Optical Trapping of a Single Molecule of Length Sub-1 nm in Solution

Open AccessCCS ChemistryCOMMUNICATIONS13 Dec 2022Optical Trapping of a Single Molecule of Length Sub-1 nm in Solution Biao-Feng Zeng†, Ran Deng†, Yu-Ling Zou†, Chun-An Huo, Jing-Yu Wang, Wei-Ming Yang, Qing-Man Liang, Sheng-Jie Qiu, Anni Feng, Jia Shi, Wenjing Hong, Zhilin Yang, Zhong-Qun Tian and Yang Yang Biao-Feng Zeng† Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 , Ran Deng† Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 , Yu-Ling Zou† Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 , Chun-An Huo Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 , Jing-Yu Wang Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 , Wei-Ming Yang Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 , Qing-Man Liang Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 , Sheng-Jie Qiu Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 , Anni Feng Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 , Jia Shi Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 , Wenjing Hong Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 , Zhilin Yang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 , Zhong-Qun Tian *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 and Yang Yang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Department of Physics, Pen-Tung Sah Institute of Micro-Nano Science and Technology, State Key Laboratory of Physical Chemistry of Solid Surfaces, IKKEM, Xiamen University, Xiamen 361005 https://doi.org/10.31635/ccschem.022.202202318 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Plasmonic optical manipulation has emerged as an affordable alternative to manipulate single chemical and biological molecules in nanoscience. Although the theoretical models of sub-5 nm single-molecule trapping have been considered promising, the experimental strategies remain a challenge due to the Brownian motions and weak optical gradient forces with significantly reduced molecular polarizability. Herein, we address direct trapping and in situ sensing of single molecules with unprecedented size, down to ∼5 Å in solution, by employing an adjustable plasmonic optical nanogap and single-molecule conductance measurement. The theoretical simulations demonstrate that local fields with a high enhancement factor, over 103, were generated at such small nanogaps, resulting in optical forces as large as several piconewtons to suppress the Brownian motion and trap a molecule of length sub-1 nm. This work demonstrates a strategy for directly manipulating the small molecule units, promising a vast multitude of applications in chemical, biological, and materials sciences at the single-molecule level. Download figure Download PowerPoint Introduction The ability to control nanoscale objects with nanometric precision is essential for various disciplines in nanoscience like chemistry,1 biological science,2 physics,3 and materials science.4 Several methods, for instance, scanning probe microscopy,5 dielectrophoresis,6,7 magnetic tweezers,8 and optical tweezers,9 have been developed to immobilize, capture, and control the motion of nanoscale objects. Among these methods, optical trapping techniques, with the advantages of remote, targeted, and parallel processing capabilities, have been extensively employed to trap objects from microns to hundreds of nanometers.10 However, further scaling of the optical trapping volume is stunted by the diffraction limit of laser beams and the ever-present Brownian motions for conventional optical tweezers.2 Several strategies have been proposed to trap the molecules of sub-100 nm size. For instance, employing the microparticle as a handle to attach to the individual molecules11,12 and enhancing the optical force by plasmonic field.13–16 For the former, manipulating individual molecules relies on the interactions between molecules and microparticles, which requires that the captured molecule be functionalized and limits applications. The latter, known as plasmon-enhanced optical trapping, is capable of confining and enhancing an electromagnetic field to the nanoscale hot spots, thus directly trapping and manipulating the nanoparticles with a laser of low intensity.13,17–22 However, to date, plasmon-enhanced optical trapping to achieve the direct trapping of single molecules with a size smaller than 10 nm remains a great challenge because of the two following issues. One is the lack of a method capable of precisely adjusting a nanostructure to gain sufficient electromagnetic field.19,23–27 The other is the lack of a method capable of simultaneously characterizing the movement of a single molecule of length sub-10 nm.28,29 Recently, investigators into dynamic nanostructures have reported that controllable metallic configurations enable the generation of a hot spot with the remarkable enhancement of the electromagnetic field. In particular, the adjustable nanogap electrodes pair in a mechanically controllable break junction (MCBJ) platform30,31 and the nanogap tip-substrate pair in a tip-enhanced Raman scattering platform were demonstrated to capture and detect 2 nm quantum dots and single molecules.32–35 Our recent work displayed the capability of synchronously trapping and sensing small molecules down to 2 nm using the single-molecule plasmonic optical trapping (POT) method, which combined the single-molecule MCBJ method and POT.36 This single-molecule technique provides promising paths toward trapping the smaller single molecule. However, the trapping of a single molecule of a length below 2 nm has still not been achieved. Herein, we present a quantitative study of the evolution of optical force for a series of sub-2 nm molecules by using the modified plasmon-enhanced break junction (PEBJ) method. The formation probabilities of molecular junctions were utilized to identify and quantify the optical trapping process. Three molecules, 4,4″-1 diamine-1,1′:4′,1″-terphenyl (TBDA), [1,1′-Biphenyl]-4,4′-diamine (DBDA), and 1,4-Benzenediamine (BDA), with molecular lengths of 14.30, 10.04, and 5.70 Å ( Supporting Information Figure S1), respectively, were employed as the probe molecules. Analysis of the conductance plateaus' length showed that the formation probabilities of molecular junctions significantly increased under focused 671 nm light illumination (30 mW, ∼5 × 107 W/m2) for all three probe molecules. Considering the length of the BDA molecule, these experimental findings indicate that our approach enables us to trap a single molecule with a size of less than 1 nm. Theoretical simulations revealed that the single-molecule trapping originated from the plasmon-enhanced optical force as large as 12.119, 7.355, and 5.897 pN for TBDA, DBDA, and BDA monomer, respectively, which had suppressed the Brownian motions. Based on the experimental results and theoretical simulations, we finally propose a single exponential model to decipher the trapping probability contributed by the synergistic effect of electromagnetic field enhancement and molecular polarizability. Results and Discussion Experimental setup Figure 1a,b shows the schematic of the modified PEBJ instrument utilized in our experiment. The core components are the laser irradiation system and MCBJ setup. Benefiting from the tailored three-pivot architecture, PEBJ provides an elegant approach to generating an adjustable nanogap between the electrode pairs. During the experiment, the pushing rod underneath the chip moved upward and downward. With such an operation, the size of the nanogap between the two gold nanoelectrodes was adjusted with a subangstrom resolution.37–40 Then, a 671 nm laser was introduced into the home-built PEBJ setup ( Supporting Information Figure S2). This excited the surface plasmons on the nanogold electrode surfaces, which further generated an enhanced electromagnetic field in the nanogap. In addition to the laser diode driver, the optical system consisted of a microscope objective (4×, 0.1 numerical aperture, 37.5 mm work distance), a beam splitter, a polarizing prism, and a charge-coupled device (CCD) camera. The linear polarized incident light illuminated with the parallel direction of the gold nanoelectrode through the polarizing prism for maximum electric field enhancement.36,41 The PEBJ-compatible chip was fabricated by the notched-wires method, as shown in Supporting Information Figure S3a,b. The CCD camera was used to ensure that the laser spot was focused on the nanoelectrodes, as shown in Supporting Information Figure S3c. If the molecules in the solution were trapped in the nanogap by the optical trapping force under the laser irradiation, there would be an increase in the formation probability of the molecular junctions. Thus, the trapping behavior was detected by means of quantifying the formation probability of molecular junction. Figure 1 | (a) Schematic of the home-built PEBJ setup with 671 nm laser for single-molecule trapping. The right panel displays the magnified view of the electrodes and the process of single-molecule optical trapping. (b) The chemical structures of three probe molecules are termed TBDA, DBDA, and BDA. Download figure Download PowerPoint Single-molecule conductance measurement without light First, we carried out the single-molecule conductance measurement using our home-built PEBJ device to show that our strategy is capable of detecting small molecules with single-molecule sensitivity.38,42 The single-molecule optical trapping events were sensed by monitoring changes in the formation probability of molecular junction, which reflects the evolution of the local concentration of molecules in the nanogap.43 Since the electric field44,45 and concentrations43 made significant impacts on the formation probability of molecular junctions, the single-molecule conductance measurements were performed in decane with a fixed bias of 100 mV at room temperature. Decane is a stable nonpolar solvent that offers a wide access window for conductance measurement (1 G0 ∼ 10−6 G0, Supporting Information Figure S4) and avoids the influence of solvent volatilization on molecular concentration under laser irradiation. The saturated molecular solutions were prepared and then diluted by a factor of 100 as the initial concentration for conductance measurement. Figure 2a,b shows the experimental results of single-molecule conductance measurement for TBDA, DBDA, and BDA without laser illumination. As shown in Figure 2a, the individual traces of conductance versus displacement distance (Δz) are presented on a semilogarithmic scale. If no molecule bridges the nanotips (black curves), the conductance traces dramatically decay to the background level from the conductance quantum G0 (G0 = 2e2/h ≈ 77.48 μS). While in the presence of molecules, the conductance plateaus appear and show the possible distribution of molecular junctions. Thousands of individual conductance traces were collected without any data selection for further statistical analysis. The data analysis was accomplished in the open-source code XME analysis ( https://github.com/Pilab-XMU/XMe_DataAnalysis). All the probe molecules showed distinct conductance with different values. We fit the one-dimensional (1D) conductance histograms of three probe molecules with Gaussian function to get the most representative conductance values. The most probable conductance states of TBDA and DBDA are 10−3.8 G0 and 10−2.96 G0, respectively. In the case of BDA, there are two conductance states located at 10−2.22 G0 (high conductance state, BDA monomer) and 10−4.0 G0 (low conductance state, BDA dimer), as shown in Figure 2b. We further analyzed the plateau length distributions of TBDA, DBDA, and BDA, which are located at 0.94, 0.65, and 0.22 nm, respectively (Figure 2b, inset). After the calibration of the 0.5 nm snap-back distance,46,47 the plateau lengths of TBDA, DBDA, and BDA are 1.44, 1.15, and 0.72 nm. The experimental plateau lengths of TBDA and DBDA agree well with the length of the theoretical calculation. Notably, the plateau length of BDA was larger than the monomer theoretical length, suggesting the formation of BDA dimer molecular junctions. This hypothesis is discussed in detail in section on the density functional theory (DFT) calculation below. Figure 2 | Single-molecule conductance measurements. (a) Typical individual conductance-distance traces in single-molecule conductance measurement, corresponding to pure solvent (black), TBDA (blue), DBDA (orange), and BDA (red for BDA monomer and pink for BDA dimer). (b) The 1D conductance histograms of TBDA (blue), DBDA (orange), and BDA (pink). The inset shows the distributions of the plateau lengths. (c) Theoretical calculations of the optimized molecular junction models for TBDA, DBDA, BDA dimer-T, and BDA monomer. (d) Calculated transmission functions versus electron energy for TBDA (blue), DBDA (orange), BDA dimer-T (pink), and BDA monomer (red). Download figure Download PowerPoint To understand the configuration of molecular junctions between three probe molecules and gold electrode tips, we calculated the transmission T(E) curves of the TBDA, DBDA, and BDA molecular junctions. The DFT and nonequilibrium Green's function with the Atomistix ToolKit package were used for the calculations. The configurations of the constructed molecular junctions were optimized, and the obtained molecular junction models were displayed in Figure 2c. For the TBDA, DBDA, and BDA monomers, the gap distances of the optimized junction models were calculated to be 1.04, 0.77, and 0.58 nm, respectively. We noticed that the gap distances of TBDA and DBDA molecular junctions were smaller than the theoretical lengths of the three molecules, due to the atomic migration on the electrode surfaces and sophisticated contact configuration between the molecules and nanoelectrodes under a strong electric field (109 V/m).48 The corresponding transmission curves (Figure 2d) show that the obtained conductance trend is consistent with the experiments: TBDA < DBDA < BDA monomer in the vicinity of the Fermi level. For BDA specifically, the single-molecule conductance measurement showed two conductance states, implying the multiple configurations of BDA molecular junctions. Previous work often attributed the high conductance state of such conjugated molecules to the molecular monomer while attributing the low conductance state to the π–π stacking dimers.43,49 However, there are various types of stacking configurations for the BDA dimer, each of which has a different size ( Supporting Information Figure S5), in particular, sandwich stacking configuration (BDA dimer-S), parallel-displaced π–π stacking configuration (BDA dimer-P), and T-shaped π–π stacking configuration (BDA dimer-T).50 To identify the configuration of the BDA dimer that accounts for the low conductance measured in our experiment, we calculated the transmission curves for all three configurations of the stacking dimers ( Supporting Information Figure S6). After configuration optimization, the BDA dimer-T molecular junction showed a tightly packed configuration with a gap distance of 0.77 nm among the three stacking configurations, indicating the most stable configuration.50 In the experiments, the plateau length of such a tightly packed configuration was more likely to overlap with the BDA monomer, as shown in Figure 2b. Besides, we found that the face-to-face stacking configurations (BDA dimer-S and BDA dimer-P) exhibited a high-conductance state while the point-to-face stacking configuration (BDA dimer-T) exhibited a low-conductance state, which was in accordance with the experimental results (Figure 2b, BDA dimer). Based on the above discussion, we attributed the observed low conductance state of BDA molecular junctions to BDA dimer-T configuration. Plasmonic optical manipulation of single molecules To perform the single-molecule POT experiment, the gold electrode pair was irradiated by the focused 671 nm laser to excite the surface plasmons. The absorption spectra of the three probe molecules range from 200 to 500 nm51–54 ( Supporting Information Figure S7), indicating that they are nonphotoresponsive to the 671 nm laser. It was thus expected that the electronic configuration of the probe molecules would not change when exposed to the 671 nm laser, and the molecular conductance would stay the same as it was when measured in the dark. We carried out the conductance measurements for these molecules with and without illumination. The peak position in the 1D conductance histograms remained constant in the control experiments ( Supporting Information Figure S8), confirming that the 671 nm laser did not change the molecular structures. We further plotted the most probable conductance with the molecular length for the three probe molecules, in the presence and absence of light, as shown in Supporting Information Figure S9. The decay of conductance of both conditions almost coincides and is consistent with the previous work.55 Figure 3a–f shows the two-dimensional (2D) conductance-distance histograms of TBDA, DBDA, and BDA collected with and without illumination. The 2D conductance-distance histograms exhibited clear conductance plateaus corresponding to the conductance peaks. Furthermore, the intensity of the conductance clouds in 2D conductance-distance histograms obtained with illumination had become more pronounced than those without illumination, which indicated an increase in the formation probabilities for the constructed molecular junctions. Figure 3 | Experimental results of single-molecule POT. The 2D conductance-distance histograms for TBDA (a, b), DBDA (c, d), and BDA (e, f) were measured with and without 671 nm laser illumination. Download figure Download PowerPoint To quantitatively analyze the evolution of the formation probabilities of the molecular junctions, we utilized the slope-based method to extract the features of molecular conductance traces and automatically count the number of curves with molecular signals,56 as shown in Figure 4a and Supporting Information Tables S1 and S2. Without illumination, the formation probabilities of TBDA and DBDA molecular junctions are approximately 35.8% and 40.2%, respectively. When exposed to a 671 nm laser, the formation probabilities of TBDA and DBDA molecular junctions have been significantly increased to 64.8% and 54.9%, respectively. For BDA molecular junctions, the formation probability of a high conductance state showed a slight increase, from 12.8% to 16.3%. While the formation probability of the low conductance state increased from 58.9% to 68.2%, more than twice that of the high conductance state in increment. Figure 4b shows that the values of trapping probability decrease with the decreased size of chemical molecules. This is rational because the trapping potential scales down with the volume of the particle that relates to the molecular polarizability in the Rayleigh limit.57 Thus the optical trapping gets more challenging as the molecular size decreases. However, the smaller captured molecules were accompanied by smaller gaps between nanoelectrodes, leading to a larger electromagnetic field in the hot spot, which in turn was more potent for POT. Thereby, to evaluate the trapping ability in the experiments, both the enhancement of the electromagnetic field and the reduction in molecular sizes need to be considered. Figure 4 | Quantitative analysis of the single-molecule POT. (a) The formation probabilities of three targeted molecular junctions with and without illumination. BDA monomer and BDA dimer-T represent the high conductance state and low conductance state of the BDA molecular junctions. (b) The single exponential model of single-molecule POT behavior. The curve reveals experimental trapping probability as a function of the theoretical length of the molecule, which is well fitted with a single exponential function. Download figure Download PowerPoint The single exponential model of single-molecule POT behavior To determine the single-molecule trapping behavior, we further established a mathematical model based on experimental and theoretical results. In theory, the trapping behavior can be evaluated by the trapping potential U = −0.5 α |E|2 (Formula 1), and the trapping probability ΔP is exponential to the depth of trapping potential (ΔP ∝ exp(−U/kBT) (Formula 2), where α is the polarizability of the particle, E is the local electric field, kB is Boltzmann's constant, and T is temperature).57 In the case of consistent test conditions, the trapping probability was determined by the molecular polarizability and the local electric field. For a microsphere particle, the optical polarizability depends on the volumes and scales with the third power of the particle radius. When the material size is reduced to the single-molecule scale, the spherical volume formula does not apply to the calculation of molecular volume. The single molecules were considered as pointlike particles with isotropic complex polarizability.15 However, this assumption did not apply to the single-molecule conductance measurement because of the applied bias and metal-molecule binding configuration. The molecular polarization was significantly influenced by the external electric field, molecule-metal contact geometry, as well as the interaction between molecules.35,58,59 Generally, the small chemical molecules were regarded as nanowires when bridging the nanoelectrodes.60 In the following treatment, we considered the nonpolar single molecule as a 1D nanowire with a length of LM. The molecules were polarized in the direction of the electric field generated by the externally applied bias, then the molecular polarizability was proportional to the molecular length, α ∝ LM. When molecules bridge the nanoelectrode pairs, the molecular length determines the gap distances, LM∝ dgap. In previous reports, Novotny et al.14 and Xu et al.15,61 proposed that in the extremely small nanogap, U ∝ (|E|2/|E0|2)−2 ∝ dgap−2, where E is the local electric field mentioned in the above formula, E0 is the incident field, and |E|2/|E0|2 is enhancement factor. In the case of the molecular junction, with the combined Formulas 1 and 2, we expect there is a relation ΔP ∝ exp((aLM−2 + bLM + c)/kBT) by assuming α ∝ LM, U ∝ (|E|2/|E0|2)−2 ∝ dgap−2 ∝ LM, where a, b, and c are constants. We defined the difference in the formation probabilities with and without light irradiation as the trapping probability ΔP and employed the theoretical length of the molecules calculated by the DFT method. As shown in Figure 4b, the exponential function fits the experimental data well with degrees of confidence R2 of about 0.999 ( Supporting Information Table S3). The excellent fitting of the experimental data to our model supports the capability of the PEBJ method to trap single molecules. Theoretical simulation of optical force and trapping potential To evaluate the trapping capability of the dynamic nanoelectrode pairs, we performed a detailed numerical analysis of plasmonic properties, the contribution of optical force, and trapping potential by considering the far field and the near field. In simulations, a normal incident plane wave with polarization along the nanoelectrode long axis was used as the light source. The parameters utilized in the finite-element simulation are shown in Supporting Information Table S4. We first simulated the extinction spectra of the coupled nanoelectrodes with 1.04, 0.77, and 0.58 nm gap distances that were obtained from geometric optimization of the junction models by DFT calculation (Figure 2c). Figure 5a shows peaks of extinction efficiency for these three gap sizes located at around 632 and 671 nm, corresponding to the two plasmonic resonance wavelengths. Under the excitation at these two wavelengths, the electromagnetic field of 671 nm excitation wavelength showed higher field enhancement by up to three orders of magnitude. Figure 5b represents the simulated electromagnetic field contour for the 0.58 nm nanogap, which is the case of the BDA monomer with a local field enhancement of up to 6000. The field enhancements approach 5000 and 3000 for the gap distances of 0.77 and 1.04 nm, respectively ( Supporting Information Figure S10). These results indicate that smaller gap distances lead to a higher field enhancement.30,31 We also considered the photothermal effect induced by the focused laser and found that the temperature change of the hot-spot region in the solution was less than 0.7 K under illumination, which is negligible ( Supporting Information Figure S11). Figure 5 | Theoretical simulations. (a) Extinction spectra of the Au nanoelectrode pairs with gap sizes of 0.58, 0.77, and 1.04 nm. The values of the gap sizes were obtained from the configuration optimization for three types of molecular junctions by using DFT calculation. (b) Spatial distribution of the electric field enhancement for the Au nanoelectrode pairs with the gap distance of 0.58 nm located at the x–y plane. (c) The optical trapping potential (U) exerted on the TBDA, DBDA, BDA dimer-T, and BDA monomer in units of kBT under 671 nm laser illumination. (d) Spatial distribution of the optical trapping potential exerted on BDA monomer under 671 nm laser illumination. (e) The optical force exerted on TBDA, DBDA, BDA dimer-T, and BDA monomer in the z-direction under 671 nm laser illumination. (f) Mapping the optical force vector exerted on BDA monomer in the x–y plane under 671 nm laser illumination. The intensity is normalized to show the complete vector flow. Download figure Download PowerPoint Subsequently, we calculated the trapping potential and optical force exerted on the probe molecules bridged across the plasmonic nanogap. The trapping potential in the nanogap can be described as U = −0.5αE2, where α is the electrodynamic dipolar polarizability of the molecule and E is the electric field.12 The formula shows that the key factor affecting the trapping potential includes the molecular polarizability and the field enhancement factor. Her