witten T.A. (2002) Europhys. Lett. J. E., 9, 405

96. Kelley T.W., Schorr P.A., Johnson K.D., et al. (1998) Macromolecules, 31(13), 4297

97. Parsonage E., Tirrell M., Watanabe H., et al. (1991) Macromolecules, 24(8), 1987

98. Fytas G., Anastasiadis S.H., Seghrouchni R., et al. (1996) Science, 274, 2041

99. Motschmann H., Stamm M., Toprakcioglu C. (1991) Macromolecules, 24(12), 3681

100. Alcoutlabi M., McKenna G.B. (2005) J. Phys. Condens. Matter., 17, 461

101. Baschnagel J., Varnik F. (2005) J. Phys. Condens. Matter., 17, 851

102. Roth C.B., Dutcher J.R. (2005) J. Electroanal. Chem., 584, 13

103. Yang, Z., Fujii Y., Lee F.K., et al. (2010) Science, 328, 1676

104. Tsui O.K.C., Russell T.P. (2008) Series in Soft Condensed Matter, ,267

105. Wallace W.E., van Zanten J.H., Wu W-L. (1995) Phys. Rev. E., 52, 3329

106. Keddie J.L., Jones R.A.L., Cory R.A. (1994) Europhys. Lett., 27, 59

107. Forrest J.A., Dalncki-Veress K., Stevens J.R., et al. (1996) Phys. Rev. Lett., 77, 2002

108. Fujii Y., Yang Z., Leach J., et al. (2009) Macromolecules, 42, 7418

109. Napolitano S., Wubbenhorst M. (2011) Nature Communications, 2:260, DOI: 10.1038/ncomms1259

110. Liu Y., Cavicchi K. (2009) Macromol. Chem. Phys., 210, 1647

111. Koutsos V., Van der Vegte E.M., Hadziioannou G. (1999) Macromolecules, 32, 1233

112. Stuen K.O, In I., Han E., et al. (2007) J. Vac. Sci. Technol. B, 25(6), 1958

113. Coulter S.K., Schwartz M.P., Hamers R.J. (2001) J. Phys. Chem. B, 105, 3079

114. Mirji S.A., Halligudi S.B., Sawant, Dhanashri P., et al. (2006) Colloids and Surfaces A: Physiochem. Eng. Aspects, 272, 220

115. Hinkley J.A., National Bureau of Standards, “Adsorption of Polystyrene on Thermally Oxidized Silicon”

116. Smith A.P., Douglas J.F., Meredith J.C., et al. (2001) Phys. Rev. Lett., 87(1), 015503

117. Efimenko K., Crowe J.A., Manias E., et al. (2005) Polymer, 46, 9329