The Surface Chemistry of Metal Chalcogenide Nanocrystals Nicholas Charles Anderson Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Graduate School of Arts and Science Columbia University 2014 © 2014 Nicholas C. Anderson All rights reserved ABSTRACT The Surface Chemistry of Metal Chalcogenide Nanocrystals Nicholas C. Anderson The surface chemistry of metal chalcogenide nanocrystals is explored through several interrelated analytical investigations. After a brief discussion of the nanocrystal history and applications, molecular orbital theory is used to describe the electronic properties of semiconductors, and how these materials behave on the nanoscale. Quantum confinement plays a major role in dictating the optical properties of metal chalcogenide nanocrystals, however surface states also have an equally significant contribution to the electronic properties of nanocrystals due to the high surface area to volume ratio of nanoscale semiconductors. Controlling surface chemistry is essential to functionalizing these materials for biological imaging and photovoltaic device applications. To better understand the surface chemistry of semiconducting nanocrystals, three competing surface chemistry models are presented: 1.) The TOPO model, 2.) the Non- stoichiometric model, and 3.) the Neutral Fragment model. Both the non-stoichiometric and neutral fragment models accurately describe the behavior of metal chalcogenide nanocrystals. These models rely on the covalent bond classification system, which divides ligands into three classes: 1.) X-type, 1-electron donating ligands that balance charge with excess metal at the nanocrystal surface, 2.) L-type, 2-electron donors that bind metal sites, and 3.) Z-type, 2-electron acceptors that bind chalcogenide sites. Each of these ligand classes is explored in detail to better understand the surface chemistry of metal chalcogenide nanocrystals. First, chloride-terminated, tri-n-butylphosphine (Bu P) bound CdSe nanocrystals were 3 prepared by cleaving carboxylate ligands from CdSe nanocrystals with chlorotrimethylsilane in Bu P solution. 1H and 31P{1H} nuclear magnetic resonance spectra of the isolated nanocrystals 3 allowed assignment of distinct signals from several free and bound species, including surface- bound Bu P and [Bu P–H]+[Cl]- ligands as well as a Bu P complex of cadmium chloride. 3 3 3 Nuclear magnetic resonance spectroscopy supports complete cleavage of the X-type carboxylate ligands. Combined with measurements of the Se:Cd:Cl ratio using Rutherford backscattering spectrometry, these studies support a structural model of nanocrystals where chloride ligands terminate the crystal lattice by balancing the charges of excess Cd2+ ions. The adsorption of dative phosphine ligands leads to nanocrystals who’s solubility is afforded by reversibly bound and readily exchanged L-type ligands, e.g. primary amines and phosphines. The other halides (Br and I) can also be used to prepare Bu P-bound, halide-terminated CdSe nanocrystals, however 3 these nanocrystals are not soluble after exchange. The change in binding affinity of Bu P over 3 the halide series is briefly discussed. Next, we report a series of L-type ligand exchanges using Bu P-bound, chloride- 3 terminated CdSe nanocrystals with several Lewis bases, including aromatic, cyclic, and non- cyclic sulfides, and ethers; primary, secondary, and tertiary amines and phosphines; tertiary phosphine chalcogenides; primary alcohols, isocyanides, and isothiocyanides. Using 31P nuclear magnetic resonance spectroscopy, we establish a relative binding affinity for these ligands that reflects electronic considerations but is dominated primarily by steric interactions, as determined by comparing binding affinity to Tolmann cone angles. We also used chloride-terminated CdSe nanocrystals to explore the reactivity of ionic salts at nanocrystal surfaces. These salts, particularly [Bu P–H]+[Cl]-, bind nanocrystals surfaces as L-type ligands, making them soluble 3 in polar solvents such as acetonitrile. This information should provide insight for rational ligand design for future applications involving metal chalcogenide nanocrystals. The strongest ligand, primary n-alkylamine, rapidly displace the Bu P from halide-terminated CdSe nanocrystals, 3 leading to amine-bound nanocrystals with higher dative ligand coverages and greatly increased photoluminescence quantum yields. The importance of ligand coverage to both the UV-visible absorption and photoluminescence spectra are discussed. Finally, we demonstrate that metal carboxylate complexes (L–M(O CR) , R = oleyl, 2 2 tetradecyl, M = Cd, Pb) are readily displaced from carboxylate-terminated ME nanocrystals (ME = CdSe, CdS, PbSe, PbS) by various Lewis bases (L = tri-n-butylamine, tetrahydrofuran, tetradecanol, N,N-dimethyl-n-butylamine, tri-n-butylphosphine, N,N,N',N'-tetramethylbutylene- 1,4-diamine, pyridine, N,N,N',N'-tetramethylethylene-1,2-diamine, n-octylamine). The relative displacement potency is measured by 1H nuclear magnetic resonance spectroscopy and depends most strongly on geometric factors like sterics and chelation, though also on the hard/soft match with the cadmium ion. The results suggest that ligands displace L–M(O CR) by cooperatively 2 2 complexing the displaced metal ion as well as the nanocrystal. Removal of surface-bound Cd(O CR) from CdSe and CdS nanocrystals decreases their photoluminescence quantum yield 2 2 and decreases the apparent intensity of the 1S -2S absorption without changing the position of e 3/2h the first electronic transition or the nanocrystal size. These changes are partially reversed upon rebinding of M(O CR) at room temperature (~60 %) and fully reversed at elevated temperature. 2 2 These results indicate that reproducible optical properties and ligand exchange reactivity may be achieved if the surface monolayer of excess metal ions is carefully controlled. In fact, using well- purified nanocrystals samples free from acidic impurities that may form L-type ionic salts on reaction with primary amines, we have synthesized the first fully stoichiometric CdSe nanocrystals though L-type promoted Z-type displacement. These amine-bound nanocrystals perform well in thin-film transistors and can be used as a synthon for Z-type addition reactions. These studies together inform a coherent picture of metal chalcogenide nanocrystal surface chemistry. In particular, they reinforce the neutral fragment model as a viable pedagogical tool for training new inorganic materials chemists. We also discuss the connection between surface chemistry and the optical properties of CdSe nanocrystals. Both L-type and Z- type ligands are required to fully passivate these materials. Most importantly, managing nanocrystal composition and optioelectronic properties demands careful analytical techniques. While nanocrystals are not simple molecules, they can be manipulated through rational design informed by accurate models. The work described here establishes a deeper understanding of nanocrystal surface chemistry and provides an example for future studies designed to explore the fundamental physics of these materials and functionalize metal chalcogenide nanocrystals for real-world applications. TABLE OF CONTENTS Acknowledgements ii Dedication xiii Chapter I. Introduction: The applications, electronics, and surface chemistry of 1 semiconductor nanocrystals Chapter II. X-type ligand exchange: Halide-termination of cadmium selenide 52 nanocrystals Chapter III. L-type ligand exchange: Relative binding affinity of Lewis bases for 84 chloride-terminated cadmium selenide nanocrystals Chapter IV. Z-type ligand displacement: Removal of metal carboxylate 153 complexes from metal chalcogenide nanocrystals Chapter V. Conclusion: Utilizing the neutral fragment model of metal 232 chalcogenide nanocrystal surface chemistry Appendix I. Comprehensive list of nanocrystal samples 247 Appendix II. Representative Spectra for cadmium selenide nanocrystals with 250 various Z-type ligand coverage i Acknowledgements There's a scent, a touch, a slight sound, a flash perhaps with a crackle. The fruit is plucked, and suddenly (in)complete. The succulent fruit is sweet on the tongue (like scotch), and a woman smiles at the taste, thinking this is fit for us both. She gives it to the Man, and sadly, he becomes merely man. This fall though is not the fault of the fruit nor the sly serpent, nor the woman, but merely fate. Significant, yes, very much so, but damning save forgiveness of some demagogue, I think not. However, this is an acknowledgement, and root-word dedications aside, I (as the author, with all my footnote fetishes and parenthetical pleasures) must deeply appreciate, with the uncharacteristically sincere dedication warbled in the heartfelt lyrics of danke schön. a vast array of folks. Like all things in the post-modern day, finding the proper place to begin can be difficult. I've read so many dedications and acknowledgements for inspiration, that my head spins like that ride at the Puyallup fair, the one where the floor drops out and you, as the rider (not the writer or necessarily the reader, for I, the writer, in terms of your prior knowledge, have little or none, and as such must rely on my very own experience to dictate my feels (mostly my thankfulness) in a long-winded way, to you, the reader, in order to disseminate my gratitude towards you and fellow audience members) stick to the sides of this giant cardboard tube as it spins round and round, the force pulling us outwards, as if to fling us uncontrollably out into the world. I then lift my arm up, gesturing perhaps, but more likely just gesturing in vain, and witness the force of the wild spin and am at awe. Finally, the ride stops, and an awkward child walks forth and demonstrates himself to the world. ii Thus we begin, comically I hope, a long list of unending thanks to a great many people who sadly, will end some day, hopefully happy. Such is the nature of tasting sweet fruit, and while I love to argue the benefits of an apple a day, today is not the time nor place for such calisthenics. Perhaps, given my obtuse (hopefully not obtrusive) introduction, I should begin with a platitude to σοφία, goddess of grace and the unmentionable. Clever audiences (see: readers) would already be aware of such praises however, so it is best that we take the tale forward to mortal creatures. To finish off with nectar infested things, I first must thank the bees for not slaying me when they had the chance. Much more significantly, for bees cannot read, my first human appreciation (an honorable adoration) is to Assistant Professor, disc golf extraordinaire, Jonathan Scharle (yes, I know your middle name) Owen, Doctorate of the Philosophy in Chemistry. I am teary-eyed as I say he is possibly the best young advisor an ambitious young graduate student could have. I almost came to work the first day, on a mid-august day in the City of New York, with a skinny tie, grey suit, and fedora in hand (because, obviously one can't wear a fedora on head inside). Jon quickly taught me however that none of such an outfit was in style in the chemistry world. In fact, he educated me in all that there is to know about the chemistry world at large, particularly with regards to the current state of affairs regarding nanocrystals and their mysteries. We used to ventilate theories of nanocrystal surface chemistry after long days of constructing laboratory equipment and battling Snyder group members out of our department-given laboratory space. I particularly recall with fondness the hours before distilling death in the form of dimethylcadmium. He had instructed me to inform my loved ones of my potentially imminent demise, and we watched curling, stark sober. Have you, my reader, ever watch curling without a single drop of alcohol? Commitment. Despite, or more likely because of, our future battles on iii both congruent and opposing sides, in conference rooms and metaphorical(?) forests, I find it a great fortune to say that Jonathan Owen has been my advisor through this ordeal, and will always be a guide in my future. He is the Yoda to my Skywalker. Next I would like to thank Professor (though I knew her best as Doctor) Brandi Cossairt. She came to (Jon and) me as an ambitious post-doc and really crystallized the Owen group. I had assigned things to their places, but she assigned them to their proper places. Without her guidance, I would have been frolicking restlessly in the waves of creation without a tide to pull me in towards the shore. I may be the first son of the Owen group, but Brandi was the first mother. For that I am thankful. Oh, and she also gave me the inspiration for distillation of 1- octadecene, essential to mass utilization of the Cao synthesis, and taught the lab to use septa for vacuum distillation of small volumes in 20 ml vials. Also, her taste in music is awesome and Jared (her husband) is a killer Husky. Go Dawgs! Speaking of firsts with regards to the Owen group, while I was the first full Owen group student, I want to thank the first half student of the Owen group, Ava Kreider-Mueller. She has the best voice in the group, and she has help me through some fairly crazy times and let me wail on her rock band drums for a night here and there with only bug-slaying as payment. Plus, she helped my undergrad, the wonderful Aya Buckley, with the VT-NMR experiments described in Chapter III. We don't even fully understand those experiments yet, but they are awesome so kudos of the supreme variety to both of you. Speaking of whom, I should thank Aya Buckley, currently a graduate student and UC Berkeley, and future powerhouse chemistry. Perhaps in the most truly honest and expiatory moment in the history of the Owen group, she brought me brownies after smashing some insignificant NMR tube the night before. For that ray of sunshine, and so many other fond memories (and P,P-dimethyl-n-octylphosphine) I am indebted. iv