Structure–Reactivity Studies, Characterization, and Transformation of

Feb 3, 2017 - ACS Editors' Choice - This is an open access article published under an ACS AuthorChoice License, which permits copying and redistributi...
0 downloads 3 Views 2MB Size
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Article pubs.acs.org/Organometallics

Structure−Reactivity Studies, Characterization, and Transformation of Intermediates by Lithium Chloride in the Direct Insertion of Alkyl and Aryl Iodides to Metallic Zinc Powder Chao Feng, Quinn T. Easter, and Suzanne A. Blum* Department of Chemistry, University of California, Irvine, California 92697-2025, United States S Supporting Information *

ABSTRACT: Employment of fluorophore-tagged alkyl and aryl iodides permitted detection of persistent surface intermediates during their direct insertion to commercially available zinc powder. The sensitivity of this subensemble microscopy technique enabled structure−reactivity studies in the formation of intermediates that are present in quantities sufficiently low as to have been undetected previously by traditional ensemble analytical techniques. These surface intermediates were transformed by lithium chloride, leading to the assignment of the mechanistic role of lithium chloride as changing the rate-determining step in the reaction by lowering the barrier for solubilization of these otherwise persistent surface organometallic intermediates. The temperature dependence/ qualitative barrier of the direct insertion step was determined independently from the solubilization step and from the barrier for the overall reaction. Detection of these zinc surface intermediates at the single-molecule level, i.e., of individual surface organometallic species, has been achieved for the first time. Energy dispersive X-ray spectroscopy (EDS) measurements of the elemental composition of the surface of the zinc powder determined that lithium chloride does not clean the surface of the oxides; instead, pretreatment of the surface with TMSCl effects partial removal of surface oxides after the 2 h pretreatment time previously reported in the empirically optimized synthetic procedure. Current limitations of this microscopy approach are also determined and discussed with respect to the addition of solid reagents during in operando imaging. Characterization of the resulting soluble fluorophore-tagged organozinc/LiCl complex by 1H NMR spectroscopy, mass spectrometry, and fluorescence spectroscopy provided insight into its solution dynamics and chemical exchange processes.



INTRODUCTION The direct insertion of organohalides into commercial metal powders would be the most efficient route for the preparation of several organometallic reagents and catalysts. Due to the recalcitrance of many commercial metal powders toward direct insertion, however, such synthetic reactions are not generally employed across much of the periodic table. The preparation of finely divided metal powders for immediate use by in situ reduction of metal halides enables direct insertion reactions with some metals in exchange for loss of convenience (e.g., with Rieke zinc1 and calcium2). Transformative recent advances by Knochel in the direct insertion of organohalides to commercial zinc,3 manganese,4 aluminum,5 and indium6 powders in the presence of lithium chloride and/or transition-metal catalysts4 have provided functional-group-tolerant access to new reagents, yet the mechanistic roles of these additives in enabling the direct insertion reactions are not well understood.7 We recently reported the detection of intermediates on the surface of zinc in the direct insertion of alkyl iodides to commercial zinc powder8 (eq 1) using a highly sensitive subensemble in operando fluorescence microscopy technique.9−27 These surface intermediates are released into solution upon addition of lithium chloride. This observation led to the assignment of the mechanistic role for lithium chloride as the solubilization of otherwise persistent surface organometallic © 2017 American Chemical Society

intermediates. We herein describe full studies of this system with the goal of understanding the role of lithium chloride. These full studies include identification of a change in ratedetermining step upon addition of lithium chloride, temperature-dependence studies, single-molecule detection of surface organometallic intermediates, and examination of the effect of additives on the zinc surface elemental composition by energy dispersive X-ray spectroscopy (EDS). The sensitivity of the fluorescence microscopy technique toward detection of surface intermediates enabled determination of the relative reaction barriers of the C−I oxidative addition step and the organozinc surface dissociation steps separately. These experiments therefore provided qualitative data on structure−reactivity effects on an intermediate that was present in a quantity sufficiently low as to have been undetected previously. Determination of structure−reactivity relationships Received: December 7, 2016 Published: February 3, 2017 2389

DOI: 10.1021/acs.organomet.6b00910 Organometallics 2017, 36, 2389−2396

Article

Organometallics on the rates of individual steps in multistep reactions is a longstanding challenge in mechanism-based reaction optimization, especially in the field of organometallic chemistry, wherein intermediates often do not build up to the quantities needed for detection by traditional analytical techniques.28 Thus, changes in structure can be difficult to pinpoint to changes in reactivity of specific intermediates rather than their effect on the overall reaction rate.29,30 This microscopy approach therefore has potential for broader application in additional systems by enabling direct observation of previously unobserved intermediates and the subsequent determination of their structure− reactivity relationships.

Probes 2 and 3 were designed such that the green BODIPY fluorophore was sufficiently removed electronically and spatially from the reactive carbon−iodide bond so as to serve as a spectator in the reaction and to avoid potential quenching from the zinc surface and electronic or steric interference with the insertion reaction. As previously reported, reaction of 2 produces surface species oxidative addition product 4,8 the chemical and physical behavior of which is then examined using in operando fluorescence microscopy. Treatment of surface species 4 with lithium chloride leads to transformation into solution-phase material 6, resulting in disappearance of the fluorescent signal from the surface of the zinc. Specifically, examination of the surface of commercial zinc powder after heat and TMSCl treatment as described by Knochel3 and soaking in a solution of fluorophore-tagged probe alkyl iodide 2 for 24 h in the absence of lithium chloride revealed the generation of surface species containing the fluorescent probe (Figure 2). In Figure 2a, the zinc particles appear as dark irregular shapes covered in bright green spots, against a light green background of solution. From the brightness and photophysical behavior of each green spot, it is possible to assign the regions each as containing many molecules rather than single molecules. This reaction was chemoselective for iodide-containing substratesi.e., the surface species did not form in control experiments with a BODIPY probe lacking a carbon−iodide bondwhich led to the assignment of this material as iodide-requiring 4 or 5.8 Structure 4 is the product of oxidative addition/direct insertion into the zinc, and structure 5 is the product of simple Lewis acid/Lewis base coordination of the iodide lone pair electrons to the Lewis acidic zinc.34 Temperature Dependence of Direct Insertion Assists with Assignment of Surface Structure 4. In order to differentiate between structures 4 (oxidative addition of 2 into Zn) and 5 (Lewis acid/Lewis base coordination of 2 to Zn), an aryl iodide probe (3) was next explored and compared in side by side experiments. Probe 3 contained an sp2 carbon−iodide bond, which was expected to have a higher barrier toward oxidative addition than the sp3 carbon−iodide bond of probe 2. Aryl iodide compounds that lack electron-withdrawing groups are reported to be significantly less reactive toward direct insertion of zinc, requiring 50 °C rather than ambient temperature for reaction.3



RESULTS AND DISCUSSION The experimental schematic is shown in Figure 1. We previously examined the reaction of commercial zinc powder

Figure 1. Experiment schematic.

1 with fluorophore-tagged organoiodides 2 and 3 in THF in the presence and absence of lithium chloride.8 Three mechanistic possibilities were considered for the role of lithium chloride in the formation of soluble organozinc reagents from alkyl iodides and zinc powder: (1) lithium chloride cleans oxides from the surface of the zinc powder31 before coordination or reaction of the alkyl iodide (similar to the role of I2 in activating magnesium metal in Grignard reagent formation32); (2) lithium chloride solubilizes surface organozinc reagents after oxidative addition, thus producing the solution-phase reagent and exposing the zinc surface to another molecule of alkyl iodide;3 (3) lithium chloride accelerates oxidative addition, as suggested previously through calculations.33

Figure 2. Structure−reactivity studies: (a) probe 2 does not react at −35 °C; (b) probe 3 requires 60 °C. The temperature dependence on the degree of reactivity is consistent with oxidative addition barriers. 2390

DOI: 10.1021/acs.organomet.6b00910 Organometallics 2017, 36, 2389−2396

Article

Organometallics

This confocal image particularly shows the power of fluorescence microscopy to characterize the surface contours and to map the spatial distribution of fluorescent material onto these contours of the irregular surfaces of commercial zinc powder, thereby highlighting the ability of this microscopy technique to avoid the requisite use of smooth surface and/or ultrapure model systems.35 Such distribution established that not all areas on the zinc surface are equally reactive toward alkyl iodide probe 2. These data lend credence to the hypothesis that intermediates are “poisoning” the most reactive locations on the zinc in ways that prevent the direct insertion from proceeding further, stalling the reaction in the absence of lithium chloride. The nonuniform reactivity may be caused by nonuniform distribution of residual oxides on the surface of the zinc, as revealed in the studies of elemental surface composition by EDS (vide infra), or alternatively by dislocations that provide initiation sites in the zinc surface, as suggested for magnesium from mechanistic studies of the formation of Grignard reagents.35 Assignment of the Surface Species 4 as an Intermediate in the Lithium Chloride Mediated Generation of Solution Organozinc Reagents. We next investigated if the surface material constituted intermediates in the lithium chloride assisted synthesis of soluble organozinc reagents. Studies of salt addition were first disclosed in our original communication:8 Addition of lithium chloride to alkyl 4 resulted in removal of this species from the surface, as seen by the lack of green spots on the surface of the zinc at t = 600 s after addition (Figure 4a). Importantly, control experiments in the absence of LiCl did not result in similar removal of the material from the surface of the zinc, as reported in our initial communication.8 The ability of lithium chloride to transform the surface species is consistent with assignment of these species as intermediates in the lithium chloride assisted generation of soluble organozinc reagents.8 These results established a mechanistic role for lithium chloride in the generation of soluble organozinc reagents: to remove otherwise persistent surface intermediates through solubilization of 4. The reaction conditions produce the expected soluble organozinc−lithium chloride complexes (eq 2, vide infra), further consistent with this assignment. Structure−Reactivity Studies and a Change in RateDetermining Step upon Addition of LiCl. Taken together, the lithium chloride addition studies and the comparison of the oxidative addition temperatures for surface reaction with alkyl iodide 2 and aryl iodide 3 permit mapping of the relative barriers of the oxidative addition step and solubilization step separately (Figure 4c). These separate measurements of single steps were previously obscured by the multiple chemical and physical changes and the previous analytical inability to detect the small quantities of intermediates that occur during the overall synthetic reaction shown in eq 1. For ease of comparison of relative reaction barriers, the energies of the starting materials, intermediates, and products were arbitrarily set as identical in the separate reactions of 2 and 3. As described previously, alkyl and aryl iodides undergo oxidative addition to commercial zinc at 25 and 60 °C, respectively, as seen in Figure 2. In the absence of lithium chloride, the subsequent barriers to solubilization are higher than either of the barriers toward oxidative addition, leading to rate-determining solubilization and persistent intermediate 4. In the presence of lithium chloride, however, intermediate 4

Thus, if oxidative addition accounted for the intermediates on the surface of the zinc, then employment of probe 3 should result in no (or less) product as seen by no (or fewer) green fluorescent spots on the surface of the zinc at ambient temperature. Probe 3, however, contains iodide nonbonding electrons like those in probe 2. Thus, if simple coordination accounted for the species on the surface of the zinc, employment of probe 3 should result in similar levels of product and thus similar levels of bright green signal on the surface of the zinc particles and would indicate the presence of Lewis acid/base coordination product 5. In a set of experiments at 25 °C reported in our initial communication,8 comparison of both probes at 25 °C clearly showed the absence of surface reaction with the aryl iodide probe 3 in contrast to the high levels of bright green signal on the surface of the zinc with probe 2. This result was consistent with the bright green spots arising from oxidative addition. Thus, the surface species were assigned as direct insertion product 4 rather than Lewis acid/ base coordination product 5, consistent with hypothesis 2. We now report that temperature-dependence studies are further consistent with the assignment of the surface material as oxidative addition intermediate 4. Specifically, reduction of the temperature to −35 °C prevented reactivity with alkyl probe 2. Imaging data in Figure 2a from this experiment showed a complete lack of detectable bright green spots on the surface of the zinc; instead, the zinc particles remained dark. Lewis acid/ Lewis base complexation, in contrast, which avoids breaking the C−I bond, should have a lower barrier than oxidative addition and would plausibly not display such marked temperature dependence. Similarly, increasing the temperature to 60 °C was sufficient to overcome a higher reaction barrier and lead to reactivity of aryl probe 3, as seen by the presence of bright green spots on the surface of the zinc after reaction at this higher temperature (Figure 2b). This observation is consistent with the aforementioned report of generation of the soluble arylzinc species at 50 °C rather than ambient temperature3 and with the insertion of 3 into Zn. Nonuniform Spatial Distribution of Surface Intermediates. The spatial distribution of 4 derived from probes 2 and 3 on the surface of the zinc particles was clearly mapped by wide-field epifluorescence microscopy (Figure 2). Many individual bright green spots and regions that are dense with organic material, with lengths of ∼0.5−4 μm, are distributed on the surface of each particle. Employment of confocal microscopy with z-axis scanning confirmed that the intermediate was nonuniformly distributed across the zinc surface in all three dimensions. Figure 3 shows a confocal image of one zinc particle of dimensions ca. 15 × 5 × 5 μm3 treated with 2.

Figure 3. Confocal microscopy image of one zinc particle imaged with z-axis scanning, showing 3D distribution of surface intermediate 4. The particle size is ca. 15 × 5 × 5 μm3. 2391

DOI: 10.1021/acs.organomet.6b00910 Organometallics 2017, 36, 2389−2396

Article

Organometallics

quantity of 4 remained on the surface that the fluorescence signal from individual species could be resolved. Specifically, 1 h after addition of lithium chloride to a sample, the surfaces of the zinc particles were reexamined. Under the original imaging settings, no bright spots were observable after this extended time. The imaging settings were then changed to higher laser excitation power to allow for the detection of faint emitters, as would arise from single molecules rather than collections of many molecules. This change in imaging conditions revealed a small number of molecules remaining on the surface of the zinc particles, a subset of which were detected as diffraction-limited green spots. The signal from some molecules continued to overlap, producing larger areas of fluorescence. The intensity vs time traces of some of these spots showed the quantized photophysical behavior that is the wellestablished fluorescent fingerprint of single molecules.36 Figure 5a shows a zinc particle after 1 h; an expansion of the region in

Figure 4. (a) Addition of LiCl to a Zn sample in THF previously prepared from alkyl iodide 2 at ambient temperature (red). Negative times indicate the time before addition of the salt. (b) Addition of LiCl to a Zn sample in THF previously prepared from aryl iodide 3 at 60 °C (black). (c) Relative barriers of direct insertion and solubilization showing the change in rate-determining step in the presence and absence of LiCl.

derived from both probes is rapidly removed from the surface at 25 °C through a low-barrier solubilization step (Figure 4a,b), and oxidative addition becomes rate determining. The current experiments do not address the relative barriers of solublization of alkyl-derived 4 and aryl-derived 4, only that these barriers are both lower than the corresponding oxidative addition reactions. Solubilization in the presence of LiCl is therefore after the ratedetermining step. Because the organoiodides in the previously optimized synthetic system react in the presence of lithium chloride,3 the different observed overall reaction rates for sp3 and sp2 organoiodides is hereby pinpointed solely to the different rates of oxidative addition of the substrates rather than different rates of solubilization of intermediates. Establishing the Absence of Significant Fluorescence Quenching upon Salt Addition. Previously published control experiments8 established that the lithium chloride additive and the reaction byproducts zinc chloride and zinc iodide do not quench the fluorescence of the BODIPY probe. Thus, the disappearance of a signal from the surface of the zinc was attributed to removal of the fluorescent species and not to fluorescence quenching. Single-Molecule Detection of Intermediate 4. Consistent with the assignment that the disappearance of a signal is due to gradual removal of fluorophore-tagged probe molecules from the surface of the zinc particles, after 1 h a small enough

Figure 5. Single-molecule detection of surface species 4: (a) zinc surface with 4, t = 1 h after addition of lithium chloride; (b) fluorescence intensity vs time graph showing four stepwise events (red arrows) that characterize the photobleaching or solubilization of four separate individual molecules of the probe.

the orange box is displayed with increased contrast, showing multiple areas of the surface that contain a small number of surface species 4. A second expansion of this region is also displayed of the area in the red box; individual species of 4 can be detected in this 1.6 × 1.9 μm2 region on the basis of their quantized events. For example, Figure 5b shows the fluorescence intensity time trace of this red-boxed region on the surface, wherein four single molecules are detected as quantized, stepwise fluorescence decreases (red arrows) that correspond to photobleaching or solubilization of these individual complexes. At t = 1 h, molecules of 4 remained at some surface locations but not others. The distribution of the remaining 4 reflects the initial spatial reactivity distribution of 2 with zinc overlaid with the spatial distribution of the insertion and/or solubilization chemistry in the presence of lithium chloride. Challenges and Limitations of Full Time Course Studies. These experiments are an early demonstration of in operando imaging in organic solvents under air-free conditions 2392

DOI: 10.1021/acs.organomet.6b00910 Organometallics 2017, 36, 2389−2396

Article

Organometallics

hypothesis 1). The effect of a second additive, TMSCl, was also examined. In the synthetic procedure, the zinc particles were treated for 2 h with 0.1 mol % of TMSCl prior to addition of the alkyl iodide substrate.3 In contrast to the lithium chloride treatment, treatment with TMSCl affected a small but statistically significant decrease in the surface oxygen composition as seen at 30 min and at 2 h (p < 0.05) on addition at 30 mol %, suggesting that a smaller effect may occur in the synthetic system at 0.1 mol %. After 2 h, detectible oxide remained at some locations on the surface but not at about half the examined locations (i.e., the remaining oxide was below the detection limit of the instrument at 5 of 11 locations with 100% Zn and 0% O), in comparison to 2 of 11 locations before treatment (see the Supporting Information for tabulated data). The heterogeneous oxide distribution that persists despite treatment with either reagent may contribute along with defect distributions35 to the observed nonuniform distribution of oxidative addition on the zinc surface by alkyl and aryl iodide probes 2 and 3 in Figures 2−5; the imaging data in these figures were acquired after a 2 h treatment with TMSCl, identical with that in the published synthetic procedure, and prior to treatment with LiCl. Taken together, the data from these studies suggest the possibility of improving the efficiency of the synthetic reaction via increasing the initial effective concentration of the reactive zinc surface for rate-determining oxidative addition by better removal of the surface oxide layer before exposure to the organohalide. Bench-Scale Characterization of Soluble Organozinc/ Lithium Chloride Complex 6. Addition of 2 to zinc powder in a bench-scale reaction in THF-d8 produced a green-yellow fluorescent solution; after 1 h the solution had turned slightly brownish yellow, consistent with a change in the fluorescence properties of the solution upon organozinc formation (Figure 6 and eq 2).

and as such required concurrent development of gloveboxcompatible air-free microscopy chambers12 and reagent addition techniques suitable for addition of solid salts to the zinc powder at the microscope. In this vein, we also determined current challenges of the imaging technique as they relate to the addition of solid reagents to the sample under imaging conditions. An ideal experiment would allow detection of the same zinc particles before and after addition of reagents, such that the effect of reagents on specific surface species could be probed throughout the course of the change in physical or chemical properties. Such imaging was not possible in our system, since the addition of reagents and even flow of blank solvent resulted in lateral and rotational motion of the particles in the study, which were swept out of the field of view. On account of this practical challenge, solid salts were measured into separate glass containers inside the glovebox and carefully placed inside the imaging chamber, which was then capped and removed from the glovebox, creating a fully sealed vessel suitable for both imaging and subsequent salt addition. This setup allowed the reaction to be performed fully under rigorously air free conditions as required to prevent reagent quenching.3 Initial images of the sample were obtained from t = −40 to 0 s. At t = 0 s, the sample was removed from the microscope objective, inverted, and shaken to dispense and dissolve the salt into the THF solution without the need to open the vial. The imaging chamber was then re-placed onto the microscope objective, the image was refocused, and the time course of the chemical reaction was monitored. This addition/shaking/refocus process required approximately 80 s; thus, data could reliably be obtained from 100 s onward. Heterogeneity of Surface Oxide Composition and Effect of LiCl and TMSCl Additives. The observation that lithium chloride transformed the surface intermediates per hypothesis 2 did not rule out that it may also have additional mechanistic roles. To examine if lithium chloride was capable of removing oxides from the surface (hypothesis 1), the elemental composition of the surface was examined via EDS before and after the addition of lithium chloride. Specifically, the molar ratio of zinc to oxygen was examined at 11 different locations spanning measurements at multiple locations on the surfaces of two or three particles per sample. The zinc particles, from the same supplier and the same mesh and heat-treated as in the reported synthetic procedure from Knochel,3 contained 92.5 ± 6.2 mol % of Zn and 7.5 ± 6.2 mol % of O (Table 1). Thus, the

Figure 6. Formation of mixed organozinc 6 and mass spectrometry detection of the diiodo and chloroiodo zincate compounds from that mixture.

Table 1. Effect of Additives on Surface Oxide Composition amt (mol %) Zn O

as received

LiCl 2 h

TMSCl 30 min

TMSCl 2 h

92.5 ± 6.2 7.5 ± 6.2

92.2 ± 4.9 7.8 ± 4.9

95.7 ± 2.6 4.3 ± 2.6

96.6 ± 3.7 3.4 ± 3.7

surface of these particles was composed of significant quantities of oxide. The high standard deviation reflected the substantial variation in surface oxide quantities at different locations on the same particle and between different particles, establishing the high heterogeneity of the surface composition of the commercial sample of zinc powder. Addition of lithium chloride in THF followed by a 2 h soaking time did not result in a measurable decrease in the oxide quantity (Table 1), ruling out an additional role for lithium chloride in also cleaning the surface (i.e., ruling out

1

H NMR spectroscopy analysis of this mixture showed complete consumption of 2 and generation of product 6 (Figure 7). The 1H NMR spectrum of 6 was broad, consistent with exchange processes. The substantial upfield shift of δ 3.1 ppm of the diagnostic α-I protons from 3.3 ppm in 2 to α-Zn protons at 0.2 ppm confirmed the successful direct insertion reaction. Electrospray ionization mass spectrometry (ESI-MS) analysis in negative ionizion mode detected diiodides 6a,b on 2393

DOI: 10.1021/acs.organomet.6b00910 Organometallics 2017, 36, 2389−2396

Article

Organometallics

LiCl). In the presence of lithium chloride, the difference in temperature required in the synthetic systems to generate organozinc reagents from alkyl iodides (ambient temperature) and aryl iodides (50 °C)3 therefore arises from the differences in barriers in the oxidative addition step and not from differences in the barriers in the solubilization step. These assignments provide a mechanistic understanding that underpins the empirically discovered lithium chloride assisted direct insertion into commercial metal powders. The broader significance of this technique is the ability to detect and determine the reactivity of intermediates that are present in quantity significantly lower than that needed for traditional ensemble analytical techniques. It thereby enables structure−reactivity and mechanistic determination on individual isolated steps in multistep reactions with such intermediates3thus addressing a longstanding analytical challenge8,11 in mechanism-based reaction design and synthetic method improvement.

Figure 7. 1H NMR spectra of key regions of probe 2 and product 6.



the basis of their mass and distinctive isotope patterns. The anions [ZnI3]−, [ZnI2Cl]−, and [ZnICl2]− were also detected. Detection of mixed organozinc halides and zinc trihalides is consistent with Koszinowski’s prior report of these anionic species in the analogous reactions of butyl iodide with zinc powder in the presence of lithium chloride.7 A Schlenk-type equilibrium containing R2ZnXLi may also be present, as reported by Koszinowski for similar zinc complexes7 and by Stalke for magnesium amide complexes in the presence of lithium chloride,37 although such species were not detected in our sample by MS. These spectral data lead to the assignment of 6 as mixed organozinc complexes of structures 6a,b, wherein the lithium counterion might be coordinated to the two bridging halides and THF solvent while in solution. These complexes may bear structural similarity to those of THFsoluble TMP-active Hauser and turbo-Hauser bases,38 as well as to other Zn-containing reagents,39 with bridging halides. The fluorescence emission maximum was not significantly shifted upon formation of the organozinc product (2, λem 510 nm; 6, λem 506 nm), consistent with the design of the spectator probe such that the BODIPY fluorophore was not in direct conjugation with the zinc center. Consistent with the lower fluorescent appearance of the solution, the quantum yield of 6 (Φ = 0.52) was less than that of starting iodide 2 (Φ = 1.0). Thus, the chemical change at the distal position does affect the quantum yield despite the electronic separation; the photophysical mechanism behind this reduction is not yet known. The quantum yield of this solution species, however, remains sufficiently high for fluorescence microscopy imaging applications.

EXPERIMENTAL SECTION

General Methods. All manipulations were carried out under a nitrogen atmosphere in dried glassware unless otherwise noted. All chemicals were used as received from commercial sources unless otherwise noted. THF was dried by passage through an alumina column under argon pressure on a Seca Solvent System (Glass Contour). Zinc powder (99.9%) was purchased from Strem and dried in vacuo while heat was applied from a heat gun for ca. 30 min. Trimethylsilyl chloride was purified by stirring over CaH2 for 24 h and was then vacuum-transferred. 3,4-Dichlorobenzaldehyde was purified by recrystallization from ethanol. Chlorodiphenylphosphine was purchased from Sigma-Aldrich and distilled under reduced pressure. Flash chromatography was conducted using a Teledyne Isco Combiflash Rf 200 Automated Flash Chromatography System. Microscopy and Image Acquisition. Imaging was performed with an IX71 inverted microscope (Olympus Corp.) and an oilimmersion objective with a 1.49 numerical aperture. Samples were illuminated with the 488 nm line of an Ar/Kr ion laser (Coherent Inc.) set to 25 mW. Illumination was done under conditions of EPI. Samples were imaged with a C9100-13 electron multiplier CCD camera (Hamamatsu Photonics). The CCD chip was a back-thinned electron multiplication type with an effective 512 × 512 array of pixels. Synthesis and characterization of probes 2 and 3 were performed as previously reported.8 Construction of Reaction Cells. Reaction cells were constructed from a 1 dram vial by cutting the bottom of the vials and adhering them to a prepared glass cover slip with Devcon 5 min Epoxy. The epoxy was allowed to cure for 1 h. The glass reaction cells were then dried under dynamic vacuum for 12 h before being brought into the glovebox. Reaction of Zinc with Alkyl Iodide Probe 2. In a glovebox filled with nitrogen, a 1 dram vial was charged with zinc (50 mg, 0.76 mmol), THF (1.0 mL), and trimethylsilyl chloride (1 drop). The mixture was then capped and agitated gently to mix the solution. After the mixture was soaked at room temperature for 2 h, the supernatant was removed and the resulting zinc particles were washed with THF (2 × 2 mL). To the pretreated zinc particles was added a solution of 2 (2.0 mM of 2 dissolved in THF, 20 drops), the solution was capped, and the vial was swirled gently to mix the solution and set aside for soaking. After the mixture was soaked at room temperature for 24 h, the supernatant was removed from the mixture and the resulting zinc particles were washed with THF (3 × 2 mL). After the third rinse THF (2.0 mL) was added to the vial, the vial was swirled vigorously to suspend the zinc particles in solution, and the resulting slurry was partitioned into three aliquots containing THF and zinc and then put into separate reaction cells. In salt addition experiments, lithium chloride (8.2 mg, 0.19 mmol) was added to a glass salt pocket; this salt pocket was gently placed inside the microscope reaction vial. The cells



CONCLUSION Subensemble fluorescence microscopy experiments with up to single-molecule sensitivity permitted interrogation of the mechanistic role of lithium chloride in the direct insertion of alkyl and aryl iodides into commercial zinc powder under synthetically relevant conditions. The significance of the measurement in this system is the assignment of the mechanistic role of lithium chloride as changing the ratedetermining step in the direct insertion of aryl and alkyl iodides to commercial zinc powder. Specifically, lithium chloride lowers the barrier for solubilization of otherwise persistent alkyl and aryl zinc intermediates on the surface of the zinc. Through this process, the rate-determining step shifts from solubilization (in the absence of LiCl) to oxidative addition (in the presence of 2394

DOI: 10.1021/acs.organomet.6b00910 Organometallics 2017, 36, 2389−2396

Article

Organometallics were then capped and carefully removed without shaking from the glovebox for imaging. The cells were imaged for 11 min each. Zinc particles with fluorescence signals were found and kept in focus. After the reaction cell was imaged for 40 s and the surface-bound intermediate was observed, the salt was added by inverting the cell and gently shaking the cell; this was repeated three times. After the cell was placed back onto the microscope, zinc particles were found and brought into focus at 100 s after the salt addition and these particles were kept in focus until 540 s after the addition, after which the stage was moved to find new zinc particles that had not yet been exposed to laser illumination at 600 s. Reaction of Zinc with Aryl Iodide Probe 3. In a nitrogen-filled glovebox, zinc particles (77.4 mg, 1.18 mmol) were weighed out into a 4 mL scintillation vial. In this vial was placed THF (1.0 mL) via syringe, followed by 1 drop of TMSCl (∼0.4 mmol, 30 mol %). The vial was capped, and the solution was allowed to stand for 2 h. Following the 2 h period, the THF was removed from the vial via syringe. The particles were washed with 2 × 2.0 mL THF portions added via syringe. In a separate 4 mL vial, the aryl iodide BODIPY (1.0 mg, 0.0022 mmol) was weighed out and dissolved in 1.0 mL of THF. This afforded a solution of 2.2 mM BODIPY, 30 drops of which were then added to the Zn particles. The vial was capped and heated to 60 °C for 24 h. After 24 h, the residual solution was removed, and the Zn particles were washed with 3 × 2.0 mL portions of THF added via syringe. The particles were agitated with 3.0 mL of THF, and one-third of this agitated solution was placed in a microscope reaction vial. In salt addition experiments, lithium chloride (8.2 mg, 0.19 mmol) was added to a glass salt pocket; this salt pocket was gently placed inside the microscope reaction vial. The vial was capped, removed from the glovebox carefully without shaking, and taken directly to the microscope. The cells were imaged for 11 min each. Zinc particles with fluorescence signals were found and kept in focus. After the reaction cell was imaged for 40 s and the surface bound intermediate was observed, the salt was added by inverting the cell and gently shaking the cell; this was repeated three times. After the cell was placed back onto the microscope, zinc particles were found and brought into focus at 100 s after the salt addition and these particles were kept in focus until 540 s after the addition, after which the stage was moved to find new zinc particles that had not yet been exposed to laser illumination at 600 s. Single-Molecule Detection. Zinc particles were prepared according to the experimental procedure for reaction of zinc with alkyl iodide probe 2: zinc particles (42 mg, 0.64 mmol) were soaked in a solution of 2 (1.0 mL, 2.0 mM in THF) over 24 h, and then LiCl (8.0 mg, 0.19 mmol) was added through a glass salt pocket inside the vial. After treatment with LiCl for 1 h at room temperature, the zinc particles were placed on a microscope for single-molecule imaging. The laser power was raised to approximately 30 times that of the original imaging conditions, and images were obtained with a 100 ms/ frame frame rate, 0 s frame duration time, and 141 camera intensification. Under these conditions of higher laser power, single molecules could be detected remaining on the surface of the zinc. Under conditions of the original lower laser power, the sample was completely dark. Surface Elemental Composition Determination via EDS. Heat-Treated3 but Otherwise Used As Received from Manufacturer. In a nitrogen-filled glovebox, Zn was placed in a 4 mL vial. This vial was capped and brought out of the glovebox for imaging. To prepare the sample for imaging, the Zn was scraped out of the vial using a spatula and placed on carbon tape mounted on a metal stand. Treatment with LiCl. In a nitrogen-filled glovebox, heat-treated Zn3 (74.8 mg, 1.14 mmol) was weighed out into a 4 mL vial. THF (1.0 mL) was added to the vial via syringe. In a separate 4 mL vial, LiCl (8.2 mg, 0.19 mmol) was weighed out and placed in the Zn/THF vial with a spatula. The vial was capped and agitated, and the solution was allowed to stand for 2 h. After the 2 h period, the THF was syringed out of the vial, and the particles were washed with 2 × 2.0 mL portions of THF added via syringe. The vial was then capped and taken out of

the glovebox for imaging. To prepare the sample for imaging, the Zn particles were scraped out of the vial using a spatula and placed on carbon tape mounted on a metal stand. Treatment with TMSCl. In a nitrogen-filled glovebox, heat-treated Zn3 (79.1 mg, 1.21 mmol) was weighed out into a 4 mL vial. THF (1.0 mL) was added to the vial via syringe, followed by 1 drop of TMSCl. The vial was capped, and the solution was allowed to stand for 30 min or 2 h. Following this 30 min or 2 h period, THF was syringed out of the vial, and the particles were washed with 2 × 2.0 mL THF portions added via syringe. The vial was capped and taken out of the glovebox for imaging. To prepare the sample for imaging, the Zn particles were scraped out of the vial using a spatula and placed on carbon tape mounted on a metal stand. Scanning electron microscopy (SEM) was accomplished by using a FEI Quanta 3D FEG Dual Beam (FEI Co.) in the Laboratory for Electron and X-ray Instrumentation (LEXI) at the University of California, Irvine. Samples were imaged on carbon tape at 1.0 kV energy, 43 pA current, and pressures lower than 8 × 10−5 mbar using the xTm software (FEI Co.) with XJ Charts (XJ Technologies). Images were collected at varying magnifications with a 1024 × 943 pixel resolution. Energy-dispersive X-ray spectroscopy (EDS) was performed on the FEI Quanta 3D using the Oxford EDS Detector for Quanta 3D (Oxford Instruments) in the LEXI facilities. Samples were imaged as prepared with 20.0 kV energy and 43 pA current at pressures below 8 × 10−5 mbar. Samples were viewed using the xTm software, and after the images were selected, data were collected and interpreted using the INCA software (v. 4.15, Oxford Instruments Analytical Limited). The Zn Kα lines were measured at 8.60 and 9.60 keV, the Zn Lα line was measured at 1.1 keV, and the O Kα line was measured at 0.62 keV.40 Synthesis of Soluble Organozinc−Lithium Chloride Compound 6. To 5,5-difluoro-10-(4-iodobutyl)-1,3,7,9-tetramethyl-5Hdipyrrolo[1,2-c:2′,1′-f ][1,3,2]diazaborinin-4-ium-5-uide (2; 15.6 mg, 0.0363 mmol) in dry d8-THF (2 mL) in a 1 dram vial were added LiCl (2.3 mg, 0.054 mmol) and zinc powder (4.7 mg, 0.073 mmol) with a stir bar, inside a glovebox, producing a fluorescent green solution with suspended gray zinc powder. The vial was capped, and the resulting reaction mixture was stirred at room temperature for 1 h, at which time the clear brownish yellow fluorescent solution above the reaction mixture was removed by pipet and transferred to a J. Young NMR tube. The tube was capped and removed from the glovebox. The 1H NMR spectrum was then acquired, showing broad peaks consistent with equilibrating species and showing complete consumption of starting material 2. The sample for MS-ESI analysis was aquired from a separate reaction mixture similarly prepared. An aliquot of the solution was removed and injected into the instrument directly via syringe to prevent quenching by atmospheric water. MS: [C17H22BClF2IN2Zn]−, m/z 529; [C17H22BF2I2N2Zn]−, m/z 621. 1H NMR spectrsocopic analysis of this second sample showed a mixture of protodemetalated material and organozinc 6. Quantum Yield and Emission Spectrum of Organozinc Compound 6 in THF. To 5,5-difluoro-10-(4-iodobutyl)-1,3,7,9tetramethyl-5H-dipyrrolo[1,2-c:2′,1′-f ][1,3,2]diazaborinin-4-ium-5uide (2; 13.8 mg, 0.0321 mmol) in dry THF (2 mL) were added LiCl (1.4 mg, 0.032 mmol), zinc powder (4.2 mg, 0.064 mmol), and a stir bar, inside the glovebox. The resulting reaction mixture was stirred at room temperature for 1 h inside the glovebox, at which time 0.2 mL of the clear solution was taken out of the glovebox by a syringe for GCMS analysis, to confirm full conversion of 2 into organozinc species 6 by detection of clean formation of protodemetalated material8 once the sample was removed from the glovebox and quenched by atmospheric water. In the glovebox, the remaining clear solution above the reaction mixture was then removed by pipet and diluted with dry THF in a 1 dram vial, which was used as the stock solution for measuring the quantum yield of organozinc species. For each concentration measurement, different quantities of drops of the stock were taken by pipet and diluted with dry THF in a cuvette fitted with a glass adapter for measuring under air-free conditions. The cuvette was capped and removed from the glovebox for quantum yield 2395

DOI: 10.1021/acs.organomet.6b00910 Organometallics 2017, 36, 2389−2396

Article

Organometallics

Mandelshtam, V. A.; Blum, S. A. J. Am. Chem. Soc. 2010, 132, 15167−15169. (17) Shen, H.; Zhou, X.; Zou, N.; Chen, P. J. Phys. Chem. C 2014, 118, 26902−26911. (18) Esfandiari, N. M.; Wang, Y.; Bass, J. Y.; Blum, S. A. Inorg. Chem. 2011, 50, 9201−9203. (19) Kubarev, A. V.; Janssen, K. P. F.; Roeffaers, M. B. ChemCatChem 2015, 7, 3646−3650. (20) Sambur, J. B.; Chen, T. Y.; Choudhary, E.; Chen, G. Q.; Nissen, E. J.; Thomas, E. M.; Zou, N. M.; Chen, P. Nature 2016, 530, 77−80. (21) Esfandiari, N. M.; Wang, Y.; McIntire, T. M.; Blum, S. A. Organometallics 2011, 30, 2901−2907. (22) Decan, M. R.; Impellizzeri, S.; Marin, M. L.; Scaiano, J. C. Nat. Commun. 2014, DOI: 10.1038/ncomms5612. (23) Sambur, J. B.; Chen, P. J. Phys. Chem. C 2016, 120, 20668− 20676. (24) Ristanović, Z.; Kerssens, M. M.; Kubarev, A. V.; Hendriks, F. C.; Dedecker, P.; Hofkens, J.; Roeffaers, M. B. J.; Weckhuysen, B. M. Angew. Chem., Int. Ed. 2015, 54, 1836−1840. (25) Ristanović, Z.; Hofmann, J. P.; De Cremer, G.; Kubarev, A. V.; Tohnke, M.; Meirer, F.; Hofkens, J.; Roeffaers, M. B. J.; Weckhuysen, B. M. J. Am. Chem. Soc. 2015, 137, 6559−6568. (26) Wang, N.; Tachikawa, T.; Majima, T. Chem. Sci. 2011, 2, 891− 900. (27) Easter, Q. T.; Trauschke, V.; Blum, S. A. ACS Catal. 2015, 5, 2290−2295. (28) For an example, see: Romero, P. E.; Piers, W. E. J. Am. Chem. Soc. 2005, 127, 5032−5033. (29) van der Eide, E. F.; Piers, W. E. Nat. Chem. 2010, 2, 571−576. (30) Hevia, E.; Mulvey, R. E. Angew. Chem., Int. Ed. 2011, 50, 6448− 6450. (31) Porter, F. C. Corrosion Resistance of Zinc and Zinc Alloys; CRC Press: Boca Raton, FL, 1994; p 121. (32) Eckert, T. S. J. Chem. Educ. 1987, 64, 179. (33) Liu, C.-Y.; Wang, X.; Furuyama, T.; Yasuike, S.; Muranaka, A.; Morokuma, K.; Uchiyama, M. Chem. - Eur. J. 2010, 16, 1780−1784. (34) Stair, P. C. J. Am. Chem. Soc. 1982, 104, 4044−4052. (35) Hill, C. L.; Sande, J. B. V.; Whitesides, G. M. J. Org. Chem. 1980, 45, 1020−1028. (36) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.; Springer: New York, 2006. (37) Neufeld, R.; Teuteberg, T. L.; Herbst-Irmer, R.; Mata, R. A.; Stalke, D. J. Am. Chem. Soc. 2016, 138, 4796−4806. (38) García-Á lvarez, P.; Graham, D. V.; Hevia, E.; Kennedy, A. R.; Klett, J.; Mulvey, R. E.; O’Hara, C. T.; Weatherstone, S. Angew. Chem., Int. Ed. 2008, 47, 8079−8081. (39) Haag, B.; Mosrin, M.; Ila, H.; Malakhov, V.; Knochel, P. Angew. Chem., Int. Ed. 2011, 50, 9794−9824. (40) Deslattes, R. D.; Kessler, E. G., Jr.; Indelicato, P.; de Billy, L.; Lindroth, E.; Anton, J. Rev. Mod. Phys. 2003, 75, 35−99. (41) Brouwer, A. M. Pure Appl. Chem. 2011, 83, 2213−2228.

and emission spectrum measurements. Rhodamine 6G in EtOH (Φ = 0.95)41 and fluorescein in 0.10 M NaOH(aq) (Φ = 0.93)41 were used as standards. Φ = 0.52; λem 506 nm.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.organomet.6b00910. Additional experimental details and imaging data from triplicate lithium chloride addition runs (PDF)



AUTHOR INFORMATION

Corresponding Author

*E-mail for S.A.B.: [email protected]. ORCID

Suzanne A. Blum: 0000-0002-8055-1405 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The bulk of this work was supported by the National Science Foundation (CHE-1464959). We thank the U.S. Department of Energy, Office of Basic Energy Sciences (DE-FG0208ER15994), for support of initial experiments, and the University of California-Irvine and the Alexander von Humboldt Foundation for funding. SEM and EDS work were performed at the UC Irvine Materials Research Institute (IRMI), using instrumentation funded in part by the National Science Foundation Center for Chemistry at the Space-Time Limit (CHE-082913). We thank Prof. Dr. Paul Knochel for helpful discussions. We thank Mr. Drew W. Cunningham for coacquiring the image in Figure 3.



REFERENCES

(1) Rieke, R. D.; Li, P. T.; Burns, T. P.; Uhm, S. T. J. Org. Chem. 1981, 46, 4323−4324. (2) Wu, T.-C.; Xiong, H.; Rieke, R. D. J. Org. Chem. 1990, 55, 5045− 5051. (3) Krasovskiy, A.; Malakohov, V.; Gavryushin, A.; Knochel, P. Angew. Chem., Int. Ed. 2006, 45, 6040−6044. (4) Peng, Z.; Knochel, P. Org. Lett. 2011, 13, 3198−3201. (5) Blümke, T.; Chen, Y.-H.; Peng, Z.; Knochel, P. Nat. Chem. 2010, 2, 313−318. (6) Shen, Z.-L.; Knochel, P. Chem. - Eur. J. 2015, 21, 7061−7065. (7) Koszinowski, K.; Boehrer, P. Organometallics 2009, 28, 771−779. (8) Feng, C.; Cunningham, D. W.; Easter, Q. T.; Blum, S. A. J. Am. Chem. Soc. 2016, 138, 11156−11159. (9) Hensle, E. M.; Blum, S. A. J. Am. Chem. Soc. 2013, 135, 12324− 12328. (10) Esfandiari, N. M.; Blum, S. A. J. Am. Chem. Soc. 2011, 133, 18145−18147. (11) Ng, J. D.; Upadhyay, S. P.; Marquard, A. N.; Lupo, K. M.; Hinton, D. A.; Padilla, N. A.; Bates, D. M.; Goldsmith, R. H. J. Am. Chem. Soc. 2016, 138, 3876−3883. (12) Cordes, T.; Blum, S. A. Nat. Chem. 2013, 5, 993−999. (13) Blum, S. A. Phys. Chem. Chem. Phys. 2014, 16, 16333−16339. (14) Chen, P.; Zhou, X.; Shen, H.; Andoy, N. M.; Choudhary, E.; Han, K.-S.; Liu, G.; Meng, W. Chem. Soc. Rev. 2010, 39, 4560−4570. (15) Roeffaers, M. B.J.; Sels, B. F.; Uji-i, H.; De Schryver, F. C.; Jacobs, P. A.; De Vos, D. E.; Hofkens, J. Nature 2006, 439, 572−575. (16) Esfandiari, N. M.; Wang, Y.; Bass, J. Y.; Cornell, T. P.; Otte, D. A. L.; Cheng, M. H.; Hemminger, J. C.; McIntire, T. M.; 2396

DOI: 10.1021/acs.organomet.6b00910 Organometallics 2017, 36, 2389−2396