Article pubs.acs.org/cm
Investigation of Indium Phosphide Quantum Dot Nucleation and Growth Utilizing Triarylsilylphosphine Precursors Dylan C. Gary, Benjamin A. Glassy, and Brandi M. Cossairt* Department of Chemistry, University of Washington, Box 351700, Bagley Hall, Seattle, Washington 98195-1700, United States S Supporting Information *
ABSTRACT: We have developed a two-phosphine strategy to independently tune nucleation and growth kinetics based on the relative reactivity of each precursor in the synthesis of indium phosphide (InP) quantum dots (QDs). This approach was allowed by the exploration of the synthesis and reactivity of a series of sterically encumbered triarylsilylphosphines substituted at the para position of the aryl group, P(Si(C6H4X)3)3 (X = H, Me, CF3, or Cl), as a contrast to P(SiMe3)3, the P3− source commonly employed in such syntheses. UV−vis absorption spectroscopy of aliquots taken during InP QD growth revealed a stark contrast between triarylsilylphosphines with electron-donating and electron-withdrawing groups in both the rate of InP formation and the final particle size. 31P{1H} nuclear magnetic resonance spectroscopy confirmed that precursor conversion remains rate-limiting throughout the nanocrystal synthesis when P(SiPh3)3 is incorporated as the sole phosphorus precursor; however, this is insufficient for effective separation of nucleation and growth in this system because of the slow nucleation rates that result. In all cases, syntheses that employ a single chemical species as the P3− source were found to suffer from a poor match in reactivity with In(O2C(CH2)12CH3)3 as they either fail to separate nucleation from growth because of slow precursor conversion rates [P(SiPh3)3 and P(Si(C6H4-Me)3)3] or preclude size selective growth from rapid precursor conversion [P(SiMe3)3, P(Si(C6H4-Cl)3)3, and P(Si(C6H4-CF3)3)3]. To balance these two extreme cases, we developed a novel approach in which two different P3− sources were introduced to segregate nucleation and growth based on the relative reactivity of each precursor.
■
deliver P3− to the In3+ without the need for distinct oxidation or reduction steps. However, this highly reactive silylphosphine is completely consumed within seconds of injection, precluding the necessary existence of monomer reserves for an extended growth period.8,10 Moreover, the reactivity of P(SiMe3)3 creates a situation in which precursor conversion, nucleation, and growth [the three stages of colloid formation according to the LaMer model of crystallization (Scheme 1)]11 all occur simultaneously. Thus, this reagent is unable to temporally separate nucleation and growth events and does not maintain adequate monomer reserves for prolonged periods of growth, therefore limiting access to InP QDs over a wide range of sizes.12−14 Therefore, new syntheses of InP aimed at controlling
INTRODUCTION Indium phosphide (InP) is a direct band gap semiconductor that absorbs in the visible region of the electromagnetic spectrum and exhibits a larger dielectric constant, lower effective e− and h+ masses, weaker phonon coupling, and lower toxicity than related chalcogenide-based semiconductors, such as cadmium selenide.1 These properties result, in part, from the greater lattice covalency of InP and lead to more pronounced quantization effects and greater photostability.2 These unique properties position InP as a prime candidate for solution-processed semiconductor applications, including biological imaging,3−5 and solid state lighting through electrically driven LED and down-conversion technologies.6 To date, however, colloidal syntheses have failed to produce InP quantum dots (QDs) in a range of sizes with narrow size distributions, a requirement for obtaining samples of uniform electronic structure and color purity. This challenge arises from the inherent difficulty in separating the nucleation and growth events during the synthesis of this and all covalent semiconductor nanocrystals, which is affected by the nature of the reagents required to make the material.2 The most successful and widely studied syntheses of colloidal InP rely on the hot injection method, where P(SiMe3)3 is introduced into an indium carboxylate salt in a high-boiling point solvent near 300 °C.7−9 P(SiMe3)3 is commonly employed because it has the correct electronic disposition to © XXXX American Chemical Society
Scheme 1. Three Stages of Colloidal Crystallization, Precursor Conversion, Nucleation, and Growth
Received: January 10, 2014
A
dx.doi.org/10.1021/cm500102q | Chem. Mater. XXXX, XXX, XXX−XXX
Chemistry of Materials
Article
monomer delivery rates using alternative precursors are needed, few of which have been documented.15 Here we unveil a new strategy to independently tune InP QD nucleation and growth using a two-phosphine approach, providing access to InP particles that absorb across a wide range of visible wavelengths. This approach was allowed by the synthesis of a series of sterically encumbered triarylsilylphosphines [P(Si(C6H4-X)3)3, where X = H, Me, Cl, CF3] to tune the monomer delivery rate by making precursor conversion the rate-limiting step in InP synthesis, creating a reservoir of monomers to facilitate growth. Our initial hypothesis was that the increased steric bulk of these silylphosphines should decrease their reactivity toward indium carboxylate salts. This should lead to longer periods of growth and, hence, the ability to grow differently sized samples of InP in a single-pot procedure. These triarylsilylphosphine precursors were designed to have similar steric properties but to have variable P−Si bond polarities through para functionalization. Indeed, we have demonstrated that precursor conversion can be made rate-limiting; however, this is not sufficient to obtain monodisperse nanocrystals over a range of sizes because of the slow rate of nucleation in this system. To overcome the slow rate of nucleation while retaining rate-limiting precursor conversion, we turned to a strategy combining two reagents with disparate reactivity profiles, namely, P(SiMe3)3 and P(SiPh3)3, which has given us the ability to independently toggle InP QD nucleation and growth to enter an artificial LaMer-like growth profile yielding temporally resolved nucleation and growth periods.
The absorbance at 310 nm, a region in which the InP extinction coefficient is relatively size-independent15,19 and thus indicative of InP concentration in all forms, revealed two distinct InP production rates for P(SiMe3)3, P(SiPh3)3 (Ph = C6H5), P(Si(C6H4-Me)3)3, P(Si(C6H4-Cl)3)3, and P(Si(C6H4CF3)3)3 rather than a continuum of InP formation rates as originally hypothesized (Figure 1A). The normalized absorbance at 310 nm shows that P(Si(C6H4-CF3)3)3 and P(Si(C6H4Cl)3)3 (panels D and F of Figure 1, respectively) resulted in rapid formation of InP with approximately 90% conversion to InP within 30 min, as indicated by the absorption at 310 nm in the UV−vis spectrum. This rapid formation of InP is also observed for syntheses solely employing P(SiMe3)3 (Figure 1B). In contrast, both P(SiPh3)3 and P(Si(C6H4-Me)3)3 require approximately 60 min to exhibit a similar degree of precursor conversion (panels C and E of Figure 1, respectively). Similarly, the maximal absorbance (λmax) for the lowest-energy electronic transition (LEET) from the UV−vis spectra of the 90 min timed aliquots of analogous syntheses reveals two distinct final particle sizes with a λmax of approximately 530 nm for the rapidly reacting precursors P(SiMe3)3, P(Si(C6H4-Cl)3)3 and P(Si(C6H4-CF3)3)3 and a λmax of approximately 585 nm for P(SiPh3)3 and P(Si(C6H4-Me)3)3 indicative of average particle diameters of approximately 3 and 4 nm, respectively.20 That we see two distinct rates rather than a continuum of rates for the para-substituted triarylsilylphosphines is suggestive of two distinct mechanistic pathways in this class of compounds (or one mechanism with two different rate-determining steps). Steric bulk was expected to play a major role in altering the reactivity of the phosphine precursors; however, all the parasubstituted triarylsilylphosphines have similar cone angles (Σ Si−P−Si angles of 335 ± 1° for all triarylsilylphosphines from DFT calculations),21 yet we see that silylphosphines with EDGs react slowly and silylphosphines with EWGs rapidly. Two potential mechanisms are presented in Scheme 3, one involving preformation of an In−phosphine complex and subsequent SN2-like reactivity (mechanism A) and the other involving P−Si bond ionization prior to reaction with indium carboxylate (mechanism B). We would suspect that mechanism A would be facilitated by EWGs in the para positions placing a partial positive charge on the silicon and facilitating attack by the oxygen nucleophile. In the case of EDGs, such a pathway might be prohibitively slow, raising the possibility of mechanism B. This makes sense from an electronic structure perspective and is provocative with respect to designing new phosphine precursors with covalent or ionic character for phosphide material synthesis. Observation of a chemical transformation of the P3− precursor prior to incorporation into InP would undermine the efforts of designing these novel reaction-limiting precursors. In the case of the sluggishly reacting P(SiPh3)3 precursor, nuclear magnetic resonance (NMR) spectroscopy should allow an opportunity to track its conversion prior to incorporation into nanocrystals. Using this technique, previous investigations in our lab have uncovered the detrimental speciation of P(SiMe3)3 to HnP(SiMe)3−n prior to InP formation in the presence of various proton sources, such as carboxylic acids, which are commonly used as ligands in QD syntheses.10 Control experiments that yield InP QDs performed with 2 equiv of In(O2C(CH2)12CH3)3, 1 equiv of P(SiPh3)3, and 0.6 equiv of HO2C(CH2)12CH3 in a sealed J-Young NMR tube heated to 285 °C demonstrate the slow, monotonic disappearance of P(SiPh3)3 over the same time period as InP
■
RESULTS AND DISCUSSION A. Synthesis and Reactivity of P(SiAr3)3 for InP QDs. We have prepared a series of sterically demanding parasubstituted triarylsilylphosphines from red phosphorus, HSiCl3, and the appropriate aryl Grignard reagent to explore their effect on monomer delivery rate in the synthesis of InP nanocrystals. Our original hypothesis was that electron-donating groups (EDGs) or electron-withdrawing groups (EWGs) at the para position in each aromatic ring would provide a means of controlling the rate of delivery of the InP monomer to the system by fine-tuning the P−Si bond polarity as well as the basicity at phosphorus. InP quantum dots were grown by injecting a room-temperature solution of silylphosphine (1 equiv) in a eutectic mixture of diphenyl ether and biphenyl16 into a solution of indium myristate (2 equiv) and myristic acid (0.6 equiv) in 1-octadecene at 315 °C. Growth was then maintained at 270 °C for 90 min (Scheme 2). The QD growth Scheme 2. Synthesis of InP Quantum Dots
was monitored by UV−vis absorption spectroscopy using aliquots taken over the course of the reaction. All InP QDs explored in this study using either the newly prepared triarylsilylphosphines or trimethylsilylphosphine exhibit photoluminescence quantum yields of
