Oriented inorganic particles prepared in a Langmuir-Blodgett film


Material Information

Oriented inorganic particles prepared in a Langmuir-Blodgett film
Physical Description:
ix, 179 leaves : ill., photos ; 29 cm.
Pike, John K., 1960-
Publication Date:


Subjects / Keywords:
Thin films multilayered   ( lcsh )
Chemistry, Inorganic   ( lcsh )
Chemistry thesis Ph.D
Dissertations, Academic -- Chemistry -- UF
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1994.
Includes bibliographical references (leaves 165-178).
Statement of Responsibility:
by John K. Pike.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002021961
oclc - 32838537
notis - AKK9455
System ID:

Full Text







To Mom and Dad.


I would like to thank Dan Talham for making a couple of completions to a big,

slow, tight-end (with great hands!) on a field in Gainesville a couple of years ago.

Looking back, I don't know which was the more courageous act--putting my graduate

education in the hands of a brand new professor with no lab and little funding or the

same young professor trusting his research to an equally new and very average graduate

student. I made a pretty good choice. Daniel R. Talham is one of the most capable,

level-headed individuals I have ever had the privilege to work with and I am glad to have

had the opportunity. He has given me nothing but good advice, made me more confident

of my ability, and shown me through his example what is required of a good scientist.

I would like to thank my fellow charter group members, Dr. Houston Byrd and

Mrs. Margaret Showalter. Special thanks go to Houston for his hard work and

innovation. He kept the lab "up and running" by providing genuine inspiration (scientific

and spiritual) when it was needed.

I wish to thank my coauthor and expert electron microscopist, Dr. Augusto A.

Morrone, for all the conversations, arguments, and "shots in the dark." Thanks are due

to Mr. Eric Lambers and the Major Analytical Instrumentation Center for trusting me with

the good XPS instrument. Thanks also go to Dr. Greg Erdos, Donna, and Lorraine at the

Core Electron Microscopy Lab for taking a wayward chemist into their darkroom.


I would like to thank my committee members, Dr. James M. Boncella, Dr. Russell

S. Drago, Dr. Mark W. Meisel, and Dr. Willis B. Person, for their encouragement and

support. I would also like to acknowledge the University of Florida for providing a nice

place in which to do chemistry, go to football games, and watch basketball. Go Gators!

I would also like to thank Jeanne Zachary for her help, kind words, and good coffee.

Very special thanks go to Jeri, Joel, Bill, Michael, Gary, Deborah, Gail, Eileen,

Ken, Margaret, Maya, Phyllip, and Alicia for always being right there. I would like to

thank my Grandma Pike and Grandma Lee for their inspiration and unfailing belief. I

would especially like to thank Virgil and Marvel Johnson for their love and


I would like to thank my parents, Bette and Ray Pike, for their unconditional love,

acceptance, and support throughout my entire life. I wish to thank my Mom, for giving

me her laughter and who never gave up on me, and thank my Dad for his gentle nature

and for instilling in me a love of reading. Thanks also for the open door policy at the

Food Palace!

Lastly, I would like to express my sincere gratitude to my wife and best friend

Sherri and profess my love for her with all that I am and all that I have. I am truly

blessed to have her delightful companionship on a voyage which, because of her, gets

better every day. And we are both very much blessed by a little traveler who arrived just

last year. He is bright and adventurous and his name is William Johnson Pike. I am very

thankful to have such a fine son.


ACKNOWLEDGMENTS ....................................... iii

ABSTRACT ................................................. viii


1. INTRODUCTION .................................. 1

Scope of Dissertation ................................ 1
The Organic Template ............................... 4
Micelles and Reversed Micelles ................... 4
V esicles .................................... 6
Bilayer Lipid Membranes ........................ 7
Tubular M icrostructures ......................... 8
Cast M ultibilayer Films ......................... 9
Floating Langmuir Monolayers ................... 10
Crystallization from supersaturated solutions ..... 12
Crystallization by chemical reaction ........... 17
Electrocrystallization ...................... 21
Deposition of pre-formed clusters ............. 21
Deposited Langmuir-Blodgett Films ................ 23
Insertion-reduction templating ............... 25
Reaction with gaseous chalcogenides ......... 26
Other Template Systems ........................ 28
Requirements for Oriented Crystallization ................. 29

FIL M ...... ... ... ...... .. .... .... .. ... ... ... 32

Introduction ...................................... 32
Objective .......................... ...... 32
Cadmium Dihalide Structure ...................... 33
Experimental Details ................................ 35
M materials ................................... 35

Substrate Preparation ..............
Instrumentation .................
Langmuir-Blodgett trough .....
Transmission electron microscope
FT-infrared spectrometer ......
X-ray photoelectron spectrometer
Procedure ...........................
Results .................. ...........
Attenuated Total Reflectance FT-Infrared
X-Ray Photoelectron Spectroscopy ....
Transmission Electron Microscopy ....
D discussion ..........................
Organic Template Reaction .........
Particle Orientation ...............
Lattice Matching .................

. . .
. .
. .
. . .
. . .
. . .

. . .
. . .
. .. .
. . .
. . .
. . .

FILM ................................. .........


O objective ...................................

Manganese Dihalide Structure .......
Experimental Details ...................
Materials and Procedures ..........
R results ..... .. ... ... .. .. .... .. ... ..
Attenuated Total Reflectance FT-Infrared
X-Ray Photoelectron Spectroscopy ...
Transmission Electron Microscopy ...
D discussion ..........................
Organic Template Reaction .........
Particle Orientation ...............
Lattice M watching .................

. . 80

. . .
. . .
. . .

. . .
. . .
. . .

. . .
. .. .
. . .

4. XPS OF TMNIN .......................

Introduction ..........................
Structure of TMNIN ...............
The Chemical Shift in the XPS ........
Experimental Details ....................
Development of the TMNIN XPS Project
Model Compounds for XPS Study .....
Method of Sample Preparation ........
XPS Instrument Conditions ..........

Results and Discussion ............................... 122
Composition from Survey Spectra .................. 122
Variable Take-Off Angle Experiment ............... 123
Argon Sputtering ............................. 131
Sputter Effects on the Ni(2p) Region ................ 133
Curve Fitting of XPS Data ..................... .. 150
Conclusion ....................................... 156

5. CONCLUSIONS AND FUTURE WORK ................. 158

Oriented Metal Dihalide Particles in Langmuir-Blodgett Films ... 158
Future W ork ...................................... 161

APPENDIX ION SPUTTER OPERATION ........................... 163

REFERENCE LIST ........................................... 165

BIOGRAPHICAL SKETCH ..................................... 179

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



John K. Pike

August, 1994

Chairperson: Dr. Daniel R. Talham
Major Department: Chemistry

For the first time, oriented arrays of inorganic particles have been synthesized in

a Langmuir-Blodgett (LB) film. Particles of MX2, where M = Cd and Mn and X = I, Br,

and Cl, are formed by the reaction of a deposited cadmium or manganese arachidate LB

film with the corresponding gaseous hydrogen halide. The reactions are monitored by

attenuated total reflectance FT-infrared spectroscopy which shows that they go to

completion and that the organic LB film remains crystalline afterward. X-ray

photoelectron spectroscopy and transmission electron diffraction are used to identify the

metal halides and show that in each case only one inorganic species is formed.

Transmission electron diffraction shows that CdI2 particles are formed exclusively with

their [001] crystal axes perpendicular to the plane of the LB film and for domain sizes

up to several im2 discrete particles have the same in-plane orientation. CdBr2 and CdCl2

particles are each observed to form with a mixture of two discrete orientations; [001] axis

parallel or [001] axis perpendicular to the LB plane. MnI2, MnBr2, and MnC12 particles

are also oriented with their [001] axes either parallel or perpendicular to the plane but to

a lesser degree than the cadmium halides. In all of the samples, domains are observed

where arrays of particles exhibit common in-plane orientations. It is proposed that the

organized organic matrix is responsible for orienting the inorganic particles, and potential

lattice matching between the inorganic particles and the organic matrix is discussed.

In a separate project, an X-ray photoelectron spectroscopic investigation was

performed on the novel one-dimensional magnetic chain compound tetramethylammonium

nickelnitrite. This material was synthesized in our lab and has an interesting crystal

structure in which there are two chemically distinct nickel centers. The results of

experiments aimed at resolving those two nickel centers are discussed.


Scope of Dissertation

This dissertation presents the experimental results from the synthesis and

characterization of metal halide particles in deposited Langmuir-Blodgett films. Chapter

1 will provide a literature review on the subject of template-directed crystallization of

particles. From experiments in our laboratory and from the literature, a list of

requirements for "good" template-directed crystallization will be presented. Chapter 2

will describe the synthesis and characterization of oriented cadmium halides in a

deposited Langmuir-Blodgett film. The work in Chapter 2 is the first observation of

oriented inorganic particle formation in Langmuir-Blodgett films. Chapter 3 will describe

the formation of oriented manganese halides in deposited Langmuir-Blodgett films.

Chapters 2 and 3 will include a summary at the end of each chapter. Also included in

Chapter 3 will be a comparison of the cadmium and manganese systems with regard to

the observed differences in their orientation as it relates to their structures. Chapter 4 will

present in detail the results of an extensive X-ray photoelectron spectroscopic

investigation of a novel magnetic compound synthesized in our laboratory by Chou Liang-



The original goal of this work was to use an organized organic template to develop

a magnetic monolayer. The magnetic monolayer would be electronically isolated from

the organic template and ideally be a monolayer analog of a bulk inorganic extended

lattice. In this way, the magnetic properties of the monolayer could be compared to those

of the bulk material. Organized organic assemblies have frequently been used in schemes

to prepare inorganic particles in which the organic structure serves as either a template

or a reaction vessel of limited size. The general goal is to produce inorganic particles

with limited size, dispersion, or orientation. This approach draws inspiration from

naturally occurring examples of organic-mediated crystallization of inorganic materials.

While we have not yet realized the goal of a monolayer material, the observation of

oriented inorganic particles formed within the organized organic template encourages us

to think that a single layer of inorganic is possible. In the laboratory, there are several

advantages to using organic molecules as templates. From a "design" standpoint, there

is more "architectural freedom" using organic molecules because of the large number of

organic compounds and structures available. The means of modifying existing organic

molecules to achieve a desired structure or property are usually within the grasp of a good

synthetic chemist. In contrast to "dry" surface preparative methods of chemical vapor

deposition or molecular beam epitaxy, the actual templating process onto organic films

usually occurs via "wet" chemistry on the benchtop. More often than not, organic

materials have the advantage of being electronically and magnetically transparent. In

contrast to "inorganic" host materials such as zeolites, aluminosilicates, and clays, organic

template materials are usually more amenable to modification. In addition, organic host

systems are seen as being more "flexible" and may accommodate a variety of inorganic

guests without requiring a perfect epitaxial match. Although it is not always the case,

laboratory analysis is often very straightforward for these organic template systems. The

advantage of using organic materials for template reactions then, lies in their overall

flexibility and accessibility.

If only it were as simple as "designing" the perfect organic template, performing

a simple benchtop reaction, analyzing the results, and harvesting the bounty of perfectly

mediated inorganic crystals! As usual, nature continues to achieve the best results and

those in the laboratory must do the best they can. The introductory chapter of this

dissertation will describe the systems in which organic have been used successfully to

control or mediate the growth of inorganic (and organic) crystals. It is worth stating that

the dimensions of the inorganic particles as formed are somewhere between the molecular

and the bulk level, usually on the order of several nanometers, and it is this size effect

which often gives rise to some interesting physical properties. The potential device

applications arising from these "quantum" size effects are the driving force for some of

the research. Some of these so-called "nanophase" particles that have been made by

organic templating include semiconductors, catalysts, magnetic materials and metallic

particles. But there is much to be learned regarding biological organic-inorganic

interfaces as well. The organic systems described herein are surfactants, having a

hydrophobic and a hydrophilic region. They may take on a variety of dimensions and

shapes; from a micelle tens of angstroms across to a vesicle thousands of angstroms in

size to a thin film with good crystalline order extending over several millimeters in two


dimensions but only a few nanometers in the third. The interfacial region between the

organic and the inorganic is not fully understood in most systems, and the defining

interactions are thought to be a combination of electrostatic, geometric, stereochemical,

and kinetic effects. Taken together, this is an exciting area of research and one which is

of interest across scientific disciplines for a variety of reasons.

The Organic Template

Micelles and Reversed Micelles

A surfactant molecule has a polar, hydrophilic part and a nonpolar, hydrophobic

part, and is then said to be amphiphilic. Irving Langmuir first discovered that a single

molecule may have two different "independent surface actions" (Langmuir, 1917). If the

hydrophobic "tails," usually alkyl chains, are short enough the polar "headgroup" will

dominate the solubility and the molecule is water soluble. In water, when the

concentration of surfactant molecules reaches a characteristic concentration, groups of 50-

100 surfactant molecules will spontaneously associate into spherical or rod-like assemblies

known as micelles (Fendler, 1987). In a micelle, the surfactant molecules are arranged

with their hydrophobic tails directed into the center of the micellar body and the

hydrophilic or polar part of the surfactant molecule is directed outward so as to be in

contact with the aqueous solution. It is a cooperative arrangement; opposing forces of

repulsion between the polar headgroups and associative van der Waals forces between the

hydrophobic tails are responsible for micellation (Fendler, 1987). Micelles are stable for

months in aqueous solutions so long as their concentration is kept above the critical level

required for micellation (Fendler, 1987). Although they may surround pockets of non-

polar solvents, micelles do not solubilize ionic molecules and traditionally are not used

as organic templates. However, a new family of mesoporous silicate-aluminosilicate

structures first reported by Beck (Beck et al., 1992; Kresge et al., 1992) employs

trimethylammonium surfactant molecules which, perhaps through a micellar intermediate,

appear to template and direct the formation of the silicate framework. A lamellar to

micellar phase transition has been observed in cetyltrimethylammonium bromide

mesophases (Auvray et al., 1989). Stucky (Monnier et al., 1993) reports a similar phase

transition in the synthesis of the new silicate mesostructures and outlines a model for their


Because of their ability to solubilize a large number of water molecules reversed

or inverted micelles are employed frequently to template inorganic particles (Kortan et

al., 1990; Chandler; Coffer, 1993; Petit et al., 1993). A reverse micelle forms when

surfactants having the appropriate hydrophobic lipophilic balance undergo self-association

in a non-polar solvent. In the reverse micelle the surfactant molecules are oriented with

their polar headgroups directed toward the center while the lipophilic tails are directed

outward toward the non-polar solvent. The interior "cavity" formed by the polar head

groups can then immobilize relatively large pools of water. For example, one molecule

of sodium bis(2-ethylhexyl)sulfosuccinate, or "AOT," is able to solubilize up to 60 water

molecules in a hydrocarbon solvent (Fendler, 1987). Spherical silver "nanoparticles" 27-

75 A in diameter were obtained by the reduction of Ag+ ions in AOT reverse micelles

using hydrazine and sodium borohydride in an isooctane solvent (Petit et al., 1993). The

silver clusters were extracted after centrifugation into water and it was found that they

had not increased in size, indicating that the clusters retained their surfactant coating.

Inverse micelle solutions provide a good medium for the synthesis of stable, size-selected

semiconductor particles of CdS (Dameron et al., 1989), and CdSe on ZnS (Kortan et al.,

1990). There remains the tendency for these small clusters to agglomerate, however, and

clever means have been reported of chemical surface capping and passivating by

thiophenol attachment to CdS clusters (Herron et al., 1990; Chandler; Coffer, 1993) and

by trimethylsilane attachment to CdSe, HgSe, and CdTe clusters (Steigerwald et al.,



In contrast to the monolayered micelle and reversed micelle, vesicles are

comprised of bilayers of surfactant molecules. The vesicle wall is made up of two

surfactant molecules oriented tail-to-tail; the polar headgroup of the outer surfactant is

directed outward, that of the inner surfactant is pointed toward the interior of the vesicle

body. Vesicles are formed by sonication of surfactants in an aqueous solution and are in

general much larger (> 10,000A in diameter), and more durable than micelles vesicless

are not destroyed by H20 dilution) (Fendler, 1987). Vesicles may have single

(bimolecular) or unilamellar walls, or multilamellar walls and several compartments with

room for solvent between the walls. Because of their similarity to biological cellular

systems, vesicles have traditionally been the workhorse of the research aimed at biological

control of mineralization or "biomineralization" (Mann, 1983; Mann et al., 1984; Mann

et al., 1993).

A classic example of vesicle-mediated inorganic crystal growth is described by

Williams and co-workers (Mann et al., 1986). Using unilamellar phospholipid vesicles

to encapsulate various solutions of Fe(II) and Fe(III), they observed by electron

microscopy that the iron was bound to the inside perimeter of the 300A diameter vesicles.

Upon raising the pH with NaOH, the iron was oxidized and the precipitate was

constrained within the vesicle, in contrast to non-encapsulated bulk iron solutions. The

bulk Fe(II) solution precipitate was a mixture of poorly ordered ac- and y-FeOOH crystals

but the intravesicular precipitate produced disk or spherically shaped Fe304 crystals with

an average diameter of 30A-50A. It was proposed that the rate of infusion of the OH-

ions into the vesicles, and subsequent Fe(II) oxidation, was controlled by the slower

effusion of the intravesicular C1- ions which favor the Fe304 product. In addition, it was

proposed that the observed spherical particle shape was due to the initial nucleation of the

oxidized iron occurring at the lipid interfacial binding sites on the vesicle interior. In

another study Fendler (Herve et al., 1984) observed an -87% increase in the alcohol

product of the photoreduction of benzophenone when the reaction was run within a

Fe304-containing vesicle.

Bilayer Lipid Membranes

Bilayer lipid membranes are formed by brushing an organic solution of a

surfactant or lipid phospholipidd) across a pinhole which separates two aqueous phases

(Fendler, 1987). The resulting single bimolecular layer spans the pinhole and, like the

wall of a unilamellar vesicle, has the polar headgroups directed outward. Anionic bilayer

lipid membranes of glyceryl monooleate have been observed (Zhao et al., 1989) to bind

cationic Fe304 particles exclusively at the polar headgroup region of the membrane. The

interesting result is that penetration of the particles to the hydrophobic interior of the

membrane was not observed. The authors (Zhao et al., 1989) reported a monolayer film

of magnetite particles bound so strongly to the bilayer lipid membrane by the electrostatic

interaction that the particles could not be pulled off of the film with a magnet. In another

study (Baral; Fendler, 1989), a film of CdS was grown on one side of a bilayer lipid

membrane and was used to mediate photoelectric effects. It was proposed that a vectorial

charge-transfer across the semiconductor-coated membrane was enhanced due to the

asymmetric membrane potential as a result of the adsorbed CdS particles (Baral; Fendler,


Tubular Microstructures

Phospholipid tubules are hollow, open-ended structures with a high aspect ratio

(0.5plm in diameter and 50-100pm in length) and are formed by the spontaneous self-

aggregation of chiral phospholipid molecules (Baral; Schoen, 1993) and more recently

sugar based surfactants (Archibald; Mann, 1993). The walls are similar to those of

unilamellar or multilamellar vesicles and the tubules are stable in aqueous solutions below

a critical melting temperature (Baral; Schoen, 1993). Using a sol-gel synthesis in an

aqueous dispersion of phospholipid tubules, a 50A-thick film of silica was self-assembled

onto the tubules (Baral; Schoen, 1993). The resulting silica-coated tubules were hollow

and the silica film had an amorphous structure. The authors observed that the

phospholipid could apparently be baked out of the silica tubules which retained their

shape, and the tubules were subsequently coated with nickel or palladium (Baral; Schoen,

1993). In a more recent communication (Archibald; Mann, 1993), sugar-based tubules

were used to mediate the mineralization of iron oxides. Although the resulting iron

oxidation products were mixed a- and y-FeOOH, it was found that chemical enrichment

of certain phases of the tubules prior to mineralization resulted in different morphologies

of the templated iron oxides. In addition, treatment of the y-FeOOH coated tubules with

Fe(II) under N2 resulted in the formation of Fe304 and retention of the tubular structure.

In each of the above studies (Baral; Schoen, 1993; Archibald; Mann, 1993) templating

of the inorganic product was not limited to the interior of the tubules.

Cast Multibilayer Films

Cast films of N-(&o-(Triethylammonio)undecanoyl)dioctadecyl-L-glutamate bromide

have been reported (Kimizuka et al., 1993) to give several orders of diffraction with

grazing angle X-rays indicating that the films are in the form of multibilayers.

"Dimension-diminished" clusters of anionic PbBr42- molecules have been synthesized in

these films by ion-exchange (Kimizuka et al., 1993). In this study, cross-sections of these

assemblies observed in a scanning electron microscope confirm that the films retain their

layered nature. Other research by the same group (Okada et al., 1990) reported the

formation of oriented particles of Fe304 within the cast multibilayer films. Electron

diffraction confirmed the magnetite structure, but not a preferred orientation. Scanning

electron microscopy showed the presence of layers in the particle-containing cast film,

however. Magnetic studies of the particle-containing cast film indicated some magnetic

anisotropy (Okada et al., 1990).

Floating Langmuir Monolayers

Recall that a surfactant molecule is amphiphilic in nature, having a hydrophobic

alkyl chain and a hydrophilic polar headgroup. If the alkyl chain of a surfactant molecule

is long enough (approximately greater than 13 carbons), the surfactant molecule does not

dissolve in aqueous solution but rather forms a film of monomolecular thickness on the

surface (Roberts, 1990) with the hydrophobic part of the molecule directed away from the

aqueous phase, and the hydrophilic part in contact with the aqueous phase (Langmuir,

1917). Because part of the surfactant molecule is "dissolved" in the water, the surface

tension of the water is reduced according to the Gibbs' adsorption isotherm (Gaines,

1966). Surface tension is the force, due to unbalanced molecular attractions, which tends

to pull molecules into the interior of a liquid phase. All else being equal, the surface

tension is a function of the chemical potential of the solution's components. Using the

Gibbs' isotherm, it is possible to evaluate the surface excess of a solute from the observed

variation in surface tension. If the surface tension of a solution decreases as the solute

activity increases, then there must be a positive excess or adsorption of the solute in the

surface region of the solvent (Adamson, 1967). The converse is also true: if the

concentration of surface-confined solute molecules can be varied (by changing their


available area) the result will be a change in surface tension. Langmuir observed that

certain materials effected a change in the surface tension of water which could be varied

by manipulating the "water" surface with a mechanical barrier. The surface tension of

water at 25C is 71.97 mN/m (1986). With an adsorbed layer of stearic acid it is

typically reduced to between 20-50 mN/m. Langmuir suggested that it is perhaps more

intuitive to think in terms of the lateral pressure exerted by the surface film and

developed the term "surface pressure" denoted by the symbol, t (Gaines, 1966). The

relation between surface tension, y, and surface pressure, 7t, is

R = 70- Y (1-1)

where y0 is the surface tension of the pristine liquid and y is the surface tension of the

surfactant-coated liquid. Surface pressure, then, is usually a positive number and, in fact,

the surface "pressure" effects of these monolayer films can be pretty amazing to watch

(Pike, Byrd, unpublished results).

Langmuir monolayers are formed by spreading a solution of surfactant molecules

onto the surface of an aqueous subphase. The surfactant solution solvent (usually

chloroform or hexane) evaporates leaving the surfactant molecules on the water surface.

A mechanical barrier is employed to "sweep" the surfactant molecules to one end of the

water surface, until a desired surface pressure is attained. Langmuir monolayers may be

quite durable, lasting for many hours at constant surface pressure in our lab, although

there are reports of floating monolayer films lasting for days or weeks (Fendler, 1987).

At various combinations of temperatures, surface pressures and pH, floating Langmuir

monolayers exhibit distinct and stable phases (Durbin et al., 1994). The undersurface of


the Langmuir monolayer forms a periodic, and depending upon pH, charged surface.

Recently, researchers have exploited the ability of these floating organic "structural"

surfaces to mimic crystalline habits and ionicities in order to nucleate and template

inorganic and organic materials from solution. This section will describe some of the

elegant means by which floating Langmuir monolayers are used to accomplish this, and

the results will be discussed in terms of the geometric, electrostatic and stereochemical

effects on crystallization.

Crystallization from supersaturated solutions

The objective is to use floating organic monolayers as "surfaces" to nucleate and

effect the crystallization of inorganic salts from supersaturated aqueous solutions.

Langmuir monolayers are prepared at some constant surface pressure over a

supersaturated aqueous solution (hereafter "subphase") and over time the subphase ions

nucleate onto the undersurface of the floating monolayer, eventually growing into small

particles which may remain attached to the surface film. The entire surface monolayer-

nascent-crystal assembly is lifted off of the surface and transferred intact usually to a light

or electron microscope where the crystals are studied. The surface film and particle

assembly have sometimes been observed directly on the trough by low angle X-ray

diffraction. In addition, the remaining subphase can be characterized by chromatographic

or spectroscopic methods.

In a pioneering study, Landau and co-workers (Landau et al., 1985) observed that

floating Langmuir monolayers of chiral a-amino acids induced the formation of oriented

crystals of a-glycine from a racemic subphase. Homochiral monolayers of the R


configuration of a-amino acids induced subphase glycine crystals to form with their (010)

or "R" face attached to the undersurface of the monolayer. Monolayers of the S

configuration caused the glycine to crystallize with the (010) or "S" face attached to the

monolayer. The oriented crystallization was observed to occur within seconds after the

monolayer was compressed to the limiting surface area per molecule (-25A). Monolayers

that were 100% in R or S molecules resulted in -90% "R-face" or "S-face" glycine

crystallization, respectively. Fifty-fifty R, S monolayer mixtures gave approximately 50-

50 mixtures of (010) and (010) glycine crystals. These percentages were determined by

HPLC analysis of the attached crystals. When the monolayers were replaced by

amphiphiles having the same amino acid polar headgroups (-(NH3+)-CO2") with a

modified hydrocarbon tail having a larger area per molecule (38A vs. 25A) glycine

crystallization was observed to be slow and unoriented. The authors (Landau et al., 1985)

proposed that a structural match between the face presented by the chiral monolayer and

the face of the burgeoning glycine crystals is responsible for the observed orientation.

More recent work by the same group (Weissbuch et al., 1988) describes a model in which

kinetic and hydrophobic effects determine the nature and growth of a-amino acids on the

face of glycine substrates. Recently, the same group (Landau et al., 1989) deposited the

floating amino acid monolayers onto solid substrates as Langmuir-Blodgett films and

observed the behavior of glycine crystallization. They report identical behavior in that

the glycine crystallized in an enantioselective manner whenever a favorable nucleating

surface was presented.

Floating monolayers were found (Landau et al., 1986) to induce oriented

crystallization of solid sodium chloride from a 6M NaCI subphase. The crystal

orientation of the NaCI was dependent upon the type of surfactant used. Cationic,

anionic, and zwitterionic monolayers were employed in the study (Landau et al., 1986).

Monolayers of long-chain carboxylates (R-CO2) induced crystallization in which the

(111) face of NaCl was in contact with the monolayer. The (111) NaCI face can be

constructed to present a hexagonal array of sodium ions. The authors state that the Na-Na

packing in such a projection is similar in distance and symmetry to the headgroup packing

in the floating long-chain carboxylate monolayer (Landau et al., 1986). Although the

symmetry of each face (monolayer and crystal) is hexagonally symmetrical, the Na-Na

spacing given in the paper for the (111) projection is somewhat smaller (by about 0.6A)

than that of the stearic acid monolayer. However, when the hydrocarbon chains were

replaced by bulkier steroid groups, no nucleation under the monolayer was observed.

Cationic monolayers of octadecylamine were observed to bind NaCl crystals in the (100)

projection. The (100) projection of NaCl is an alternating cubic array of (+) and (-)

charges. It was proposed that the repeat surface area of the NaCI (100) face was very

closely matched to that of the octadecylamine monolayer and it was further suggested that

the chloride ions penetrate between the octadecylammonium molecules to form such a

structure (Landau et al., 1986). Zwitterionic amphiphilic monolayers with (R-(NH3+)-

CO2-) headgroups induced the formation of the (110) face of NaCI with 70% a selectivity

at a pH of 1-3. The (110) NaCl face consists of rows of alternating (+) and (-) charges.

When the pH was increased to 4-6, the percentage of (110) crystals dropped to about 40-

50% and (111) crystals started to appear. The authors (Landau et al., 1986) mentioned

that the (110) and (111) faces do not occur naturally, rather, bulk NaCI crystallizes

exhibiting the (100) face preferentially. In each case, when the limiting mean molecular

area of the monolayer was increased, either by the addition of bulky side groups or by

decreasing the surface pressure, crystallization was slow and preferred orientation was not

observed. These results suggest that preferential crystallization at floating monolayers is

a combination of structural and electrostatic effects. Oriented crystallization is determined

by the extent to which the symmetry, periodic spacing, and charge distribution of the

"host" template can mimic a face of the nascent "guest" crystal.

Using floating stearic acid monolayers, another group (Mann et al., 1988) observed

the role of the organic template in determining the orientation, morphology, and the

crystal structure of CaCO3 grown from a supersaturated solution. Crystallization in the

absence of the monolayer resulted in the formation of the rhombohedral calcite form of

CaCO3 whereas in the presence of the monolayer the vaterite form was observed (Mann

et al., 1988). Calcite and vaterite are polymorphs of calcium carbonate. In the semi-

compressed monolayer, the immature vaterite (monolayer-induced) crystals were very well

oriented with their (001) face parallel to the monolayer and had a narrow particle size

distribution (2.2pm 0.2pm) whereas in the fully compressed monolayer the authors

report that the vaterite crystals are much smaller (1.3pm 0.5pm). The authors suggest

that since the Ca-Ca packing density is smaller in the vaterite than in the fully

compressed stearic acid monolayer by about 0.5A and certainly smaller still than the

semi-compressed stearic acid monolayer, geometric matching is not as important as are

stereochemical and electrostatic matching. Apparently the orientation of the carboxylate

group is of importance; the carboxylate groups are perpendicular to the vaterite [110] axes

and parallel to the calcite [110] axes. The authors report that the stearic acid monolayer

carboxylate groups are oriented similar to those in vaterite, and even though there is a

closer lattice match between calcite and the monolayer with respect to packing density,

the crystals grow with the vaterite structure (Mann et al., 1988; Mann et al., 1990). In

addition, the more liquid-like (semi-compressed) stearic acid monolayer promotes more

uniform vaterite crystallization, indicating that some flexibility is required of the organic

template in this case.

A more complete study was presented by the same authors (Rajam et al., 1991;

Heywood et al., 1991) in which the effect of different surfactant molecules on CaCO3

crystallization was observed. Floating Langmuir monolayers of octadecanol,

octadecylamine, cholesterol in addition to stearic acid were employed to induce CaCO3

crystals to grow from a supersaturated solution (Rajam et al., 1991; Heywood et al.,

1991). The crystals were observed by transmission electron diffraction. To complement

the above-mentioned study (Mann et al., 1988; Mann et al., 1990) stearic acid was found

to grow oriented calcite or vaterite depending on the subphase calcium concentration

(high [Ca2+] gives calcite, low [Ca2+] gives vaterite). Octadecylamine monolayers grew

two vaterite orientations independent of calcium concentration. Octadecanol inhibited

crystallization completely, and cholesterol monolayers gave rise to random, non-oriented

calcite crystals (Rajam et al., 1991; Heywood et al., 1991). To account for the

observations, the authors cite geometric and stereochemical matching between the

monolayer carboxylate headgroups and the ions in the (110) crystal face of calcite. For

vaterite on stearic acid, there is no geometric match, but there is a stereochemical

relationship similar to the 1988 paper (Mann et al., 1988). For octadecylamine, Heywood

and Mann (1991) suggest that bidentate binding of HCO3- may be important in

determining the preferential orientation of the vaterite. This report (Rajam et al., 1991;

Heywood et al., 1991) in two parts and a very recent paper by the same group (Heywood;

Mann, 1994) shows how geometric, stereochemical, and electrostatic factors in floating

organic templates may be manipulated extensively to mediate the growth of an inorganic

crystal. Recently, Heywood and Mann have grown BaSO4 crystals under floating

Langmuir monolayers of long-chain carboxylates and sulfates (Heywood; Mann, 1992;

Heywood; Mann, 1992). Again, oriented crystals were observed in both systems, but the

authors observed crystals of irregular shape in the carboxylate-grown BaSO4 and suggest

that, because no geometric or stereochemical lattice match exists, the kinetic product is

being formed as a result of electrostatic interactions or charge accumulation at the

template-crystal interface. Heywood and Mann publish with an eye on the biological

aspects of organic-template directed crystallization and are leaders in the field of

crystallization from supersaturated solutions onto floating monolayers. The interested

reader is invited to peruse Mann, who has written several good reviews on the subject

(Mann, 1983; Mann, 1988; Mann et al., 1990; Mann, 1993).

Crystallization by chemical reaction

This section will outline the work in which crystallization is induced by proactive

chemical means at the undersurface of a floating Langmuir monolayer. Normally this is

accomplished by the slow infusion of a gas (usually an acid chalcogenide) over the

surface of a Langmuir trough while a Langmuir monolayer is maintained at constant

surface pressure. The gas reacts with ions in the aqueous subphase causing crystals to

nucleate and form along the monolayer undersurface. There are reports of particles

generated in situ by wet chemical means, however, and they are included in this section.

More often than not, the purpose of the studies that follow is to produce semiconducting

particulate thin films for potential device applications, rather than elucidate the nature of

bioorganic-mediated crystallization. Size quantization (limitation) of the semiconducting

particles is often observed in these systems, which may be transferred intact to solid

substrates by Langmuir-Blodgett deposition methods. With two exceptions (Zhao et al.,

1992; Yi et al., 1994) the formation of oriented particles is not observed in these systems,

possibly due to the equilibrium process being "pushed" too rapidly by the addition of the

reacting gas. Perhaps any orienting influence the organic template might have is

overwhelmed here by the rapid growth and large size of the nascent crystal.

Fendler and co-workers have grown particles of semiconducting CdS and ZnS

(Zhao et al., 1990; Zhao; Fendler, 1991a; Zhao; Fendler, 1991b) at floating Langmuir

monolayers using a variety of surfactants with essentially the same result. They observe

the formation of particles at the film-water interface which may be size mediated by

limiting their exposure to the gas. The particles begin as isolated clusters and, over

several hours, grow laterally along the undersurface of the monolayer until they run into

a neighboring particle. Only then do the particles grow downward into the subphase. By

controlling the rate of gas infusion, it is apparently possible to control the morphology

of the particles from a spheroid shape 25-30A in diameter, to a disk shaped 50-100A

across and 25-30A thick. Thicker particles (up to 1800A) are observed to form over

longer time periods, but in each case the whole monolayer-particle arrangement may be

transferred intact onto a solid substrate by Langmuir-Blodgett deposition. The particulate

films as deposited are described as a porous network of disk-shaped particles with the

broad face parallel to the monolayer (Zhao et al., 1990; Zhao; Fendler, 1991a; Zhao;

Fendler, 1991b). The films are well characterized by a variety of techniques and the

absorption spectra (Zhao; Fendler, 1991a) confirm a size-quantization of the particles, a

layer by layer deposition, and a linear increase in intensity as the particles themselves

become thicker. Characterization by scanning probe microscopy (Zhao et al., 1990; Zhao;

Fendler, 1991a) and transmission electron diffraction (Zhao; Fendler, 1991a) indicates that

the particles grow in an unorganized fashion, and no preferential orientation is observed.

In another study, Fendler and co-workers (Zhao et al., 1991) generated CdSe particles in

situ at Langmuir monolayers and observed size-quantized behavior.

H2S exposure of a Pb(II) subphase coated with an arachidic acid Langmuir

monolayer resulted in the formation of oriented PbS crystals at the monolayer solution

interface (Zhao et al., 1992). In the presence of the monolayer, the cubic crystals were

oriented with their [111] direction perpendicular to the surface; in the absence of the

monolayer, there was observed a mixture of (111), (hOO) and (hkO) orientations. The PbS

crystals are observed to be evenly-spaced, well-formed equilateral triangles. In addition,

the fact that coherent electron diffraction was observed over sample areas containing

several particles (selected areas of diffraction or SAD = 2pm in diameter) indicates that

discrete crystals are oriented with respect to one another within the plane of the

monolayer (Zhao et al., 1992). Crystals formed equally well in uncompressed (7t = 0

mN/m) and compressed (nt = 25mN/m) arachidic acid films. However, particles formed

in a circular Langmuir trough exhibited polycrystalline orientation similar to that of

particles formed in the absence of a monolayer; those formed in a rectangular trough

exhibited excellent orientation (Zhao et al., 1992). The authors do not rationalize the

latter observations, but infer that the organization of floating amphiphilic molecules is

dependent upon the shape of each Langmuir trough and the uniformity of compression.

Slow H2S infusion was reported to give better results with regard to particle morphology,

orientation, and size distribution. The results are rationalized in terms of a near perfect

lattice match between the Pb-Pb distance in PbS (4.20A) and the do00 spacing of the

arachidic acid headgroups (4.16A).

Very recently, Fendler and co-workers (Yi et al., 1994) formed silver particulate

films beneath monolayers of various surfactants by slow exposure of the Ag+ subphase

to formaldehyde vapor. Silver particle formation was found to be influenced by silver

concentration, pH, the type of monolayer used, and the presence of ammonia or copper(II)

in the subphase. The morphology of the size-quantized silver particles was described as

fractal in nature and the immature particles in some cases were oriented with their (111)

face parallel to the monolayer (Yi et al., 1994). As the particles grew they became non-

fractal and orientations other than the (111) were observed. The mature films were

described as similar to those of particulate semiconductors discussed previously (Zhao;

Fendler, 1991b) except that transfer by vertical Langmuir-Blodgett deposition was

accomplished only with difficulty. The results were not rationalized in terms of any

complementarity between the organic monolayer and the particles.


We are aware of only one report (Kotov et al., 1993) describing the formation of

metallic particles on the undersurface of a floating monolayer by electrochemical

reduction. Electrochemical silver reduction occurred at the solution-monolayer interface

resulting in a "two-dimensional" silver particulate film. Size-quantization of silver

particles was observed (500-700A) and no silver particles were observed to form under

monolayers of positively charged or bulky headgroups. Transmission electron diffraction

recorded no preferred orientation. Silver reduction at floating monolayer interfaces was

reported to occur at more negative potentials than is required at electrode surfaces in bulk

solution (Kotov et al., 1993) and particulate formation occurred only at the frontier

regions of the advancing two-dimensional silver particulate film beneath the organic


Deposition of pre-formed clusters

An interesting hybrid method in which floating Langmuir monolayers are used to

manipulate inorganic particles is described in this section. Although not strictly a

"templating" technique, it is included here for completeness. Pre-formed inorganic

clusters are incorporated into Langmuir-Blodgett films by two means. First, the pre-

formed clusters are put into the subphase solution and are electrostatically bound to either

the floating monolayer or the hydrophilic side of a deposited monolayer. The substrate

is withdrawn and another particle-surfactant layer assembly is deposited resulting in a

bilayer "sandwich". These techniques have been demonstrated by Fendler and co-workers

to make Langmuir-Blodgett films of size-quantized hexametaphosphate-stabilized CdS

clusters between bilayers of dioctadecyldimethylammonium chloride (Xu et al., 1990) and

unstabilized cationicc" Fe304 particles within arachidic acid Langmuir-Blodgett films

(Zhao et al., 1990). The second technique is to spread pre-formed, organic stabilized

particles onto the surface of the Langmuir-Blodgett trough. The clusters are then treated

as a monolayer--compressed and deposited directly onto the solid substrate without an

accompanying surfactant molecule. It is presumed that the on-trough behavior of such

materials is "ornery" because there is no clear hydrophobic-hydrophilic aspect to the

stabilized particles, which is generally a requirement for a well-behaved Langmuir

monolayer. Nevertheless, Fendler and co-workers (Kotov et al., 1994) report the

deposition of monoparticulate layers of dodecylbenzenesulfonic acid-stabilized size-

quantized CdS clusters from a Langmuir-Blodgett trough without an accompanying

surfactant. Brewster angle microscopy shows that the Langmuir monolayer consists of

isolated islands that aggregate into a close-packed film upon compression. Absorbance

versus number of layers indicates a linear increase in intensity with increasing number of

layers as well. In addition, the position of the absorption edge, which is often used to

determine the size of the so-called "q-state" clusters, does not shift with the increase in

the number of layers. The results are rationalized by the authors (Kotov et al., 1994) to

mean that the individual CdS particles do not agglomerate into larger clusters upon

deposition and drying of the film. This work was repeated with similar results by a

another group (Du et al., 1992) using hexametaphosphate-stabilized CdS colloids in a

mixed behenic acid-octadecylamine LB film.

Deposited Langmuir-Blodgett Films

Katherine Blodgett showed that it is possible to transfer floating Langmuir

monolayers onto solid supports (Blodgett, 1935). The Langmuir-Blodgett (LB) method

is experiencing a revival of sorts due to its potential in device applications. In addition,

the LB method is attractive because it provides a way to have some (slight!) mechanical

control over what molecules do. LB films are good experimental platforms. The

observation of low-dimensional magnetic phenomena (Pomerantz et al., 1978),

measurement of electron mean-free pathlengths (Brundle et al., 1979) and inorganic

extended lattice formation (Byrd et al., 1993) are examples of the experimental variety

possible with Langmuir-Blodgett films. The advent of computer controlled film

deposition equipment (Ulman, 1991; Roberts, 1990) has resulted in better quality films.

The increased accessibility of physical methods such as scanning probe microscopy,

electron microscopy, and FTIR gives us a better understanding of LB films. These

factors will continue to spur interest in LB films and the associated academic and perhaps

industrial applications.

The Langmuir-Blodgett method has been described by this author (Pike, 1992) and

others (Blodgett, 1935; Gaines, 1966; Roberts, 1990; Ulman, 1991). Briefly, if an

appropriate solid support (hereafter substrate) is moved through a floating Langmuir

monolayer that is held at constant surface pressure the floating monolayer can be


transferred intact to the substrate. By repeated insertions and withdrawals, multiple layers

can be built up on the substrate. The most common configuration of the deposited film

is the Y-type (Roberts, 1990) layer in which the polar headgroup layers are in contact

with one another and the hydrophobic alkyl tails are in contact with one another as well.

In this way the Y-type LB film is composed of bilayers; the center of which is where the

polar headgroups face one another. For some amphiphilic molecules, it is possible to

incorporate metal ions into the interstitial region between the polar headgroups by

adjusting the pH of the subphase so that it is above the pKa of the amphiphile's

headgroup but below the pKb of the metal hydroxide. Proper pH is usually sufficient for

metal incorporation into LB films of the carboxylic acids, CH3(CH2)nCOOH. For the

alkyl phosphonic acids, CH3(CH2)nPO3H2, the situation can be more complex (Byrd,

1994). The deposited LB film will then contain the metal as the salt of the amphiphile.

Typically, the amphiphile of choice is a long chain carboxylic acid such as stearic acid

(CH3(CH2)16COOH) or arachidic acid (CH3(CH2)18COOH). Metal ions may also be

inserted into the film by leaving the deposited film of the unionized acid in a solution of

metal ions at the appropriate pH for a few minutes (Leloup et al., 1985). An ion

exchange mechanism will give rise to a film of the metal salt. However, given that the

films are often easier to deposit as the metal salt of the fatty acid than as the fatty acid

alone, the latter method of metal insertion into LB films is not often used.

This section will describe the work in which deposited Langmuir-Blodgett films

have been used as organic templates for the synthesis of inorganic particles and clusters.

Typically the LB films are deposited as the metal-fatty acid salt then subjected to further

reaction by exposure to gas-phase reactants.

Insertion-reduction templating

Barraud and co-workers (Leloup et al., 1985; Ruaudel-Teixier et al., 1986;

Zylberajch et al., 1988) and Toumarie (Belbeoch et al., 1985) have used the insertion

method discussed previously to ion-exchange Ag+ ions into deposited LB films of behenic

acid (CH3(CH2)20COOH). The films were observed by X-ray diffraction to be in the

form of well ordered layers before and after insertion of the Ag+ (Belbeoch et al., 1985).

Infrared spectroscopic analysis shows the films to be in the form of the protonated fatty

acid prior to silver insertion (Leloup et al., 1985). After insertion IR showed the film to

be the fatty acid salt (Leloup et al., 1985). The silver behenate films were then exposed

to hydrazine (N2H4) vapor which reduced the Ag+ to the metal. X-ray diffraction showed

that the film remained as well-ordered layers (Belbeoch et al., 1985) and IR analysis

showed the return of the protonated behenic acid form (Leloup et al., 1985). Electron

diffraction and microscopy showed the presence of small silver particles 300-500A in

diameter in the film. There was not any preferred orientation observed in the silver and

the authors observe that the clusters are not coherent with the matrix of behenic acid

(Belbeoch et al., 1985).

In an interesting communication (Sastry et al., 1992), a Langmuir-Blodgett film

of lead arachidate deposited in the usual way was bombarded by 4.5 keV electrons. The

authors observed in the XPS a shift toward lower binding energy of the Pb(4f)

photoelectron peak and speculate that the Pb2+ in the LB film is being reduced to lead

metal. The authors do not speculate on the fate of the organic template.

Reaction with gaseous chalcogenides

A recent report (Leloup et al., 1992) describes the insertion of copper and

subsequent conversion to Cu2S within an LB film of behenic acid. The conversion is

effected by exposure of the deposited copper-containing film to gaseous H2S. No data

is presented regarding the physical structure of the films, however (Leloup et al., 1992).

In a pioneering experiment, researchers at the University of Texas at Austin

(Smotkin et al., 1988) deposited LB films of cadmium arachidate onto solid substrates and

exposed the film to H2S. The resulting CdS particles were determined, by UV-vis

spectroscopy, to be about 50A in size (Smotkin et al., 1988). The CdS-containing films

were observed to have a linear increase in absorbance (at 380 nm) with increasing number

of layers. Ellipsometric measurements showed the films to be thicker by about 3A per

bilayer after conversion to CdS. Electron microscopy of the converted films showed the

presence of small particles, but electron diffraction did not confirm the CdS structure

(Smotkin et al., 1988). We now suspect that the authors were observing spurious particles

of copper sulfide in the electron microscope which came about possibly due to a side

reaction of the H2S with the copper electron microscope grid. We attempted to reproduce

their results using the same type of materials and confirmed the formation of copper

sulfide particles by electron diffraction. Scanning electron micrographs of the copper

microscope grid showed the formation of large particles on the surface of the grid after

exposure to H2S. However, the observation of size-limited or quantized semiconducting

particles formed in a Langmuir-Blodgett films is evidence in support of the templating

ability of LB films. A more complete study of CdS particles formed in deposited films

was recently reported by Fendler and co-workers (Geddes et al., 1993). In this paper, the

time-dependent growth of the CdS is monitored by a quartz crystal microbalance, UV-vis,

and surface plasmon resonance. In a 20-layer cadmium arachidate film the reaction with

H2S appears to go to completion in about 20 minutes (Geddes et al., 1993).

Unfortunately, the authors do not report the details of the amount of H2S gas required to

effect complete reaction. The size of the CdS particles and the changes in film thickness

are consistent with the results reported by Smotkin et al. (1988).

Another group (Scoberg et al., 1991) describes the synthesis of semiconducting

CdS, CdSe, CdTe, and CdSxSex_ in deposited LB films of polymerizable long-chain

diynoic acids. The cadmium-containing LB films were exposed to H2X (X = S, Se, or

Te) for several minutes in a sealed container. The resulting films were studied by UV-vis

spectroscopy. Size-quantization was observed on the order of 40-50A for all particles and

did not change after the diynoic LB film is polymerized (Scoberg et al., 1991). The

authors reported a reduction in the particle size when dihexadecylphosphate was mixed

into the diynoic acid monolayer prior to deposition.

There are two reports (Zhu et al., 1992; Peng et al., 1992) of size-quantized lead

sulfide particles formed in LB films of stearic acid. The first group (Zhu et al., 1992)

synthesizes the particles in the usual manner and observes the resulting particles by UV-

vis and scanning tunneling microscopy (STM). The STM measurements confirm the 1.5-

2.0 nm PbS particle size derived from UV-vis measurements (Zhu et al., 1992).

Interestingly, the author's interpretation of STM measurements conducted on the films

after the organic portion of the LB film had been pyrolyzed at 1200C was that the

nanoparticles remained discrete and intact (Zhu et al., 1992). However, UV-vis analysis

was not performed after pyrolysis to confirm this result. In the second report (Peng et

al., 1992), the authors report the formation of non-stoichiometric PbS in a stearic acid LB

film. The initial 1 to 1.5 Pb to S ratio was found by XPS to increase over a period of

two months of exposure to air. The authors report that the sulfur is lost as H2S or SO2

and the Pb remains in the form of the stearic acid salt (Peng et al., 1992).

Other Template Systems

To be sure, there are other systems which have been used to template and mediate

the growth of inorganic particles. Size-quantized CdS and PbS clusters have been formed

in zeolites (Herron et al., 1989; Wang; Herron, 1987; Moller et al., 1989). Layered metal

phosphonate hosts (Cao et al., 1991) have templated the formation of oriented, size-

quantized CdSe, CdS, ZnSe, and PbS. In the latter system, particle growth and

orientation occurs cooperatively with the layered host lattice, but the host lattice is

destroyed in the process. In a very interesting letter (Golan et al., 1992), CdSe

nanocrystals were electrodeposited onto { 111} gold surfaces. An epitaxial match between

the gold (111) face and the CdSe (001) face was proposed for the observed exclusive

orientation of the CdSe. Size-quantization has also been observed in the ion-exchange

membrane "Nafion" by cadmium doping then reacting with H2S (Smotkin et al., 1990).

Finally, there are two groups which report the formation of size-quantized particles of PbS


(Mahler, 1988) and CdS (Bianconi et al., 1991; Lin; Bianconi, 1991) in polymeric

matrices. Bianconi (Bianconi et al., 1991; Lin; Bianconi, 1991) observed size-quantized

CdS particles in a dip-coated polyethyleneoxide film after a 7-day immersion in an

organic sulfur solution. Oriented CdS crystals were observed by electron diffraction after

the sulfur solution had been treated with the surfactant sodium bis(2-

ethylhexyl)sulfosuccinate (Bianconi et al., 1991). The authors observed that the size of

the CdS particles could be controlled by adjusting the initial Cd(II) concentration. It was

also observed that the use of different polymer matrices altered the morphology of the

CdS (Bianconi et al., 1991).

Requirements for Oriented Crystallization

In each of the above studies, organized organic assemblies were used in schemes

to prepare inorganic particles. The organic structure serves as either a template or a

reaction vessel of limited size (Heuer et al., 1992). The goal is generally to produce

inorganic particles with limited size, dispersion, or orientation (Stucky; MacDougall,

1990; Henglein, 1989) The approach draws inspiration from naturally occurring examples

of organic-mediated crystallization of inorganic materials (Berman et al., 1988; Dameron

et al., 1989; Addadi; Weiner, 1985).

It is possible, then, to lay out the requirements for templating inorganic particles

by organic structures. First, the organic must provide a surface or a region which

concentrates one of the eventual reactants. This requirement is concomitant with

electrostatic complementarity. In most cases it is the metal ion which is concentrated,

although in the studies by Mann (Rajam et al., 1991) it was suggested that the initial

binding of HC03- was a key step in orienting the calcium carbonate. Second, this initial

"binding region" should have ion-affinities high enough to hold the initial participants in

place for reaction, but low enough to allow the particle-forming reaction to occur. In the

deposited Langmuir-Blodgett films this is accomplished by a mechanism in which the

stronger acid, HnXn-, is able to protonate the metal-carboxylate, (R-COO-)mMm+ forming

the inorganic Mn X t and the carboxylic acid, RCOOH. Third, for size-quantization to

occur, there should be some means of limiting a) the initial total number of potential

reactants in a given region, or b) the spatial area in which the nascent particles have to

grow. The latter limitation is probably less important than the former, as we have seen

how the growing guest can outgrow and destroy the host structure (Cao et al., 1991), even

if the host structure is a covalent compound. However, there is supporting evidence for

(b) in that size-quantization was observed in deposited Langmuir-Blodgett films, which

are probably not the most durable structures. Fourth, for dispersion limitation to occur,

there should be some means of limiting the rate, pathlength, or direction of the diffusing

species within the host template. This requirement is most easily seen in the chemical

capping and passivating of the nanoparticle surfaces which, if left to themselves, would

otherwise agglomerate into bulk structures (Herron et al., 1990). Also, from the literature,

it is conducive to "good" crystallization if the rate of the particle-forming can be slowed

so as not to overwhelm the provisional initial ordering forces of the template; witness the

rate limiting properties of the vesicle walls in the formation of the magnetite particles

(Mann et al., 1986). Fifth, for orientation to occur, the requirement of geometric and

stereochemical lattice matching between the host and the guest should be met, but it is

perhaps more clear in some cases (Pike et al., 1993) than in others (Bianconi et al., 1991)

exactly how, and to what degree, this requirement is satisfied.




An objective of this work is to develop methods for preparing single layers of

inorganic extended lattice solids (Byrd et al., 1994a; Byrd, Pike et al., 1993; Byrd, Pike

et al., 1994a; Byrd, Pike et al., 1994b; Byrd et al., 1994b; Pike, Byrd et al., 1993; Pike,

Byrd et al., 1994). Such systems should allow for investigation of materials properties

such as magnetism, conductivity, semiconductivity or even superconductivity in the two-

dimensional limit of a monolayer. The approach uses layered organic assemblies, such

as LB films or self-assembled monolayers, as templates for forming inorganic monolayers

in analogy to how the organic assemblies described Chapter 1 have been used to form

inorganic particles. For example, in a preparation that uses an LB film as a component

of a mixed organic-inorganic layered solid (Day, 1985), we have reported monolayer and

multilayer film analogues of the layered transition-metal phosphonates (Byrd, Pike et al.,

1994a; Byrd, Pike et al., 1994b; Byrd, Pike et al., 1994c; Byrd, Pike et al., 1993; Byrd

et al., 1994a; Byrd et al., 1994b). A different approach is to use the organic assembly to

prearrange the metal ions into a layered array for further reaction that produces an

extended inorganic lattice inside the organic matrix (Pike, Byrd et al., 1993; Pike, Byrd

et al., 1994). The idea is that the layered organic template should limit diffusion of the

metal ions to individual planes of the organic matrix. If the inorganic solid has a layered

structure, then the lattice energy may favor formation of single layers of the inorganic


Cadmium Dihalide Structure

The initial attempt (Figure 2-1) at this approach targets the cadmium dihalides,

CdX2 (Pike, Byrd et al., 1993). The "CdI2" structure (Wells, 1984) is described by a

hexagonal-close-packing of the anions in which every other layer of octahedral holes is

occupied by a metal ion. The CdCl2 and CdBr2 structures are analogous to that of CdI2

except for a cubic-close-packing of the chloride and bromide ions (Wells, 1984). The

structure of a single layer, however, is the same in both structure types. In these layered

structures, the bonding within the anion-metal-anion layer is ionic-covalent, while the

anion-anion interaction between layers is van der Waals in nature and easily cleaved

(Wells, 1984).

Although single layers have not yet been identified, this chapter will present

electron diffraction studies that show the template-formed-CdI2 is oriented exclusively

with the CdI2 layers parallel to the LB layers (CdI2 [001] axis normal to the LB layer).

In addition, for domain sizes up to several pm2, arrays of discrete particles are aligned

with respect to one another about the [001] axis. The results suggest that the organic

matrix is responsible for both orienting the CdI2 particles and favoring crystal growth

Cdr Cd+ Cd"_ -1
0-00-00-0 OOHO OHOOH

X=l, CI, Br

Figure 2-1: Cartoon representation depicting reaction of Langmuir-Blodgett film
with gaseous HX.

parallel to the LB planes. In this chapter we will present data on the LB-formed CdBr2

and CdC12 particles formed by the same method. As it happens, the situation is a little

more complex with these two systems. As with the formation of CdI2 particles, CdBr2

and CdCl2 particles are produced with preferred orientations in the organic template,

although two orientations of both the bromide and chloride are observed. The systems

are characterized by attenuated total reflectance FT-infrared spectroscopy (ATR-FTIR),

X-ray photoelectron spectroscopy (XPS), transmission electron microscopy (TEM), and

transmission electron diffraction (TED).

Experimental Details


Octadecyltrichlorosilane (OTS, C18H37SiC13, 95%) was obtained from Aldrich

Chemical Co. (Milwaukee, WI) and stored under N2 until use. Octadecanoic acid stearicc

acid, C17H35COOH, 99.5%) was used as received from EM Science (Cherry Hill, NJ).

Eicosanoic acid (arachidic acid, C19H39COOH, 99%) and cadmium chloride

hemipentahydrate (CdC12 2 V2 H20, 98%) were purchased from Aldrich and used

without further purification. Hydrogen bromide, 99.8%, technical hydrogen chloride,

99.0% and hydrogen iodide, 98.% were obtained from Matheson Gas Products (Atlanta,

GA). Spectrograde chloroform, hexadecane and hexane were used as received. 300-mesh

titanium electron microscope grids were purchased from the Ted Pella Company

(Redding, CA). Formvar resin was purchased from Ted Pella and dissolved in

spectrograde 1, 2-dichloroethane before use. Silicon monoxide (325 mesh powder) was

purchased from Aldrich. The water used as subphase in the LB experiment was purified

via a Sybron/Barnstead Nanopure (Boston, MA) system immediately before use and had

an average resistivity of 18 MQ-cm.

Substrate Preparation

For all XPS measurements, single crystal (100) 20 x 15 x 1 mm3 n- and p-type

silicon wafers were purchased from Semiconductor Processing Company (Boston, MA).

For the infrared experiments, silicon ATR crystals, 450 50 x 10 x 3 mm3, were purchased

from Wilmad Glass Company (Buena, NJ). The silicon XPS and IR substrates were

cleaned using the RCA procedure (Kern, 1990) and dried under N2. OTS was self-

assembled (Netzer; Sagiv, 1983; Maoz; Sagiv, 1984) onto the silicon substrates by placing

the clean substrates in a 2% solution of OTS in hexadecane for 30 minutes. Substrates

were then rinsed in a chloroform Soxhlett extractor for 30 minutes. Upon removal from

the Soxhlett, the substrates were hydrophobic. TEM substrates were electron microscope

grids sandwiched between a 600 A thickness of Formvar film and a 75 x 25 x 1 mm3

glass microscope slide. A home-built electron-beam evaporator system was used to

deposit 300 A of SiO onto the Formvar layer at a rate of 1-2 A/s as monitored by a

quartz crystal microbalance. Pressure within the evaporator chamber was maintained at

4 x 10-8 atm. The basic TEM substrate preparation procedure is described in the

literature (Fischer; Sackmann, 1986). The entire slide was then made hydrophobic with

OTS as described above except that hexane was substituted for chloroform in the rinsing



Langmuir-Blodgett trough

The LB films were deposited using a KSV Instruments (Stratford, CT) Model

5000 LB system with a PTFE trough and hydrophobic barriers (Pike, 1992). The surface

pressure was measured with a platinum Wilhelmy plate suspended from a KSV


Transmission electron microscope

TEM and TED analyses were performed on a JEOL (Peabody, MA) JEM 200CX

electron microscope. In all cases, the electron beam was normal to the sample LB basal

plane. Beam exposure was purposefully kept to a minimum to avoid sample degradation.

Exposure times were typically less than 5 s for images and less than 11 s for diffraction

patterns. The instrument was operated at 80, 100 and 200 kV as necessary and the

electron beam was collimated with a 200-pm-diameter (hereafter pm-d.) condenser

aperture. In some cases, focussing and other conditions were set in areas immediately

adjacent to the region of interest, which was then shifted to the center of the image plane

for photographic recording.

FT-infrared spectrometer

Infrared spectra were recorded with a Mattson Instruments (Madison, WI)

Research Series-1 FTIR spectrometer using a narrow-band mercury cadmium telluride

detector. A Harrick (Ossining, NY) TMP stage was used for the ATR experiments. All

spectra consisted of 1000 scans at 2.0 cm-' resolution and were referenced to the clean

OTS-coated Si ATR crystal.

X-ray photoelectron spectrometer

X-ray photoelectron spectra were obtained using a Perkin-Elmer (Eden Prairie,

MN) PHI 5000 Series spectrometer. All spectra were taken using the Mg Ka line source

at 1253.6 eV. The spectrometer has a typical resolution of about 1.7 eV, with anode

voltage and power settings of 15 kV and 300 W, respectively. Typical operating pressure

was less than 5 x 10-9 atm. Survey scans were performed at a 700 take-off angle with

respect to the sample surface parallel using a pass energy of 89.45 eV. Multiplex scans,

20 scans over each peak, were run over a 45-50 eV range with a pass energy of 17.90 eV.

Sample exposure to the X-ray beam and detector dead time were minimized to avoid

unnecessary sample degradation.


Langmuir-Blodgett films were prepared by spreading 100 pL of a 4.5 mM solution

of stearic or arachidic acid in chloroform onto the subphase and allowing 15 minutes for

the solvent to evaporate prior to compression. The subphase temperature was maintained

at 16 + 1 C. For stearic acid and arachidic acid films, a pure water subphase was used

(pH 5.2). Stearic acid and arachidic acid Langmuir monolayers were linearly

compressed at a rate of 10 mN/(m min) to a surface pressure of 24 mN/m and deposited

at rates of 25-50 mm/min. Transfer ratios were unity. Cadmium arachidate LB films

were deposited from a 4 mM Cd2+ subphase that was pH-adjusted upward with KOH to

pH 6.2 0.1. The subphase was allowed to equilibrate for several hours prior to

spreading while the pH was constant. The cadmium arachidate Langmuir monolayer was

compressed linearly at a rate of 10 mN/(m min) to a surface pressure of 30 mN/m and

deposited at rates of 25-50 mm/min with 10 min between successive dips for drying.

Transfer ratios were unity.

Metal-halide-containing films were produced by exposing the LB films to the

gaseous hydrohalic acids. The sample was placed within a glass test tube fitted with a

sidearm stopcock valve and sealed with a rubber septum. The 80 mL tube containing the

sample was evacuated and backfilled with dry Ar several times before 5 mL of gaseous

HX was introduced through the rubber septum via gastight syringe. After a period of

time the tube was purged with a positive pressure of dry Ar. Typical reaction conditions

were 6.4 x 10-2 atm partial pressure of HX in argon for 5-15 minutes. Great care was

taken throughout the procedure to avoid sample contamination.


Attenuated Total Reflectance FT-Infrared Spectroscopy

Reactions of the cadmium arachidate (CdArac) LB films with hydrogen halide

(HX) gas were monitored by ATR-FTIR. Figure 2-2 shows the ATR-FTIR spectra from

1500-3500 cm-1 of 10-bilayer CdArac LB films before and after exposure to HCI, HBr,

and HI. Peaks at 2953 cm-1, 2917 cm-1 and 2849 cm-1 in the CdArac spectrum (Figure

2-2a) are the asymmetric methyl Va(CH3), asymmetric methylene Va(CH2) and symmetric

methylene vs(CH2) stretches, respectively, of the close-packed, all-trans hydrocarbon

chains in the LB film (Wood et al., 1989; Porter et al., 1987). The band at 1545 cm-1

is the asymmetric carboxylate stretch, Va(COO-), characteristic of the metal carboxylate

salt. The small peak at 1681 cm-1 is a carbonyl stretching band, v(C=0), and appears

as a result of incomplete salt formation in the LB film. At a pH of 6.2, used to deposit

the CdArac films, we estimate 90-95% salt formation (Ahn; Franses, 1991; Peltonen et

al., 1994; Schwartz et al., 1993). Upon exposure to HX (Figure 2-2b-d) the Va(COO-)

band completely disappears, and in all three cases a new band at 1701 cm-1 (v(C=0)) is

observed resulting from protonation of the amphiphile and formation of the carboxylic

acid. There is no change in position, intensity, or fwhm of the methyl and methylene

stretching bands indicating that the organic layers do not undergo any gross changes in

structure during the reaction, and that there is no loss of organic material under the

reaction conditions used. In a series of control experiments, 10-bilayer films of metal-free

arachidic acid were exposed to each of the hydrogen halides and no changes in the ATR-

FTIR spectra were observed. In the arachidic acid and HX-exposed CdArac LB films,

a broad absorbance between 3300 and 2500 cm-1 is observed which can be attributed to

the acid O-H stretch. We do not observe any lattice or metal-coordinated water in the

HX-exposed CdArac LB films (Nakamoto, 1963).

X-Ray Photoelectron Spectroscopy

Control experiments were run on stearic acid LB films to determine whether the

gaseous HX is incorporated into either the non-metal-containing LB film, the OTS layer,

or the silicon substrate. Stearic acid LB films were exposed to each of the HX gases then

placed in the XPS for study. XPS multiplex spectra over the appropriate halogen regions

show that no halides are present in the non metal-containing LB films.

v CH2
a 2

v CH
s 2

v CH
a 3

(a) 20 Layer CdArac LB Film

=0.5 a.u.

(b) After Rxn with HI Gas


(c) After Rxn with HBr Gas

(d) After Rxn with HCI Gas




Figure 2-2:

ATR-FTIR spectra of 10-bilayer LB films of cadmium arachidate: (a)
before reaction; and after 5-min exposure to (b) HI; (c) HBr; (d) HCI. In
spectra b-d, the disappearance of the VaCOO- band indicates that the
reaction goes to completion.

v COO-

2800 2400 2000
Wavenumbers (cmf1)





Figure 2-3 presents XPS multiplex spectral data for 10-bilayer CdArac LB films

after reaction with HCI, HBr and HI. In each case, XPS reveals that the halogen is

incorporated into the metal-containing film. The binding-energy peak areas are then

integrated and corrected with an atomic sensitivity factor (5000 Series ESCA Systems

Version 2.0 Instruction Manual, 1989; Wagner et al., 1981) to yield the observed relative

concentrations for each element of interest (Table 2-1). The observed relative

concentrations presented in Table 2-1 do not change over a wide range of HX-exposure


When using XPS to quantify elemental percentages in heterogenous samples, the

effects of photoelectron attenuation by an overlayer must be considered (Seah; Dench,

1979). The inelastic mean free path, Xm, of a photoelectron depends upon the

photoelectron's kinetic energy and upon the material through which the escaping

photoelectron travels. The XPS signal intensity, I, for any given element in the sample

is reduced to the extent that that element's photoelectrons are attenuated by an overlayer.

The relation is given by the attenuation equation (Briggs; Seah, 1990):

I = (I")exp(-dm/(XmsinO)) (2-1)

where I" is the peak area normalization or sensitivity factor (Wagner et al., 1981), dm

is the overlayer thickness of material m, Xm is the inelastic mean free path of the

photoelectron through material m, and 0 is the take-off angle with respect to the surface



415 405 395



4I I 3 I
415 405 395


630 620 610


205 195 185


415 405 395 75 65 55
Binding Energy (eV)

XPS spectra of 10-bilayer LB films of cadmium arachidate after 5 min
exposure to: (a) HI; (b) HCI; and (c) HBr. The observed relative M:X
concentrations calculated in the text are (a) 32:68; (b) 45:55; and (c) 40:60.

Figure 2-3:

The relative concentration for element A, CA, is then given by:

Cexp(- dmA
CA= Am, sm (2-2)
Ap d 'a d
Eexp(- mA ) + exp(- ) ..
A 'ms sin B m, sinm

where the denominator is summed over all the elements of interest (Holloway; Bussing,

1992). To model the cadmium halide-LB assembly, we assume a Y-type (Roberts, 1990)

matrix of tail-to-tail arachidic acid bilayers separated by CdX2 monolayers where the

topmost layer is an arachidic acid monolayer. For organic overlayers, values of Xm are

generally longer than those of inorganic overlayers (Seah; Dench, 1979) and estimates in

the literature vary by about an order of magnitude (Seah; Dench, 1979; Akhter et al.,

1989; Laibinis et al., 1991; Sastry et al., 1991; Ohnishi et al., 1978). For Xm > dm

however, the predicted relative concentrations for the Cd3d, Cl2p, Br3d, and I3d

photoelectron peaks are nearly insensitive to changes in number of bilayers and changes

in Xm and vary by only about 1% over the range of Am values in the literature (Seah;

Dench, 1979; Akhter et al., 1989; Laibinis et al., 1991; Sastry et al., 1991; Ohnishi et al.,

1978). Using the model described above and dm = 27.6 A for the thickness of a CdArac

monolayer (Brundle et al., 1979; Tippmann-Krayer et al., 1992; Jark et al., 1989), relative

concentrations of 35.0% for Cd and 65.0% for I are predicted for a 1:2 Cd:I system in

Table 2-1:

Summary of XPS multiplex data for 10 bilayer films of cadmium
arachidate after exposure to HX.

LB Film/HX Peak KEa (eV) Observed Predictedb
Relative Relative
Concentration Concentration
(%) (%)
CdArac/HI Cd3d 840 32 3c 35

_3d 624 68 65

CdArac/HBr Cd3d 839 40 2d 32

Br3d 1049 60 68

CdArac/HCl Cd3d 839 45 4e 31

C12p 1179 55 69

aKinetic energies are not corrected for peak-shifting due to sample charging effects.
bCalculated for MX2 stoichiometry from Equation 2-2 in text using XCd = 319 A; XI=275
A; XBr= 356 A; XCI = 378 A. c95% confidence limits N=5. d95% confidence limits N=9.
e95% confidence limits N=5.

the LB matrix. For Cd-I, the predicted relative concentrations compare well with the

observed relative concentrations of 32% Cd to 68% I for the HI-treated samples. For the

CdArac-HBr system the observed relative concentrations, 40% Cd and 60% Br, are

exactly centered between the model's predictions for 1:2 and 1:1 Cd:Br stoichiometries.

For the Cd:C1 systems, the observed relative concentrations are 45:55, while the model

predicts relative concentration values of 32:68 and 49:51 for the 1:2 and 1:1 ratios,

respectively. The XPS data are tabulated in Table 2-1 and are compared to the predicted

relative concentrations calculated using the layered model with a CdX2 stoichiometry.

Transmission Electron Microscopy

Control experiments were performed on LB films of arachidic acid to verify that

the hydrogen halides do not react with either the un-ionized carboxylic acid, SiO,

Formvar or the Ti support grid. TEM images of multilayer arachidic acid LB films after

HX exposure were featureless, and we have been able to record only diffuse scattering

in electron diffraction measurements from the non metal-containing LB films. Titanium

support grids were used because copper grids react with the hydrogen halides. In another

control experiment, a film of bulk CdArac was cast from hexane onto an electron

microscope grid and then exposed to HI. Prior to reaction, the cast film appears in Figure

2-4 as a random assortment of needles in the TEM, and only diffuse scattering is

observed in diffraction (Figure 2-5). After exposure of the cast CdArac film to HI,

distinct particles are formed and appear in the micrograph (Figure 2-6) as darker regions

Figure 2-4:

Figure 2-5:

TEM image of a cast cadmium arachidate film prior to reaction with HI
(30K magnification, 200kV).

TED pattern of 2.6 pm-d. area in Figure 2-4.





Figure 2-6:

TEM image of a cast cadmium arachidate film after exposure to HI (30K
magnification, 200 kV).

Figure 2-7: TED pattern of 2.6 pm-d. area in Figure 2-6. Diffraction originates from
a polycrystalline array of CdI2.



against the lighter organic material. The particles have an average size of about 0.1 pm.

Electron diffraction from a 2.6-pm-d. area of the HI-exposed cast film is shown in Figure

2-7. The diffraction pattern can be assigned to the CdI2 structure, arising from a sample

of randomly oriented particles.

As seen in Figures 2-8 and 2-9, CdArac LB films are featureless in the TEM, but

yield crystalline diffraction in TED with the expected in-plane hexagonal symmetry and

an observed d100 spacing of 4.26 0.06 A consistent with the literature values (Reigler,

1989; Inoue et al., 1989; Garoff et al., 1986). After exposing the CdArac LB film to HI,

bright field TEM (Figure 2-10, 2-12) shows the appearance of particles. The observed

particles range in size from 100 nm up to several pm across with an average size of about

0.4 pm. Some of the smaller particles have a hexagonal shape, while the larger

crystallites lack well defined edges. TED from the HI-exposed CdArac LB film is shown

in Figures 2-11 and 2-13. During the diffraction experiment, Bragg spots from both the

organic lattice and inorganic lattice are seen. The organic lattice diffraction pattern fades

away within several seconds, while that of the inorganic lattice does not change position

or orientation upon extended exposure to the electron beam (Pike, Byrd et al., 1993).

TED from the large, continuous particle (2 0.7pm) in Figure 2-10 is shown in Figure 2-11

and yields a single-crystal diffraction pattern corresponding to the CdI2 [001] zone axis.

Bragg spots corresponding to d-spacings of 3.63 0.02 A, 2.12 0.02 A, and 1.38

0.02 A can be assigned to the {100}, 110}, and {120} reflections of CdI2, and several

orders of each reflection are seen (Pinsker, 1941; Wyckoff, 1982; Finch; Wilman, 1937).

Table 2-2 indexes the diffraction spots for the template-formed particles by comparing

Figure 2-8:

Figure 2-9:

TEM image of a 10-bilayer LB film of cadmium arachidate (10K
magnification, 200kV).

TED pattern in Figure 2-8 reveals the expected in-plane hexagonal
symmetry of cadmium arachidate. Diffraction spots correspond to (a)
{100}; (b) {110} reflections and a yield a d100 spacing of 4.26 0.06 A.

Figure 2-10: TEM image of a single CdI2 particle > 0.7 mrn in size formed in a 10-
bilayer cadmium arachidate LB film after reaction with HI (30K
magnification, 200kV).

Figure 2-11:

200 kV TED pattern from Figure 2-10. The pattern corresponds to the
[001] zone axis of CdI2. Diffracted spots can be assigned to the (a) { 100};
(b) {110}; (c) {120}; and (d) {140} reflections of CdI2.


them with the bulk CdI2 d-spacings (Pinsker, 1941; Wyckoff, 1982; Finch; Wilman,

1937). The TED pattern in Figure 2-13 arises from a 2.6-pm d. area in Figure 2-12, and

dark field TEM of the same region shows that several discrete crystals contribute to this

diffraction pattern. Like the single particle in Figure 2-10, the observed diffraction from

a group of several particles corresponds to the CdI2 [001] zone axis. The diffraction

spots have an angular variance of 40 indicating that each particle has nearly the same

in-plane orientation about the [001] zone axis.

Figures 2-14 through 2-17 present typical TEM images and a TED pattern from

particles resulting from the reaction of a 10-bilayer CdArac LB film and HBr. The

particles can be grouped into two size categories: (1) large particles several hundred

nanometers to one micrometer across, and (2) small particles tens of nanometers in size.

Figure 2-14 is an electron micrograph of several large particles which have been

identified as CdBr2 from electron diffraction (Table 2-2). Like the CdI2 particles,

diffraction from the large CdBr2 particles, shown in Figure 2-15, gives rise to several

orders of {hkO} reflections in a hexagonal pattern which originates from the [001] zone

axis. At higher magnification, areas are observed where two types of smaller particles

are grouped together in a relatively homogenous distribution (Figure 2-16). In these areas

there are "needles", generally 10-50 nm long with an aspect ratio of about 10, and

"round" particles 10-50 nm in diameter. Throughout the area shown in Figure 2-16, the

needle-like particles are oriented at about 1200 relative to one another. The corresponding

TED pattern is shown in Figure 2-17 and presents evidence for two different orientations

of the same crystallographic phase.

Table 2-2:

Summary of transmission electron diffraction data for 10-bilayer cadmium
arachidate Langmuir-Blodgett films after exposure to HX.

hkl CdI2 CdI2 CdBr2 CdBr CdC12 CdC12
d(Obs) d(Lit.)a d(Obs) d(Lit.)' d(Obs) d(Lit.)c
(A) (A) (A) (A) (A) (A)
100 3.63 3.69 3.4 3.42 3.31 3.33
110 2.12 2.12 1.99 1.98 1.93 1.93
120 1.38 1.38 1.29 1.29 1.26 1.26
200 1.8 1.83 1.67 1.71 1.66 1.67
220 1.05 1.06 0.99 0.99 0.96 0.96
130 1.01 1.02 0.94 0.95
300 1.23 1.22 1.16 1.14 1.11 1.11
003 6.3 6.22 5.84 5.82
006 3.1 3.11 2.92 2.91
009 1.95 1.94
0012 1.59 1.56 1.46 1.45

aPinsker, 1941; Finch; Wilman, 1937 (C3m) bPinsker,
1941 (RIm).

1942 (RIm) Cpinsker; Tatarinova,

Figure 2-12:

Figure 2-13:

TEM image of a several CdI2 particles formed in a 10-bilayer cadmium
arachidate LB film after reaction with HI (10K magnification, 200kV).

200 kV TED pattern from Figure 2-12. The pattern corresponds to a 2.6
pm-d. area in Figure 2-12. Diffracted spots can be assigned to the (a)
{100}; (b) {110}; and (c) {120) reflections of CdI2.

. 1,1:Alil

.5 ail~



. I..

Figure 2-14:

TEM image of a typical "large" CdBr2 particle formed in a 10-bilayer
cadmium arachidate LB film after reaction with HBr (30K magnification,
200 kV).

Figure 2-15: TED pattern corresponding to a 0.7 pm-d. area in figure 2-14. Diffracted
spots can be assigned to the (a) (1001; (b) {110}; (c) (120}; (d) {130);
and (e) {140} reflections of CdBr2.


amp *A

^.y k.'.

Figure 2-16: TEM image of "smaller" CdBr2 particles (50K magnification, 200 kV).
Note the 1200 relative rotation of the needles.

Figure 2-17:

200 kV TED pattern of a 0.7 pm-d. area in Figure 2-16. Diffracted spots
can be assigned to the (a) {003}; (b) {100); (c) {006); (d) {110}; (e)
{200}; (f) {0012}; and (g) {120} reflections of CdBr2.

* S. *
-I -

-*ai '

' r -


r A

*. -t S .


Diffraction spots correspond to either {hk0} or {001) CdBr2 reflections, and several orders

of each type of reflection are visible. The {hk0} reflections have an angular variance of

10' and originate from crystals that have the [001] axis perpendicular to the LB plane.

The limited angular variance indicates that the crystals giving rise to the {hk0} reflections

are oriented relative to one another within the ab plane. The {001} reflections are

produced by the CdBr2 particles that are oriented with the [001] zone axis parallel to the

LB plane.

TEM and TED results of HCl-exposed CdArac LB films are shown in Figures 2-

18 through 2-21. Large particles (Figure 2-18) ranging from 100 to 600 nm in size are

mixed with regions of homogeneously distributed smaller particles (Figure 2-20) that

range from 10 to 50 nm in size. Some of the larger particles are textured and appear to

be aggregates of small particles. TED patterns can be indexed to the CdCl2 structure and,

like the CdBr2 samples, two discrete sets of reflections are observed. Electron diffraction

from some of the isolated larger particles gives rise to only {hk0} reflections and is

shown in Figure 2-19, while others give rise to a mixture of {hkO} or {001} reflections.

Electron diffraction patterns from large areas (2 2.6-gm d.) of evenly distributed smaller

particles (Figure 2-21) gives rise to strong {100) and {110) reflections as well as

reflections from the {001} planes. The {110} Bragg spots have an angular rotation of

120. The complete indexing for the CdC12 TED data is listed in Table 2-2 (Pinsker;

Tatarinova, 1941; Wyckoff, 1982).

Figure 2-18:

TEM image of typical "large" CdCl2 particle formed in a 10-bilayer
cadmium arachidate LB film after exposure to HCI (30K magnification,
100 kV).

Figure 2-19: TED pattern of Figure 2-18 can be assigned to the CdCI2 structure.
Diffraction spots correspond to the (a) {100}; (b) {110}; (c) (120}; and
(e) { 130) reflections of CdC12.


t .


Figure 2-20: TEM image of "smaller" CdCl2 particles formed in a 10-bilayer Langmuir-
Blodgett film after exposure to HCI (30K magnification, 100 kV).

Figure 2-21: TED pattern of a 2.6 pm-d. area in Figure 2-20. Diffracted spots can be
assigned to the (a) (003); (b) { 100}; (c) {006); and (d) {110} reflections
of CdCl2.

I G,
A)~ "

qI l ,:'



'p. 1"
~~ U

''*.I I

d r


Organic Template Reaction

ATR-FTIR and XPS results confirm that the reaction of CdArac LB films with

hydrogen halide gases goes to completion. The carboxylate is re-protonated in the

reaction and the halide is incorporated into the film forming the cadmium dihalide. The

C-H stretching bands are unchanged after reaction suggesting that the organic lattice is

intact with the hydrocarbon chains remaining crystalline and in an all-trans conformation.

The reaction does not require high pressures of hydrogen halide and appears to be

complete within a few seconds. All of the inorganic particles that are formed can be

identified by electron diffraction as the corresponding cadmium dihalide, CdX2. No other

inorganic products are observed in the films by XPS and TED measurements. In addition,

the XPS data coupled with the TED results indicate that all of the halogen in the films

is in the form of the metal halide. When an un-ionized film of arachidic acid is reacted

with HX, no halogen is observed in the film by XPS measurements. Halogens are only

seen when the metal ion is present, and in each system, the cadmium to halogen ratio is

independent of sample history (Table 2-1). For the CdI2 system, the observed Cd:I ratio,

determined by XPS, is exactly the ratio expected for the CdI2 stoichiometry. In the

CdBr2 and CdCl2 systems, the observed Cd:halogen ratios are greater than that predicted

by the layered model (Equation 2-2). A possible reason for the difference is oxide

formation at the surface of the particles (Peng et al., 1992) which would increase the ratio

of metal to halide observed in the XPS. Literature values of the Cd(3d5/2) photoelectron

binding energies in CdI2 and CdCl2 show a 0.7 eV difference (Handbook of X-Ray

Photoelectron Spectroscopy, 1989). Relative to the average CdX2 Cd(3d5/2) binding

energy, there is a 0.6 eV difference for the CdO and a 0.8 eV difference for Cd(OH)2

(Handbook of X-Ray Photoelectron Spectroscopy, 1989). If oxide formation is a factor

in reducing the observed metal to halide ratio it is possible that the shift would not be

resolved by our instrument. Oxidation is more likely to occur at crystal edges, and as the

average crystal size is reduced in the CdBr2 and CdCl2 samples, edge effects at crystal

boundaries will have a greater effect on the observed relative elemental concentrations in

these systems than in the CdI2 samples. The LB films may also suffer damage due to

photoelectrons generated in the XPS experiment (Graham et al., 1993). We have

observed higher rates of carbon loss for films containing the protonated fatty acid than

for films containing the metal-carboxylate salt (Pike, 1992). Evidence for the desorption

of whole molecules from protonated fatty acid LB films in the XPS has been reported by

Kobayashi et al. (Kobayashi et al., 1988). Loss of organic molecules in the LB film

would tend to invalidate the layered model used to account for the effects of the

photoelectron escape depth. Similarly, TEM images show that discrete particles of the

cadmium dihalides are formed, indicating that the film geometry deviates further from the

alternating layered matrix used to predict the relative elemental concentrations. The most

important observation from the XPS data is that the cadmium-to-halogen ratios do not

vary from experiment to experiment (Figure 2-3, Table 2-1), and are independent of

reaction times and conditions. The fixed metal-to-halogen ratio for each system is

consistent with the formation of only one inorganic species and with all of the halogen

existing as the corresponding metal halide identified by TED.

Particle Orientation

It is well known that CdArac LB films give rise to electron diffraction from

domains of crystalline order up to several pm2 in area (Reigler, 1989; Garoff et al., 1986;

Inoue et al., 1989). In contrast to the solvent-cast films, exposure of CdArac LB films

to HI, HBr or HCI results in cadmium halide particles that exhibit preferential orientations

relative to the substrate plane and to other particles. In the case of the CdI2 system we

find that, throughout the film, the particles are oriented exclusively with the [001] axis

normal to the LB film basal plane. In other words, the metal halide layers are always

parallel to the LB layers. By selecting an area of diffraction that includes many particles

(Figure 2-13), we observe single-crystal-like diffraction rather than polycrystalline-type

rings. Not only do the particle basal planes share a common orientation, but domains

exist as large as 2.6 pm-d. where the corresponding in-plane crystal axes have a common


To account for the observed orientation in this experiment, we can consider two

mechanisms. First, the CdArac precursor film organizes the cadmium ions into an

arrangement similar to that found in the cadmium halides, and thereby directs the growth

of the inorganic lattice by facilitating diffusion parallel to the LB basal plane and limiting

diffusion between planes. Such a mechanism could account for the cadmium halide

particles growing preferentially with the metal halide planes parallel to the LB plane.

However, the observation of domains in which discrete cadmium halide particles have the

same in-plane orientation suggests that some degree of lattice matching exists between

the LB template and the cadmium halide, and that this lattice match is responsible for

directing the growth of particles. This mechanism is analogous to how oriented particles

are thought to grow at solution-monolayer interfaces (Mann et al., 1988; Heywood; Mann,

1992a; Heywood; Mann, 1992b; Rajam et al., 1991; Landau et al., 1985; Weissbuch et

al., 1988; Landau et al., 1989; Landau et al., 1986; Popovitz-Biro et al., 1991; Zhao et

al., 1992; Mann et al., 1990). As discussed in Chapter 1, directionally-mediated

crystallization is thought to occur when there is geometric (Heywood; Mann, 1992a;

Heywood; Mann, 1992b; Zhao et al., 1992), stereochemical (Mann et al., 1988; Landau

et al., 1985; Weissbuch et al., 1988; Landau et al., 1989; Mann et al., 1990), or

electrostatic (Rajam et al., 1991; Landau et al., 1986; Mann et al., 1988) complementarity

between the floating monolayer and the crystalline solid. At this point, we can only

speculate about a potential lattice match between the CdI2 particles and the LB film.

Using a shorthand based on the Wood (Wood, 1964) notation, a (43 /2 x 3 /2)300

relation is found for the (001) surfaces of CdI2 and the CdArac precursor, where 2aCd12

= 43 aCdArac (a is the unit cell of the parent lattice; aCdI2 = 4.24 A and aCdArac = 4.92

A) (Wyckoff, 1982; Reigler, 1989; Garoff et al., 1986; Inoue et al., 1989; Pinsker, 1941).

The superposition of these two surface nets is shown in Figure 2-22 and is very nearly

commensurate, having only a 0.5% lattice mismatch where the lattice mismatch is defined

as (43-aCdArac 2aCdI2 / 43.aCdArac) x 100. Of course, after reaction with HI, the

organic layer is no longer present as the cadmium salt and there may be some

reorganization of the organic assembly to conform to the CdI2 (001) surface. This

reorganization would have to be cooperative and over a fairly long-range to result in the

observed orientation of discrete particles over a range of several uim. Calculating the

same (43 /2 x 43 /2)30 lattice match between an organized arachidic acid layer and the

(001) face of CdI2 results in a 0.9-2% mismatch, depending on the values of aarachidic acid

used (Zhao et al., 1992).

In the CdBr2-containing films a larger range of particle shapes and sizes is

observed. The large CdBr2 particles shown in Figure 2-14 give rise to many orders of

{hk0} reflections indicating that these particles are oriented with their [001] axis

perpendicular to the LB plane. Electron diffraction from a collection of many discrete

particles confirms that, within the LB plane, they are oriented with respect to each other

as well. As in the CdI2 case, there may be a lattice match between CdBr2 and the

organic lattice that accounts for the preferential orientation of the CdBr2 particles. The

same (43 /2 x q3 /2)300 relationship suggested for the CdI2 system results in a > 7%

mismatch in the case of CdBr2. A potentially closer match is a (4/5 x 4/5)00 relationship

between the (001) face of CdBr2 and the basal plane of the precursor CdArac which

would result in only a 0.4% mismatch, where SaCdBr2 = 4aCdArac (aCdBr2 = 3.95 A and

aCdArac = 4.92 A) (Reigler, 1989; Garoff et al., 1986; Inoue et al., 1989; Wyckoff, 1982;

Pinsker, 1942). Electron diffraction from regions containing the smaller "needle-like" and

"round" particles (Figure 2-17) gives rise to two sets of reflections that can be indexed

as the {hk0) set and the {001} set of CdBr2 reflections.


0.426 nm
d c= 0.426 nm

Figure 2-22:

Line representation of the proposed (43 /2 x 43 /2)300 relation of the
(001) surfaces of CdI2 and cadmium arachidate, drawn to scale. The d-
spacings correspond to the distances between (100) planes and there is a
30 angle between the two lattice nets.

In a single-crystal pattern, these reflections cannot be produced simultaneously because

their respective zone axes are normal to one another (Edington, 1976). The observed

diffraction pattern then, is produced from two sets of CdBr2 crystals with distinct

orientations. One set is oriented with the [001] direction perpendicular to the LB plane,

producing the {hk0} reflections, while the other set of crystals is oriented with the [001]

direction parallel to the LB plane, producing the {0011 reflections. Attempts to isolate

each set of crystals using dark field imaging were unsuccessful due to the low intensity

of the scattered beams. Figure 2-17 shows that the {001} reflections are aligned with the

directions and also produce a hexagonal pattern. Since the sets of planes that give

rise to the {001) reflections have only twofold symmetry, a hexagonal pattern can only

be produced if there are multiple orientations of the crystals related by threefold

symmetry. In Figure 2-16, the needle-like particles are clearly pictured with three

preferred orientations rotated by 1200 with respect to one another. This threefold

orientation of the needle-like particles may be a consequence of the in-plane hexagonal-

close-packing of the alkyl chains in the LB film. If indeed the needle-like particles give

rise to the {001} reflections then the small "round" particles must give rise to the {hkO}

reflections and are oriented, like the large CdBr2 particles, with their [001] axis

perpendicular to the LB basal plane. The round particles are also oriented with respect

to one another within the LB plane, although the angular variance is greater than that

observed in the CdI2 system.

CdC12, like CdBr2, is also formed in oriented arrays. The CdCI2 particles grow

with their [001] axis either parallel or perpendicular to the LB plane. Unlike the CdBr2,

the observation of the two types of CdCl2 particles is not as dramatic in the electron

micrograph. For each set of crystal diffraction, however, the absence of polycrystalline

rings requires that large groups of discrete particles share a common orientation within

the LB plane. Also, like the CdBr2, the CdCl2 {001} reflections have threefold symmetry

in the diffraction pattern indicating that the crystals giving rise to these reflections must

have three orientations.

Lattice Matching

Comparing the three systems, we observe larger, more contiguous CdI2 particles

than either of the other metal halides. In some cases, single CdI2 particles up to 5 pm

across are observed. When selecting arrays of several particles, electron diffraction from

the CdI2 sample is sharper and more single-crystal-like than from similar arrays of CdBr2

and CdC12. The CdI2 particles sampled in Figure 2-12 are oriented by 40 about the

[001] zone axis while those particles giving rise to the same reflections in the CdBr2 and

CdC12 systems, although still oriented, show a wider variance of in-plane alignment, up

to 120 in the CdCl2 sample. If lattice matching with the organic template is responsible

for the observed particle orientation, perhaps the lattice match is closer for CdI2 than for

either CdBr2 or CdCl2. Comparing the (43 /2 x 43 /2)300 relationship between the (001)

faces of the metal halide and the hexagonal close packing seen in CdArac, it is clear that

the match is closest for CdI2. Of course, the potential lattice match does not need to be

the same in all three cases. With all three systems, however, the organic template is

expected to conform somewhat to the inorganic extended lattice. There is a limit to the

extent the LB film can deform before strain develops resulting in defects and domain

boundaries. If the LB layer does play a role in mediating particle growth, such strains

could limit particle size as well as the extent of particle alignment and may account for

the smaller particles and lower degree of orientational order observed for the bromide and

chloride samples. Defect sites or domain boundaries in the LB film could also be

effective nucleation sites for crystals to grow with alternative orientations such as those

observed in the CdBr2 and CdC12 systems.

Figure 2-23: Cartoon representation of the observed orientation of CdI2 within an LB
film of cadmium arachidate. Flat, hexagonal "plates" represent the layered
planes of the metal dihalide (parallel to the LB basal plane) and do not
imply particle shape or geometry. The particles are portrayed with their
"unit cell" axes aligned with respect to one another.

Figure 2-24: Cartoon representation of the observed orientations of CdBr2 and CdC12
within an LB film of cadmium arachidate. Flat, hexagonal
"plates"represent the layered planes of the metal dihalide (parallel and
perpendicular to the LB basal plane) and do not imply particle shape or
geometry. The particles are portrayed with their "unit cell" axes aligned
with respect to one another.




The idea of using the Langmuir-Blodgett method to build an organic template

which in turn limits the growth of an inorganic crystal was discussed in Chapter 1 and

further validated in Chapter 2. With the proper choice of organic and inorganic materials

oriented growth may be attained, and the cadmium halide study was very effective as a

proof of this concept. The next step is to apply the methods learned in that system to

more interesting materials, namely those with potentially interesting magnetic properties

such as the manganese halides. The manganese halides have structures and reactivities

similar to the cadmium halides. In addition, the manganese halides have the added

advantage of being paramagnetic. One of our long-term goals is the realization of a

monolayer analog of a bulk ferromagnetic material. The synthesis and characterization

of organic templated manganese halides is then one means to that end. As it happened,

the choice of the cadmium halides as model systems for the manganese halides was a

fortuitous one. The cadmium systems were occasionally difficult to deposit as Langmuir-

Blodgett films but were very well behaved in every other respect. In contrast, manganese

arachidate is very easy to manipulate and deposit as a Langmuir-Blodgett film. In terms

of their structure and amenability towards characterization, the manganese-containing LB

films and the manganese halides were a bit more difficult than the cadmium systems. In

this chapter we will describe the synthesis and characterization of the manganese halides

within a deposited Langmuir-Blodgett film.

Manganese Dihalide Structure

The manganese dihalides, MnX2 have structures similar to the cadmium dihalides

(Wells, 1984). They have either the CdI2-type or the CdCl2-type layer structures where

the anions are hexagonally close packed or cubic close packed, respectively. The metal

ions occupy the octahedral holes in every other layer of anions. In the layered manganese

halide structures, the bonding within the anion-metal-anion layer is ionic-covalent, while

the anion-anion interaction between layers is van der Waals in nature and easily cleaved

(Wells, 1984). Manganese chloride adopts the c.c.p. structure while the iodide adopts the

h.c.p. structure (Wells, 1984; Wyckoff, 1982). The dibromide of manganese, however,

may adopt either structure (Wells, 1984; Wyckoff, 1982; Ferrari; Giorgi, 1929). The

h.c.p. form of manganese bromide has been reported to be metastable at room temperature

and may undergo a c.c.p.- to h.c.p.-type phase transition at about 623 K (Schneider et al.,

1992). The authors (Schneider et al., 1992) mention that the phase transition in MnBr2

is the only one occurring for these types of layered structure. The electron diffraction

patterns of the two structures are similar, however.

Experimental Details

Materials and Procedures

Manganese(II) chloride tetrahydrate, MnC12* 4H20, 99.6% was obtained from

Fisher Chemical and used without further purification. All other chemicals were

described in Chapter 2. The respective substrates for XPS, IR, and TEM were the same

as those described in Chapter 2.

Langmuir-Blodgett depositions of the manganese arachidate were made using

slightly different conditions than for the cadmium systems. The manganese(II)

concentration was made to 1 x 10-3 M and then the subphase pH was adjusted to 6.4 with

the addition of dilute KOH. We did not observe any precipitate in the subphase at this

pH due to manganese oxide or hydroxide formation. The Langmuir monolayers were

compressed to a surface pressure, 7t, of 30 mN/m at a linear compression rate of 5

mN/m/min. At this pressure, the floating monolayers were very stable and easy to

deposit. Dipping speeds of 20 mm/min up and down resulted in very steady and uniform

deposition as judged by transfer ratio which was 1.02 0.02 in all cases. The films

deposited on all substrates (OTS coated) were very dry upon withdrawal from the

subphase, indicating that the films were of good quality. A good (and rare) study of

conditions for depositing manganese carboxylates is found in the article by Peltonen et

al. (1994).

The deposited films were exposed to the gaseous hydrogen halides, HCI, HBr, and

HI in various concentrations and for various periods of time. The exposure times ranged

from five to fifteen minutes. Partial pressures of HX in argon were typically 6.2 x 10-2

atm. Thicker films (60 layers) required longer HX-exposure times and higher HX

concentrations to effect complete reaction--closer to 15 minutes and 2.5 x 10-1 atm. In

all of the experiments that are reported, the higher exposure times have been used such

that the reaction has gone to completion.


Attenuated Total Reflectance FT-Infrared Spectroscopy

Reactions of the manganese arachidate (MnArac) LB films with hydrogen halide

(HX) gas were monitored by ATR-FTIR. Figure 3-1 shows the ATR-FTIR spectrum from

1500-4000 cm1 of a 10-bilayer MnArac LB film. Peaks at 2955 cm-1, 2915 cm-1 and

2850 cm-1 in the MnArac spectrum are the asymmetric methyl Va(CH3), asymmetric

methylene Va(CH2) and symmetric methylene vs(CH2) stretches, respectively, of the

close-packed, all-trans hydrocarbon chains in the LB film (Wood et al., 1989; Porter et

al., 1987). The band at 1562 cm-1 is the asymmetric carboxylate stretch, Va(COO-),

characteristic of the metal carboxylate salt. The small peak at 1686 cm-1 is a carbonyl

stretching band, v(C=O), and appears as a result of incomplete salt formation in the LB

film. From XPS measurements, Peltonen (1994) estimates a pKa of a manganese stearate

LB film to be 6.2. Using Peltonen's estimates (1994), our film (deposited at pH = 6.4)

should be about 60% manganese arachidate, the remainder being the protonated arachidic

acid. Judging from the IR spectrum Figure 3-1 we observe that the v(C=O) at 1686 cm-

is not very intense while the Va(COO-) peak at 1562 cm-1 is higher in intensity. Unless

TU 2850
= 0.2 au
.. 1562
Q 2955

4000 3500 3000 2500 2000 1500
Wavenumber (cmf1)

ATR-FTIR spectrum of a 10-bilayer MnArac LB film.

Figure 3-1:

there are orientational or structural effects leading to a reduction of the observed v(C=0)

band intensity, Peltonen's pKa value of 6.2 may be a little high. The film in Figure 3-1

was deposited at a pH of 6.4 and we estimate that the film is predominately in the form

of the metal-carboxylate salt.

Upon exposure to HI (Figure 3-2) the Va(COO-) band disappears from the IR

spectrum, and several new bands appear. The intense band around 1697 cm-1 (v(C=0))

results from formation of the carboxylic acid. This band is actually a doublet with

maxima at 1701 cm-1 and 1692 cm-1. It is possible that this split is due to a cis-trans

isomerization of the alkyl chains about the hydrogen-bonded dimeric carboxylic acid

headgroups (Rabolt et al., 1983) or otherwise two different crystal forms in the LB film

(Bellamy, 1975). There is no change in position, intensity, or fwhm of the methyl and

methylene stretching bands indicating that the organic layers do not undergo any gross

changes in structure during the reaction, and that there is no loss of organic material

under the reaction conditions used. In Figure 3-2, a broad absorbance between 3300 and

2500 cm-1 is observed which can be attributed to the acid O-H stretch. In contrast to the

HX-exposed cadmium arachidate films, we do observe two sharp peaks at 3502 cm-1 and

3418 cm-1 indicative of some water in the lattice. These peaks, and a small peak at 1581

cm-1, are also seen in the HCI- and HBr-exposed films and will be treated in the

discussion section. Figure 3-3 shows the IR spectrum of an HBr-exposed 10-bilayer

manganese arachidate LB film. Visible are the bands due to the alkyl chains near 2900

cm-1 unchanged from their positions in the unexposed film (Figure 3-1) and the v(C=O)

band at 1699 cm-1. Figure 3-4 presents the spectrum of an HCl-exposed 10-bilayer manganese


S T 2849
}8= 0.2 au

-" 1581
U' 2953

3502 3418

4000 3500 3000 2500 2000 1500
Wavenumber (cmf1)

Figure 3-2: ATR-FTIR spectrum of a 10-bilayer MnArac LB film after exposure to HI.


= 0.2 au


3516 3446


30 I







Wavenumber (cm 1)

ATR-FTIR spectrum of a 10-bilayer MnArac LB film after exposure to




Figure 3-3:


= 0.2 au


3518 3453







Wavenumber (cm1)

ATR-FTIR spectrum of a 10-bilayer MnArac LB film after exposure to





Figure 3-4:

arachidate LB film. The skeletal peaks near 2900 cm-1 are unchanged and the v(C=O)

peak at 1702 cm-1 shows that the conversion to the carboxylic acid is complete. The

smaller peak at 1605 cm-1 and the two peaks at 3518 cm- and 3453 cm-1 are most

probably due to water in the crystal lattice and are shifted from their values in the HI-

and HBr-exposed films. In addition, the intensity of these peaks has increased relative

to those in Figures 3-2 and 3-3.

X-Ray Photoelectron Spectroscopy

Figure 3-5 presents XPS multiplex spectral data for 10-bilayer MnArac LB films

after reaction with HCI and HBr. Unfortunately, the Mn(2p) photoelectron binding energy

region overlaps the I(3d) region making interpretation of the HI-exposed manganese

arachidate spectrum difficult even after repeated attempts at curve fitting and

deconvolution. In each case, XPS reveals that the halogen is incorporated into the metal-

containing film. As before with the cadmium dihalides, the binding-energy peak areas

are then integrated and corrected with an atomic sensitivity factor (5000 Series ESCA

Systems Version 2.0 Instruction Manual, 1989; Wagner et al., 1981) to yield the observed

relative concentrations which are presented, along with the predicted values derived from

the attenuation equation (Equations 2-1 and 2-2), in Table 3-1. In the case of the HC1-

and HBr-exposed MnArac films, the observed values are in line with the predicted values

of the relative concentrations for MX2 stoichiometries.

S6- I -I I 6 I 6 I
S670 660 650 640 630
- T-r

85 75 65 55 45

670 660 650 640 630 220 210 200 190 180

Binding Energy (eV)

Figure 3-5: XPS spectra of 10-bilayer MnArac LB films after exposure to HCI and
HBr. Spectrum of HI-exposed film is not included as the Mn(2p) region
overlaps the I(3d) region (see text).

Summary of XPS multiplex data
arachidate after exposure to HX.

for 10-bilayer films of manganese

LB Film/HX Peak KEa (eV) Observed Predictedd
Relative Relative
Concentration Concentration
(%) (%)
MnArac/HI Mn2p 601 33

_3d 624 67

MnArac/HBr Mn2p 601 34 4b 28

Br3d 1049 66 72

MnArac/HCI Mn2p 601 33 3C 27

Cl2p 1179 67 73

aKinetic energies are not corrected for peak-shifting due to sample charging effects.
b95% confidence limits N=3. c95% confidence limits N=2. Calculated for MX2
stoichiometry from Equation 2-2 using XMn = 270 A; XI = 275 A; XBr = 356 A; X1 =
378 A.

Table 3-1: