Upscaling Colloidal Nanocrystal Hot-Injection ... - ACS Publications

Dec 22, 2016 - Figure 2F. The temperature drop associated with addition of the cold injection ..... electrics, phase-change memories, etc.). □ ASSOC...
1 downloads 0 Views 1MB Size
Subscriber access provided by University of Newcastle, Australia

Article

Upscaling Colloidal Nanocrystal HotInjection Syntheses via Reactor Underpressure Maksym Yarema, Olesya Yarema, Weyde M.M. Lin, Sebastian Volk, Nuri Yazdani, Deniz Bozyigit, and Vanessa Wood Chem. Mater., Just Accepted Manuscript • DOI: 10.1021/acs.chemmater.6b04789 • Publication Date (Web): 22 Dec 2016 Downloaded from http://pubs.acs.org on December 26, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Chemistry of Materials is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

Maksym Yarema,*,† Olesya Yarema,† Weyde M. M. Lin,† Sebastian Volk,† Nuri Yazdani,† Deniz Bozyigit,† Vanessa Wood*,† †

Laboratory for Nanoelectronics, Department of Information Technology and Electrical Engineering, ETH Zurich, CH-8092 Zurich, Switzerland ABSTRACT: We report an approach to linearly upscale hot-injection syntheses of colloidal nanocrystals by applying mild underpressure to the flask reactor prior to the injection such that rapid addition of large volumes of precursor is facilitated. We apply this underpressure-assisted approach to successfully upscale synthetic protocols for metallic (Sn) and semiconductor (PbS, CsPbBr3 and Cu3In5Se9) nanocrystals by one-to-two orders of magnitude to obtain tens of grams of nanocrystals per synthesis. Here, we provide the technical details of how to carry out underpressure-assisted upscaling and demonstrate that nanocrystal quality is maintained for the large-batch syntheses by characterizing the size, size distribution, composition, optical properties, and ligand coverage of the nanocrystals for both small- and large-scale syntheses. This work shows that fast addition of large injection volumes does not intrinsically limit upscaling of hot injection-based colloidal syntheses.

INTRODUCTION Synthesis of monodisperse colloidal nanocrystals (NCs) relies on fast homogeneous nucleation from supersaturated solution.1 Instantaneous formation of a large number of seed nuclei is ensured either by strong heating (i.e., non-injection heating-up approach) or by fast addition (i.e., injection) of an elemental precursor or reducing agent to the hot reaction mixture. The latter methods, often referred to in literature as hot-injection techniques, are convenient and fast one-pot processes, which are capable of providing ultra-narrow size distributions (e.g., In NCs with σ = 2.2% size distribution)2 due to effective time separation between nucleation and growth processes.1 However, it is this fast addition of large volumes that presents a serious challenge for upscaling hotinjection protocols. Several ways to tackle the problem have been provided. These, however, imply either an entire re-design of the recipe (e.g. slowing down the pre-nucleation reaction3 or using highly-concentrated precursor solutions)4 or the use of specially designed jet injectors.5, 6 Despite this progress in chemistry and engineering, it is increasingly stated that hotinjection technique is suitable only for small-scale exploratory research, providing typically less than 100 mg of NC product, and is not scalable in the lab environment.7 Here we focus on the possibility to upscale injection-based syntheses of colloidal NCs without modifying the original protocol or using specially designed jet equipment. We contend that such upscaling is possible at least by one-to-two orders of magnitude. This paper presents an easy and universal solution for linear upscaling of hot-injection synthesis (Figure 1). Applying a mild vacuum to the reaction mixture prior the injection enables an injection rate of 100-150 mL·s1 such that large volumes of 200-500 mL can be introduced

into the reaction flask within few seconds. For comparison, conventional syringe-assisted injection rate lay between 2 and 10 mL·s-1. We used this approach previously to produce sufficient amounts of high quality PbS NCs for inelastic neutron scattering experiments, which require up to 10 grams of material per sample.8 In this paper, we study an underpressure-governed hot-injection method in detail. In particular, we examine possible effects from upscaling, comparing large-scale and conventional small-scale series of syntheses for welldocumented PbS synthesis.9-11 We then demonstrate the versatility of the underpressure-governed hot-injection technique by upscaling three other synthetic protocols, which generate materials of industrial interest and also present different synthetic challenges. Synthesis of metallic NCs is of interest for lithium-ion batteries, catalysis, and phase-change applications, and preparing Sn NCs12 presents an interesting case by requiring two sequential injections. Synthesis of the inorganic perovskite NCs, CsPbBr3,13 is of interest for LEDs and requires injection of warm solution. Synthesis of another ternary semiconductor, Cu3In5Se9,14 which is of interest for photovoltaic, lighting and bio-applications, presents a system where the injection volume exceeds the volume of reaction mixture. We show that the multigram-scale quantities of PbS, Sn, CsPbBr3 and Cu3In5Se9 NCs prepared via underpressure-assisted hot-injection method are of high quality exhibiting narrow size distributions, excellent optical properties, and reproducible compositions. This work demonstrates that the speed of injection does not represent an intrinsic limit for the upscaling of injection-based syntheses of colloidal NCs.

ACS Paragon Plus Environment

Chemistry of Materials

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1. Schematic illustration of set-ups and obtained products for (A) conventional syringe-assisted hot-injection and (B) underpressure-governed hot-injection methods.

EXPERIMENTAL SECTION General remarks. All syntheses are carried out under inert atmosphere using a typical vacuum gas manifold. Smallscale syntheses follow original recipes9, 12-14 and same reagents and purity grades are used for both the large and small scale syntheses. Large-scale syntheses were carried out in a 2 L three-neck flask reactor, connected via a 50 cm doublewall reflux condenser to the vacuum gas manifold. An addition funnel was connected to one side of the flask, containing pre-prepared injection mixture. Injection was made applying a mild vacuum to the flask reactor for few seconds prior the injection, followed by opening the stopcock of the addition funnel (see Supporting Videos 1 and 2). Stirring of the reaction mixture was facilitated with 5 cm elliptic magnetic stirrer with moderate rotating speeds of 600-900 rpm. A list of glassware is provided in Table S1. Detailed protocols are given below and complete purification procedures are described in the Supporting Information (Figure S1). Safety notes. (I) The flask reactor contains relatively large amounts of reagents, which might include heavy metal salts, harmful byproducts and hot solvents. Therefore all handling should be carried out in well-ventilated fume hood. The fume front window should be closed as much as possible during the process. (II) Applying a mild vacuum to the flask reactor induces a vigorous boiling of the reaction mixture (see Supporting Videos 1 and 2). The condenser with internal cooling coil must be used in order prevent an introduction of chemicals into the vacuum gas manifold. Optionally a cold Dewar trap (see Table S1) can be installed on top of

Page 2 of 8

condenser. (III) Under no circumstances should the 2 L flask reactor be left unsupported from below. (IV) When cooling the reactor flask to room temperature, an operator must keep in mind do not exceed the certified temperature difference between the glass reactor temperature and external coolant (typically 150-200°C, depending on the glass type). The use of pressurized air and a hot water (or oil) bath is recommended for cooling the reactor down to ~ 200°C. Large-scale synthesis of PbS nanocrystals. PbO (18 g, 80 mmol) is mixed with 70 mL oleic acid and 730 mL 1octadecene in the 2 L three-neck flask reactor. The addition funnel is loaded with 8.4 mL (40 mmol) bis(trimethylsilyl)sulfide and 400 mL of dried 1-octadecene. The system is connected to the vacuum gas manifold and heated to 150°C under vacuum. The turbid yellow solution turns colorless, indicating formation of lead (II) oleate solution. After 2 hours, the set-up is backfilled with nitrogen and left for another 20 min for the temperature stabilization. Prior to injection, the heating mantle is removed from around the flask reactor and vacuum is applied for few seconds, until pressures reaches 1-10 mbars. After the vacuum valve is closed, the stopcock of the addition funnel is opened, resulting in fast injection of the sulfur precursor. The injection is finished within 3 seconds and the stopcock to the addition funnel is closed. An established 3 stage temperature profile is followed: (i) natural cooling to 100°C for 4 min (initial temperature drop due to the injection, followed by slow cooling in an absence of heating mantle), (ii) stabilization at 100°C for 5 min (heating mantle is reinstalled), and (iii) reaction termination with water bath.9 The size of PbS NCs is tuned by the concentration of oleic acid, initially added to the reaction mixture. Because the total volume of reaction mixture is kept constant, the amount of 1-octadecene is adjusted accordingly. The initial concentration of oleic acid is calculated as a ratio between oleic acid and a sum of oleic acid and 1-octadecene volumes. The scaling factor compared to the original protocol is 20. Large-scale synthesis of Sn nanocrystals. Oleylamine (1 L) is placed in the 2 L three-neck flask reactor. Two addition funnels are prepared containing 24.04 g (144 mmol) of LiN(SiMe3)2 dissolved in 80 mL of anhydrous toluene in one and 16 mL of 1 M diisobutylaluminium hydride in tetrahydrofuran in second one. The system was assembled to the vacuum gas manifold using Claisen adaptor before the condenser and a cold trap on top of condenser (Table S1, Supporting Video 2). Following the original recipe,12 oleylamine is heated to 140°C under vacuum for 1.5 hours and then cooled down to 50°C, at which point 3.8 g (20 mmol) of anhydrous SnCl2 are added under a N2 stream. The system is sealed again and heated for another 30 min at 110°C/vacuum conditions. Finally, flask reactor is backfilled with N2 and heated to set injection temperature (180°C). Prior to the injection, mild vacuum is applied to the flask reactor for few seconds, reaching 1-10 mbars. After the vacuum valve is closed, two injection mixtures are sequentially added with a delay of 10 s, as described for small-scale protocol.12 The reaction solution turns brown after addition of reducing agent, indicating fast formation of Sn NCs. For optimal size distribution, the reaction mixture is held at 180°C for 1 h and

ACS Paragon Plus Environment

Page 3 of 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

then cooled down with ice-water bath. The scaling factor, comparing to original protocol is 40. Large-scale synthesis of CsPbBr3 nanocrystals. PbBr2 (13.8 g, 37.6 mmol) is mixed with 1 L of 1-octadecene in the 2 L three-neck flask reactor and heated to 120°C under vacuum. In 1 hour, 100 mL of dried oleylamine and 100 mL of dried oleic acid are injected to the PbBr2-octadecene mixture, which lead to quick dissolution of PbBr2. Afterwards, reaction mixture is heated to set injection temperature (180°C). Meanwhile, Cs-oleate injection mixture is prepared by dissolving 1.63 g (5 mmol) of Cs2CO3 and 3.2 mL (10 mmol) of oleic acid in 80 mL of 1-octadecene at 120°C under vacuum. The Cs-oleate solution is then transferred to the warmed addition funnel (lab heater) prior the injection. Next, the heating mantle is removed from the flask and mild vacuum is applied to the flask reactor for few seconds, reaching 1-10 mbars. After vacuum valve is closed, warm Cs-oleate injection mixture is added by opening the funnel stopcock. The reaction is terminated 5 s after Cs-oleate injection with ice-water bath. The scaling factor, comparing to original protocol is 200. Large-scale synthesis of Cu3In5Se9 nanocrystals. CuCl (4.95 g, 50 mmol) and InCl3 (11.05 g, 50 mmol) are dissolved in 300 mL of tri-n-octylphosphine and the solution is transferred air-free to the 2 L three-neck flask reactor. The addition funnel is loaded with 15.8 g (200 mmol) of Se and 50.1 g (300 mmol) of LiN(SiMe3)2 dissolved in 300 mL of anhydrous toluene. The system was assembled to the vacuum gas manifold and heated to 100°C under vacuum. In 1 hour, the set-up is backfilled with N2 and heated to set injection temperature (290°C). Prior the injection, mild vacuum is applied to the flask reactor for few seconds, reaching 1-10 mbars. After the vacuum valve is closed, the stopcock of addition funnel is opened, allowing fast injection of Se and amide precursor mixture. The reaction temperature decreases to 200°C and is slowly raised to 220°C during 5 min of growth time. Reaction is terminated by hot-water bath and then by an ice-water bath. The scaling factor, comparing to original protocol is 50. Purification of nanocrystals. As-synthesized NCs are purified with a typical solvent/non-solvent approaches. Rotanta 460 Lab Benchtop Centrifuge (Hettich), equipped with 4×750 mL swing-out rotor is used. Details of the purification steps are in the Supporting Information. Characterization. The NCs are characterized with FEI Tecnai F30 transmission electron microscope (TEM images), FEI Quanta 200 SEM microscope (EDX spectroscopy), Rigaku SmartLab 9 kW System (SAXS), Agilent Cary 5000 spectrophotometer (absorption spectra), custombuild PL set-up, using 532 nm CW laser and Ocean Optics QE65000 spectrometer (photoluminescence spectra) and Bruker V70 FTIR spectrometer (FTIR data). Size distributions are measured with ImageJ software. SAXS spectra are fitted with NanoSolver software (Rigaku). Temperature profiles are collected with KEM-net software (J-KEM).

Figure 2. (A) Position of excitonic peak as a function of oleic acid initial concentration for small-scale and large-scale syntheses of PbS nanocrystals. (B) Absorption spectra, (C,D) TEM images, size distributions (inset (C,D)), and temperature profiles (complete in (E) and zoomed region and its first derivative in (F)) for comparable PbS syntheses, prepared with 8.75 vol.% of oleic acid. Data for large-scale syntheses in red and for small-scale syntheses in blue.

RESULTS AND DISCUSSION In this section, we present comparisons of NCs obtained with the small- and large- scale syntheses for the four different materials, and use the findings to draw conclusions about the versatility of the upscaling approach. Large-scale synthesis of PbS nanocrystals. Synthesis of PbS NCs is based on fast reaction between Pb oleate and the organosulfur compound, bis(trimethylsilyl)sulfide. Although PbS NCs form within seconds, the size distribution was found to improve when growth times are extended up to 10 min at elevated temperatures.9 In this paper, we synthesize two size series of PbS NCs – one for a conventional reaction volume (250 mL) and another one using 2 L flask reactor and underpressure-governed hot-injection method (Figure 2A). We tune the size (and position of the excitonic peak) of PbS NCs by the concentration of oleic acid. Both syntheses use the same injection temperature (Tinj = 150°C), Pb and S precursors concentrations, and the same post-injection temperature profile strategy.9 For both syntheses, the size of the PbS NCs increases (optical band gap of PbS NCs narrows) with increasing oleic acid, but PbS NCs from the large-scale synthesis have smaller average size (Figure 2B). To explain this, we examine a small- and large-scale synthesis for a constant concentration of oleic acid (COA = 8.75%, shaded diamonds in Figure 2A). Supporting Video 1 shows time-synchronized videos of two processes. TEM images reveal small size

ACS Paragon Plus Environment

Chemistry of Materials

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 8

Figure 3. (A,B) Absorption spectra of size series for PbS nanocrystals prepared with small- and large-scale synthesis approaches. (C) Energies and full-width-of-half-maxima of first excitonic peaks extracted from (A,B). (D) Small angle X-ray spectra of PbS nanocrystals with excitonic peak energy of 1.33 eV (offset for clarity), and (E) size distribution functions from the fits in (D). (F) Estimated concentration of long hydrocarbon chain molecules, normalized to the total surface area of PbS nanocrystals as a function of purification steps. Data for largescale syntheses in red and for small-scale syntheses in blue.

distributions for both (Figure 2C and 2D). Since all initial photon energies (i.e., the size of PbS NCs decreases). The reaction parameters are kept constant, we track the temperadependency for small-scale and large-scale batches is the ture profiles throughout the synthesis (Figure 2E). The temsame for entire range of PbS sizes (Figure 3C). The FWHM perature profile consists from three stages: (i) natural coolof the PL spectra exhibit the same trends (Figure S2). These ing of the flask reactor (i.e., the heating mantle is removed), findings indicate that the quality of PbS optical properties is not influenced by upscaling the synthesis via underpressure(ii) maintenance of T ~ 100°C (the heating mantle is placed governed hot-injection method. Likewise, SAXS measureback), and (iii) reaction termination by cooling to room temments on large-scale and small-scale batches (Figure 3D), perature.9 which exhibit excitonic peak energies of Eexc = 1.33 eV, The faster injection for large-scale synthesis systematishow equivalent ~ 9% size distributions (Figure 3E). cally lowers process temperature during first 1.5 min after injection with the main difference observed during first tens We further test reproducibility of the large-scale syntheses of seconds after injection (Figures 2E). A zoomed region of by applying same reaction conditions to several batches the first 30 s after injection and a time derivative is shown in (Figures 3C and S2B). We tested the largest and the smallest Figure 2F. The temperature drop associated with addition of sizes of PbS NCs (Figure S3) since it is these syntheses cold injection solution to the hot reaction mixture, occurs which frequently pose difficulties for consistency.15 Altquicker and is larger for the large-scale synthesis. A temperhough the large-scale set-up shows certain deviations of absorption and PL peak energies and widths, the reproducibilature drop rate of 4°C·s-1 is achieved for large-scale syntheity is comparable to that of small-scale PbS synthesis or to sis 3 s into the reaction, while for small-scale synthesis the other hot-injection protocols.6 fastest temperature drop is only 3°C·s-1 and is observed at the reaction time of 7-8 s. Since decreased PbS size has been To allow NCs to be used in an electronic device, it is most reported for lower growth temperatures,10, 11 the different often necessary to purify the NCs after their synthesis to retemperature profile explains the smaller average sizes of PbS move excess organics. To complete the comparison between NCs prepared by underpressure-governed hot-injection underpressure-governed and syringe-assisted hot-injection technique (Figure 2A). Furthermore, the minima in the timetechniques, we compare batch purification of NCs from derivatives of temperature profiles correspond to the times small- and large-scale syntheses by monitoring the amount at which injections end (see Supporting Video 1), indicating of organic constituents after each washing cycle (i.e., after a much faster injection rate for the underpressure-governed centrifuging the colloidal solution). Details for the washing injection vs. conventional syringe-assisted approach (~ 130 of PbS NCs and evaluation of organic molecules concentramL·s-1 and ~ 5 mL·s-1, respectively). tion can be found in the Supporting Information. FTIR spectroscopy was used to estimate the molar concentrations of Next, we compare the quality of PbS NCs obtained from long hydrocarbon chain compounds by means of calibration small-scale and large-scale syntheses. Figures 3A, 3B, S2, curves for oleic acid in tetrachloroethylene, while sizes and and S3 show absorption and PL spectra for both size series molar concentrations of PbS NCs are extracted from the poof PbS NCs. It has previously been reported that full-widthsition of first excitonic peaks (Figure S4 and Table S2).17 at-half-maximum (FWHM) of first excitonic peak is broader 3, 15 Figure 3F shows concentration of organic molecules (i.e., for small-size PbS NCs. Possible explanations stem from ligands), normalized to the total surface area of PbS NCs. the nonlinear dependence between optical band gap and the For both syntheses, the concentration of organics decreases size of NCs15 as well as intrinsically broader single NC linwith each subsequent washing cycle. First two washing cyewidths for small size NCs.16 Figure 3C shows that that the cles removes most of unbound 1-octadecene and oleic acid FWHM increases as the excitonic peak shifts towards higher ACS Paragon Plus Environment

Page 5 of 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Chemistry of Materials

molecules, and clean samples with density of 4-5 ligand molecules per 1 nm2 of PbS NC surface are achieved (Figure 3F, Table S2). This highlights that the washing protocols developed for small-scale syntheses can be linearly up-scaled and directly applied to large-scale batches.

Figure 4. TEM images (A,B), SAXS spectra (C) and size distributions (D) of Sn nanocrystals prepared via underpressure-governed hot-injection (large-scale synthesis, data in red) and syringe-assisted hot-injection (small-scale synthesis, data in blue) methods. Size distributions in (D) are derived from SAXS fits in (C). SAXS spectra are offset for clarity.

Large-scale synthesis of Sn nanocrystals. Sn NCs are prepared by reducing the in-situ formed Sn amide precursor. The synthetic protocol involves two sequential injections: (i) solution of lithium silylamide, which induces the formation of Sn amide intermediates, and (ii) solution of diisobutylaluminium hydride solution, which facilitates reduction of Sn precursor to metallic Sn NCs. The time between two injections should be kept as short as possible (~ 10 s) to eliminate possible side reactions (e.g., slow decomposition of Sn amide).12 To upscale the synthesis of Sn NCs, we modify the set-up by adding the Claisen adapter between three-neck flask and condenser. This enables us to connect two addition funnels to side joints and a thermocouple to the remaining opening (optionally, four-neck 2 L flask could be purchased). Supplementary Video 2 demonstrates two underpressure-governed injections for large-scale Sn synthesis. Following original recipe, the reaction is then kept at elevated temperature for 1 h to improve the size distribution of Sn NCs.12 Figures 4A,B compare TEM images of Sn NCs, prepared by small-scale and large-scale syntheses. For both cases, narrow size distributions are achieved, comparable to those of the original recipe.12 High-resolution TEM images (Figures 4A,B and S5) indicate the formation of 4-nm-thick natural SnOx shell and high crystallinity of Sn cores. SAXS spectra show that identical size distributions are achieved for large-scale and small-scale Sn NC syntheses (Figure 4C,D). As shown for small-scale synthesis,12 the size distribution of Sn NCs is defined by the growth time, which is kept constant in this work (1 h). Achieving Sn NCs shows that it is possible

to use underpressure-governed hot-injection technique to enable two large-volume injections with very short time delays of 10 s.

Figure 5. TEM images (A,B), size distributions (C), and absorption and photoluminescence spectra (D) of CsPbBr3 nanocrystals prepared via underpressure-governed hot-injection (large-scale synthesis, data in red) and syringe-assisted hot-injection (small-scale synthesis, data in blue) methods.

Large-scale synthesis of CsPbBr3 nanocrystals. The synthesis of inorganic perovskite CsPbBr3 NCs relies on fast co-precipitation of Pb bromide and Cs oleate in a mixture, containing oleic acid, oleylamine, and 1-octadecene. Apart from fast injection of large volumes, this synthesis presents two difficulties for the large-scale synthesis: (i) the Cs precursor solution should stay warm during injection and (ii) the reaction should be terminated 5 s after injection.13 While a warm injection solution can be maintained by different means (e.g., using shielded additional funnel, gently preheating the addition funnel with conventional lab heater, etc.), the necessity of fast cool down is an intrinsic problem for large-scale syntheses. We use an ice-water bath, but still observe slower cooling of large-scale synthesis with respect to the small-scale protocol. Consequently, we obtain slightly larger CsPbBr3 NCs from the up-scaled synthesis (9.7 and 9.3 nm for large- and small-scale syntheses, respectively, Figures 5A-C and S6). However, absorption and photoluminescence spectra remain similar for both samples (Figure 5D). Such moderate size-effects for >9 nm CsPbBr3 NCs have been explained by relatively small Bohr diameter of Wannier–Mott excitons in CsPbBr3 (~ 7 nm).18 Importantly, underpressure-governed hot-injection synthesis yields multigram-scale perovskite CsPbBr3 NCs with excellent photoluminescence properties, such as, narrow linewidth of photoluminescence spectrum (~ 22 nm) and high PL quantum yield of 76% (Figure 5D).

ACS Paragon Plus Environment

Chemistry of Materials

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 6. TEM images (A,B), size distributions (C), composition (D) and absorption and photoluminescence spectra (E) of Cu3In5Se9 nanocrystals prepared via underpressure-governed hot-injection (large-scale synthesis, data in red), and syringe-assisted hot-injection (small-scale synthesis, data in blue) methods. Large-scale synthesis of Cu3In5Se9 nanocrystals. Indium-rich Cu-In-Se NCs are prepared via amide-promoted synthesis, where a lithium amide superbase is co-injected with a selenium precursor into the reactor with a solution containing metal halides and tri-n-octylphosphine. The injection volume exceeds the volume of the reaction mixture, resulting in large temperature drops and relatively high injection temperature requirements.14 For large-scale synthesis of Cu3In5Se9 NCs, we inject ~ 400 mL of selenium and amide precursor into ~ 300 mL of solution containing metal halides. An underpressure-governed injection ends in 3 s, and the process temperature drops from 290°C to 200°C. The temperature is then raised to 220°C during 5 min for the growth of Cu3In5Se9 NCs. We perform the small-scale synthesis14 using same reaction conditions. Figures 6A-D show the structural properties of the obtained materials. For both syntheses, we achieve small average diameters (3.4-3.5 nm), narrow size distributions (9-11%), and the target composition (Cu3In5Se9). Large-scale synthesis provides slightly larger Cu3In5Se9 NCs, which can be associated with differences in the temperature profiles. Optical properties (Figure 6E) reflect this size difference, showing a small red shift of absorption and photoluminescence spectra for the largescale batch of Cu3In5Se9 NCs.

Page 6 of 8

CONCLUSIONS In this paper, we discuss an upscaling of injection-based syntheses of colloidal NCs. We demonstrate the use of mild vacuum to improve the injection rate up to 100-150 mL·s-1. Using the underpressure-governed hot-injection technique, we show that a wide range of conventional hot-injection syntheses can be linearly up-scaled by a factor of 20-200. We demonstrate that the quality of obtained metallic and semiconducting nanocrystals (i.e., their structural and optical properties) is preserved during upscaling. Differences between NCs prepared via large-scale and small-scale syntheses arise from different injection rates and/or different temperature profiles. For PbS NCs, the underpressure-governed injection is faster than for a conventional syringe-assisted addition. This causes a faster temperature drop and lower process temperatures during the first minute after injection, resulting in smaller NCs from the large-scale synthesis. For other materials (CsPbBr3, Cu3In5Se9, and Sn NCs), the difference between products obtained via large-scale and small-scale syntheses is smaller. This can be related to the lower sensitivity to temperature variances during these processes as well as to similar injection end-times for the large-scale and small-scale protocols. To design a large-scale synthesis of an arbitrary colloidal NC system, the temperature profile for original small-scale recipe (including the influence of temperature drop rate, the maintenance of a specific temperature, and the reaction termination) must be understood. Furthermore, the underpressure-governed hot-injection should also be designed to take the same amount of time as the original small-scale synthesis. Finally, it is important to consider the mixing dynamics of the large-scale process – a fast injection in an up-scaled synthesis could lead to conditions where the process is controlled by convective motion of mixing liquids (more details in the Supporting Information).19, 20 Reactor underpressure eliminates one of the limits facing an upscaling of the hot-injection synthesis and thus provides an easy solution for obtaining tens of grams of high-quality colloidal nanocrystals. However, further linear upscaling of hot injection syntheses will meet other limits such as reactor heat exchange and heat control, mixing dynamics, as well as purification of the obtained materials. Addressing these questions will be important to reach kilogram-scale throughput of the synthesis without losing in quality of colloidal nanomaterials, which will reinforce the commercial viability of electronic, photonic, and electrochemical devices that use large numbers of colloidal NCs (e.g., solar cells, lithium-ion batteries, thermoelectrics, phase-change memories, etc.).

Purification details of large-scale syntheses, list of used glassware, mixing time calculations, additional spectra, high-resolution TEM images, and videos of underpressure-governed hot-injections. This material is available free of charge via the Internet at http://pubs.acs.org

ACS Paragon Plus Environment

Page 7 of 8

Chemistry of Materials

* [email protected]; [email protected]

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The manuscript was written through contributions of all authors. This work is financially supported through ETH Research Grant (ETH-4212-2), Swiss National Science foundation Quantum Sciences and Technology (QSIT) NCCR (51NF40-160591), Sinergia (CRSII2-147615), and Ambizione Fellowship (PZ00P2-161249).

Authors thanks to Mario Mücklich and Dr. Alla Sologubenko for technical assistance. TEM and EDX measurements are performed at the Scientific Center for Optical and Electron Microscopy (ScopeM) of the Swiss Federal Institute of Technology, Zurich.

1. Kwon, S. G.; Hyeon, T., Formation Mechanisms of Uniform Nanocrystals via Hot-Injection and Heat-Up Methods. Small 2011, 7, 26852702. 2. Yarema, M.; Pichler, S.; Kriegner, D.; Stangl, J.; Yarema, O.; Kirchschlager, R.; Tollabimazraehno, S.; Humer, M.; Häringer, D.; Kohl, M.; Chen, G.; Heiss, W., From Highly Monodisperse Indium and Indium Tin Colloidal Nanocrystals to Self-Assembled Indium Tin Oxide Nanoelectrodes. ACS Nano 2012, 6, 4113-4121. 3. Hendricks, M. P.; Campos, M. P.; Cleveland, G. T.; Jen-La Plante, I.; Owen, J. S., A Tunable Library of Substituted Thiourea Precursors to Metal Sulfide Nanocrystals. Science 2015, 348, 1226-1230. 4. Cademartiri, L.; Malakooti, R.; O'Brien, P. G.; Migliori, A.; Petrov, S.; Kherani, N. P.; Ozin, G. A., Large-Scale Synthesis of Ultrathin Bi2S3 Necklace Nanowires. Angew. Chem. Int. Ed. 2008, 47, 3814-3817. 5. Nienow, A. W.; Edwards, M. F.; Harnby, N. Mixing in the Process Industries; Butterworth-Heinemann, 1997. 6. Patent WO 2012158832 A2; Method for Preparing Semiconductor Nanocrystals, 2011. 7. Zhang, L.; Xia, Y., Scaling up the Production of Colloidal Nanocrystals: Should We Increase or Decrease the Reaction Volume? Adv. Mater. 2014, 26, 2600-2606. 8. Bozyigit, D.; Yazdani, N.; Yarema, M.; Yarema, O.; Lin, W. M. M.; Volk, S.; Vuttivorakulchai, K.; Luisier, M.; Juranyi, F.; Wood, V., Soft Surfaces of Nanomaterials Enable Strong Phonon Interactions. Nature 2016, 531, 618-622.

9. Hines, M. A.; Scholes, G. D., Colloidal PbS Nanocrystals with SizeTunable Near-Infrared Emission: Observation of Post-Synthesis SelfNarrowing of the Particle Size Distribution. Adv. Mater. 2003, 15, 18441849. 10. Shrestha, A.; Spooner, N. A.; Qiao, S. Z.; Dai, S., Mechanistic Insight into the Nucleation and Growth of Oleic Acid Capped Lead Sulphide Quantum Dots. Phys. Chem. Chem. Phys. 2016, 18, 14055-14062. 11. Choi, H.; Ko, J.-H.; Kim, Y.-H.; Jeong, S., Steric-Hindrance-Driven Shape Transition in PbS Quantum Dots: Understanding Size-Dependent Stability. J. Am. Chem. Soc. 2013, 135, 5278-5281. 12. Kravchyk, K.; Protesescu, L.; Bodnarchuk, M. I.; Krumeich, F.; Yarema, M.; Walter, M.; Guntlin, C.; Kovalenko, M. V., Monodisperse and Inorganically Capped Sn and Sn/SnO2 Nanocrystals for High-Performance Li-Ion Battery Anodes. J. Am. Chem. Soc. 2013, 135, 4199-4202. 13. De Roo, J.; Ibáñez, M.; Geiregat, P.; Nedelcu, G.; Walravens, W.; Maes, J.; Martins, J. C.; Van Driessche, I.; Kovalenko, M. V.; Hens, Z., Highly Dynamic Ligand Binding and Light Absorption Coefficient of Cesium Lead Bromide Perovskite Nanocrystals. ACS Nano 2016, 10, 20712081. 14. Yarema, O.; Bozyigit, D.; Rousseau, I.; Nowack, L.; Yarema, M.; Heiss, W.; Wood, V., Highly Luminescent, Size- and Shape-Tunable Copper Indium Selenide Based Colloidal Nanocrystals. Chem. Mater. 2013, 25, 3753-3757. 15. Weidman, M. C.; Beck, M. E.; Hoffman, R. S.; Prins, F.; Tisdale, W. A., Monodisperse, Air-Stable PbS Nanocrystals via Precursor Stoichiometry Control. ACS Nano 2014, 8, 6363-6371. 16. Cui, J.; Beyler, A. P.; Coropceanu, I.; Cleary, L.; Avila, T. R.; Chen, Y.; Cordero, J. M.; Heathcote, S. L.; Harris, D. K.; Chen, O.; Cao, J.; Bawendi, M. G., Evolution of the Single-Nanocrystal Photoluminescence Linewidth with Size and Shell: Implications for Exciton–Phonon Coupling and the Optimization of Spectral Linewidths. Nano Lett. 2016, 16, 289-296. 17. Cademartiri, L.; Montanari, E.; Calestani, G.; Migliori, A.; Guagliardi, A.; Ozin, G. A., Size-Dependent Extinction Coefficients of PbS Quantum Dots. J. Am. Chem. Soc. 2006, 128, 10337-10346. 18. Protesescu, L.; Yakunin, S.; Bodnarchuk, M. I.; Krieg, F.; Caputo, R.; Hendon, C. H.; Yang, R. X.; Walsh, A.; Kovalenko, M. V., Nanocrystals of Cesium Lead Halide Perovskites (CsPbX3, X = Cl, Br, and I): Novel Optoelectronic Materials Showing Bright Emission with Wide Color Gamut. Nano Lett. 2015, 15, 3692-3696 19. Nere, N. K.; Patwardhan, A. W.; Joshi, J. B., Liquid-Phase Mixing in Stirred Vessels:  Turbulent Flow Regime. Ind. Eng. Chem. Res. 2003, 42, 2661-2698. 20. Patwardhan, A. W.; Joshi, J. B., Relation between Flow Pattern and Blending in Stirred Tanks. Ind. Eng. Chem. Res. 1999, 38, 3131-3143.

ACS Paragon Plus Environment

Chemistry of Materials

Page 8 of 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment

8