Metallic Degenerately Doped Free-Electron-Confined Plasmonic Nanocrystal and Infrared Extinction Response

Abstract

In this paper, synthetically scaled-up degenerately n-type doped indium tin oxide (Sn:In₂O₃) nanocrystals are described as highly transparent conductive materials possessing both optoelectronic and crystalline properties. With tin dopants serving as n-type semiconductor materials, they can generate free-electron carriers. These free electrons, vibrating in resonance with infrared radiation, induce strong localized surface plasmon resonance (LSPR), resulting in efficient infrared absorption. To commercialize products featuring Sn:In₂O₃ with localized surface plasmon resonance, a scaled-up synthetic process is essential. To reduce the cost of raw materials during synthesis, we aim to proceed with synthesis in a large reactor using industrial raw materials. Sn:In₂O₃ can be formulated into ink dispersed in solvents. Infrared-absorbing ink formulations can capitalize on their infrared absorption properties to render opaque in the infrared spectrum while remaining transparent in the visible light spectrum. The ink can serve as a security ink material visible only through infrared cameras and as a paint absorbing infrared light. We verified the transparency and infrared absorption properties of the ink produced in this study, demonstrating consistent characteristics in scaled-up synthesis. Due to potential applications requiring infrared absorption properties, it holds significant promise as a robust platform material in various fields.

1. Introduction

The synthesis of doped metal oxide nanocrystals (NCs), such as indium tin oxide (Sn:In₂O₃), begins with colloidal synthesis methods [1,2]. This technique aims to synthesize NCs capable of absorbing light in the infrared (IR) range through localized surface plasmon resonance (LSPR). Sn:In₂O₃ is known for its electrical conductivity and visible light transparency due to free-electron carriers delocalized within the material bulk lattice. The introduction of Sn(ac)₄, a substitution dopant, into the indium oleate grating n-type aliovalent induces free carrier density in the concentration range ~10²⁰ cm⁻³. The free-electron carrier is fixed inside the NC core. With nano-sizes smaller than the incidence of the electromagnetic wavelengths of infrared rays, the free-electron carrier can vibrate in the NC core and absorb light [3]. Under nanoscale material sizes, Sn:In₂O₃ can be used in various fields as semiconductor nanocrystals (NCs) [4,5,6]. It can be deposited as thin films on various substrates, such as transparent electrodes for displays, thin-film transistors, and smart windows [7,8]. Transparent inks can be produced utilizing the infrared absorption characteristics of Sn:In₂O₃ [9,10]. Furthermore, engineering scale-up synthesis enables marketability and cost-effectiveness, allowing for larger quantities compared to lab-scale production [11]. Even with scale-up synthesis, infrared absorption properties through LSPR can be maintained consistently. We were able to assess transparency in the visible light spectrum and light absorption in the infrared spectrum using both a regular camera and an infrared camera. Additionally, we verified the efficiency of the infrared ink and paint in absorbing light in the infrared region [12,13].

We developed plasmonic nanocrystals using degenerately n-type doped metal oxides acting as metallic free-electron carriers, confined coherently oscillating within the NC lattice matrix akin to classical gold or silver nanoparticles [14]. This current study was chemically scaled up 20 times compared to previous research [15]. This presents an advantage in terms of process feasibility for industrialization. Materials such as gold and silver, which are rare commodity elements, come with high unit costs, making them less applicable for large-scale or real-world use [14]. However, metal oxide Sn:In₂O₃ is significantly cheaper compared to gold and silver [16]. Moreover, as a metal oxide, Sn:In₂O₃ is in an oxidized state, resulting in high stability against oxidation and sulfur poisoning compared to other metallic based plasmonic nanoparticles materials. The high stability reduces chemical processing complexity, thereby increasing the economic feasibility in scale-up and process design [17,18,19].

2. Materials and Methods

Indium (III) acetate (In(ac)₃, 99.99%) was purchased from Uniam (Tokyo, Japan), and tin (IV) acetate (Sn(ac)₄) was purchased from Sigma-Aldrich (St. Louis, MO, USA). Oleic acid (OA, 90%, technical grade), Oleyl alcohol (OlAl, 85%, technical grade), Hexane (95.0%), and isopropyl alcohol (99.5%) were purchased from Daejung Chemicals (Siheung-si, Republic of Korea). All chemicals were used as is without extra purification.

Injection Precursor Synthesis: A scale-up colloidal synthesis was performed with a modified synthesis procedure from Sn:In₂O₃ NCs synthesis [20]. To prepare the tin (Sn)-doped Sn:In₂O₃ NC injection precursor solution, Sn(ac)₄ (1.77 g), In(ac)₃ (27.73 g), and oleic acid (200 mL) were transferred to a large 1 L volume 3-neck flask. Synthesis was conducted with a 5% at. Sn dopant concentration. Colloidal NC synthesis precursor was put into the flask with a magnetic stirring bar and connected with a glass condenser on the main neck and a Schlenk line vacuum and N₂ gas manifold. The thermocouple was attached to the flask neck, and the rest of the neck was sealed with a rubber stopper. When the precursor was ready to be reacted in the flask, the precursor mixture was degassed under vacuum. The precursor mixture stirring speed was set to 600 RPM, and after reaching 120 °C, the flask was degassed for 15 min. The flask was ventilated from vacuum to inert N₂ gas. The finished injection precursor solution was divided into several 19-gauge needle glass syringes under an inert N₂ gas flow at a temperature of 120 °C.

ITO (Sn:In₂O₃) Scale-Up NCs Synthesis: When synthesizing Sn:In₂O₃ NCs, continuous injection colloidal synthesis was conducted for NC growth. Oleyl alcohol (260 mL) and a magnetic stirring bar were placed in another 1 L large 3-neck flask. The thermocouple and glass condenser were connected to the flask and sealed with a glass stopper. The flask was heated toward 230 °C under vacuum, and the stirring speed was set to 600 RPM. When the reaction solution reached 120 °C, gas flow was switched to inert N₂ gas, and the prepared precursor solution in a glass syringe was ready to be injected. Oleyl alcohol was injected into the flask at a rate of 4.0 mL/min, while maintaining a reaction vessel temperature of 230 °C. Once the injection with one syringe was completed, it was immediately replaced with the next syringe for consecutive injections. After all injections were completed, the reaction flask was cooled by removing the heating mantle and allowing air to enter until the solution temperature reached approximately 60 °C. Once the solution was cooled, the reaction mixture was transferred and divided into 50 mL centrifuge tubes in 15 mL volume portions. Hexane was added to the centrifuge tube to disperse the total solution volume to 30 mL each.

Purification of ITO (Sn:In₂O₃) Scale-Up NCs: In the purification step of Sn:In₂O₃ NCs, isopropyl alcohol (IPA) antisolvent was added to the solution dispersed in hexane in a ratio of hexane to IPA of 2:1 by volume. The solution was centrifuged at 3600 RPM for 3 min, and the upper organic solution supernatant was discarded. The solid NC pellet remaining at the bottom of the centrifuge tube was redispersed in 20 mL hexane using a vortex mixer. After redispersion, 10 mL of IPA antisolvent was added to the mixture until the solution became opaque, precipitating the NCs. The solution was centrifuged again at 3600 RPM for 3 min. This washing process was repeated three times, followed by discarding the upper organic solution and adding 20 mL hexane to the remaining solid NC pellet at the bottom of the centrifuge tube, which was redispersed using a vortex mixer. The redispersed NCs were centrifuged at 2000 RPM for 3 min to remove any undispersed aggregates, and the supernatant containing the NCs sample was collected.

ITO (Sn:In₂O₃) Ink Production: The supernatant containing the collected NC samples was collected in glass vials. Additional hexane was added to dilute and redisperse the desired concentration of 10 mg/mL. The redispersed solution at the required concentration was the finished ink product.

SEM Sample Preparation: The concentration of Sn:In₂O₃ NCs was measured by weight. A glass slide was placed on an electronic balance. A total of 200 μL of synthesized NC solution dispersed in hexane was drop-cast onto a glass slide. The hexane solvent was completely dried, and the remaining solid NCs were measured. Hexane was added to dilute the solution until it reached a concentration of 0.1 mg/mL. A 50 μL aliquot of the diluted solution of NCs at 0.1 mg/mL was drop-cast onto a silicon wafer SEM grid piece measuring 1 cm × 1 cm. A microscopy image was imaged using a Hitachi SU8230 (Chiyoda City, Japan) scanning electron microscope (SEM) at a 15.0 kV accelerating voltage, ×150 k magnification. An additional infrared analysis was conducted through Nicolet Summit FTIR (Green Bay, WI, USA) transmission mode, and a Widy SenS 320V-ST InGaAs SWIR (Prague, Czech Republic) infrared camera.

X-ray Diffraction Analysis: The sample preparation involved drop-casting a solution of scaled-up Sn:In₂O₃ NC infrared ink dispersed in hexane at a concentration of 30 mg/mL onto Si substrates, followed by complete solvent evaporation. Data were collected using a Rigaku (Tokyo, Japan) MiniFlex 600 X-ray diffractometer. The X-ray source used was Cu Kα radiation with an input voltage of 50 kV and power of 1.0 kW.

3. Results and Discussion

Scaled-up Sn:In₂O₃ NC infrared ink functionality is enabled through the introduction of an n-type tin dopant under colloidal synthetic conditions. Sn dopants are incorporated into the In₂O₃ lattice, substituting In atomic sites as degenerately doped ${\mathrm{Sn}}_{\mathrm{In}}^{•}$. Free-electron carriers are generated within the NC lattice following the Kröger–Vink notation [2,21]. The intentional doping of n-type Sn dopant induces strong LSPR optical functionality in the near-infrared spectrum. Scaled-up Sn:In₂O₃ NCs similarly exhibited powerful LSPR optical functionality [22].

$${\left(2{\mathrm{Sn}}_{\mathrm{In}} · {\mathrm{O}}_{\mathrm{i}}\right)}^{\times }\to 2{\mathrm{Sn}}_{\mathrm{In}}^{•}+{\displaystyle \frac{1}{2}}{\mathrm{O}}_{2 \left(\mathrm{g}\right)}+2{\mathrm{e}}^{\prime }$$

Lab-scale Sn:In₂O₃ NCs were synthesized in a 50 mL 3-neck flask with the resulting solution adjusted to a concentration of 22 mg/mL with the addition of a hexane solvent (Figure 1a, left). The scaled-up Sn:In₂O₃ NCs were obtained and were then transferred to a 1 L total volume container. Hexane solvent was added to adjust the concentration to 30 mg/mL (Figure 1a, right). The synthesis process of Sn:In₂O₃ involves reaction with oleic acid with indium (III) acetate (In(Ac)₃) to produce indium oleate (In(Oleate)₃) and then form indium oxide. At this time, indium oxide was doped with tin to synthesize Sn:In₂O₃. Indium acetate and oleic acid were coordinated to form indium oleate precursor, and when injected into oleyl alcohol using the continuous injection method, it undergoes NC growth through the condensation reaction of indium oleate and oleyl alcohol [23]. At this time, all other conditions are the same, only the injection rate is changed to 0.2 mL/min for synthesis (Figure 1b, left). A continuous precursor injection batch reactor setup for the scaled-up synthetic reaction is shown (Figure 1b, right).

NCs were observed using SEM imaging, and in a scaled-up synthetic batch, well-defined NC particle appearance was observed (Figure 2a). The Sn:In₂O₃ NC solution spectrum was obtained and measured via FTIR, complemented with a Drude model analysis using MATLAB (2023b) fitting [24]. The experimentally collected scaled-up LSPR peak value (2537 cm⁻¹) and MATLAB fitted value (2586 cm⁻¹) well matches within the infrared absorption region of interest. The experimentally collected lab-scale LSPR peak value (1951 cm⁻¹) and MATLAB fitted value (1962 cm⁻¹) well matches within the infrared absorption region of interest. The LSPR extinction resonance frequency is dependent on the dopant activation of the Sn dopant precursor. As the n-type dopant activation concentration within the confines of the NC lattice increases, the free-carrier electron density increases and moves to higher frequencies. Due to a higher reactor volume in the scaled-up reactor, dopant activation may be varied due to degassing synthetic conditions and scalability, but the intrinsic LSPR absorption property is still realized. Therefore, scaled-up Sn:In₂O₃ NC inks at a dopant precursor concentration of 5% are observed in the higher frequency region due to higher population of free carriers within the NCs. And variations may be tamed through dopant activation control through precursor concentration fine-tuning. Because LSPR is an interfacial and dielectric phenomenon, the LSPR extinction response can be simulated and fitted through the Drude model (Figure 2b) [25]. Fourier-transform infrared spectroscopy (FTIR) was used to measure infrared absorption over a wider wavelength range. FTIR was employed to analytically characterize the infrared extinction LSPR response by extracting optoelectronic material property parameters [26].

$${ϵ}_p={ϵ}_{\infty }-{\displaystyle \frac{{\omega }_p^2}{{\omega }^2+i\omega \gamma }}$$

The Drude model was used to extract the concentration of degenerately doped free-electron carriers confined within Sn:In₂O₃ NCs [27,28]. ${ϵ}_p$ represents the optical extinction characteristics of plasmonic metal oxides in the equation above, and $\omega$ is a function of the wavenumber. The Drude model is operated by an external incident electromagnetic field beyond the infrared range. It is treated as a vibration of a free-electron gas [27]. The position of the extinction peak depends on the plasma frequency ${\omega }_p$ according to the concentration of free electrons $n_e$. The absorption coefficient α is a function of the wavenumber ω. It is determined by the imaginary loss coefficient part of ${ϵ}_p=n-ki$. The relationship between the absorption and imaginary loss coefficients is α = 4πk/λ. The wavelength λ can be converted to the wavenumber 1/ω (cm⁻¹) [14,27,28]

$${\omega }_p=\sqrt{{\displaystyle \frac{n_e{e}^2}{{\epsilon }_0{m}^{*}}}}$$

The bulk plasma frequency is determined by the influence of ${\omega }_p$. The free-electron concentration is $n_e$, and the constant e is the fundamental electron charge. ${\epsilon }_0$ is the permittivity of free space, and m* is the electron effective mass. The damping parameter $\gamma$ is a function of frequency-dependent $\omega$. An extended Drude model is thus employed, where ${\gamma }_L$ is the low-frequency damping constant, ${\gamma }_H$ the high-frequency damping constant, ${\gamma }_X$ the crossover frequency, and ${\gamma }_W$ the cross-over width.

$$\gamma \left(\omega \right)={\gamma }_L-{\displaystyle \frac{{\gamma }_L-{\gamma }_H}{\pi }}\left[ta{n}^{-1}\left({\displaystyle \frac{\omega -{\gamma }_X}{{\gamma }_W}}\right)+{\displaystyle \frac{\pi }{2}}\right]$$

We estimated the relationship between the observed LSPR peak and the free carrier density confined within the NCs (

Table 1. Drude model LSPR properties.

Properties	Scale Up Parameter	Lab Scale Parameter
LSPR Peak [cm−1]	2537	1962
Free Carrier Density $n_e$ [cm−3]	2.53 × 1020	1.44 × 1020
Bulk Plasma Frequency ${\omega }_p$ [cm−1]	7525	5670
Low-Frequency Damping Constant ${\gamma }_L$ [cm−1]	1569	1582
High-Frequency Damping Constant ${\gamma }_H$ [cm−1]	360	86
Crossover Frequency ${\gamma }_X$ [cm−1]	4447	3507
Crossover Width ${\gamma }_W$ [cm−1]	277	311

) [29]. The free carrier density of Sn-doped scaled-up Sn:In₂O₃ NCs was estimated to be 2.53 × 10²⁰ cm⁻³. And the free carrier density of Sn-doped lab-scale Sn:In₂O₃ NCs was estimated to be 1.44 × 10²⁰ cm⁻³. The free carrier density of doped Sn:In₂O₃ NCs varies according to the degree of the doping of the Sn. Scaled-up Sn:In₂O₃ NCs were confirmed to be metallically degenerately doped and incorporated into the NC lattice adequately. Even with the scaled-up synthetic procedure, the material characteristics in infrared absorption through the LSPR effect is observed [30]. FTIR measurements were characterized in a mid-infrared transparent solvent. The selected dielectric medium in the FTIR liquid cell is suitable for dispersion and has a dielectric constant of 1.505, which is tetrachloroethylene (TCE) [31]. Spectral fitting using the Drude model via MATLAB was conducted under solution dispersion in this dielectric medium solution. The XRD pattern of scaled-up Sn:In₂O₃ NC infrared ink shows peaks corresponding to (222), (400), (440), and (662) bixbyite phase indium oxide, confirming nanoparticle crystallinity after synthesis (Figure 3) [31,32].

Scaled-up Sn:In₂O₃ NCs exhibited optical properties similar to those at lab scale in previous studies demonstrating a free carrier density range of 10²⁰ cm⁻³ [33]. Free electrons plasmonically resonate with infrared waves, resulting in intense mid-infrared LSPR absorption. A dispersion solution of Sn:In₂O₃ NCs in hexane at a concentration of 10 mg/mL visually appears blue, attributed to the infrared absorption tail absorbing the red visible light spectrum (Figure 4a, left). The identical Sn:In₂O₃ NCs solution observation was made using a near-infrared InGaAs camera covering the range from 900 nm to 1700 nm, exhibiting strong opacity to infrared (Figure 4a, right) [34]. We further utilized Sn:In₂O₃ NC ink in a colorant surfactant application by painting a square shape on a 130 cm × 97 cm art canvas surface by brush. With the naked eye, in the visible spectrum, no apparent marking was observed on the canvas surface (Figure 4b, left). With the same canvas surface when photographed with an infrared camera, the painted square pattern surface area is optically discrete and discernable over a wide surface area (Figure 4b, right) [35].

4. Conclusions

The successful scale-up of Sn-doped Sn:In₂O₃ NC infrared ink functionality was achieved through the introduction of n-type tin dopants under controlled large-batch colloidal synthetic conditions. The Sn dopants are incorporated into the In₂O₃ lattice, substituting for In atomic sites as degenerately doped ${\mathrm{Sn}}_{\mathrm{In}}^{•}$ generating free-electron carriers within the NC lattice, procreating metallic degenerately doped NCs. This doping process induces strong LSPR optical functionality in the near-infrared spectrum, which is maintained in scaled-up batches of Sn NCs. Expanded Sn:In₂O₃ NCs showed infrared LSPR activity, observed in laboratory-scale synthesis study procedures. With a near-infrared InGaAs camera, we were able to see the state of dispersion in hexane and the absorption of infrared rays after painting on an art canvas.

The optical properties of scaled-up Sn:In₂O₃ NCs with an estimated free carrier density of 2.53 × 10²⁰ cm⁻¹ mirror those observed in a lab-scale synthetic sample. This consistency demonstrates that the scale-up process does not compromise the material optical performance as a metallic degenerately doped plasmonic nanocrystal. The NC ink as a dilute thin film is visibly transparent, while the concentrated NC solution dispersed in hexane exhibits a unique blue color due to visible red wavelength absorption. Strong opacity to the infrared electromagnetic spectrum was confirmed using the InGaAs camera. As a practical application, Sn-doped NC ink proves effective as a colorant surfactant, producing visible markings only under infrared imaging. SEM imaging confirms the well-defined NC particle distribution in hexane, and FTIR, complemented by Drude model analysis, verified the LSPR peak values and the infrared absorption characteristics. The synthetic process for scale-up involves a continuous precursor injection batch reactor, ensuring uniform doping and consistent NC quality. The scaled-up Sn-doped NCs exhibit crystallinity, as confirmed by XRD patterns, with peaks corresponding to the bixbyite phase of indium oxide. Scanning electron microscopy (SEM) and XRD were used to observe the formation of spherical Sn:In₂O₃ NCs with crystallinity.

Even as the scale of the experiment increases, the same LSPR phenomenon can be observed. This study suggests the feasibility of the mass production of Sn:In₂O₃ NCs. Mass production can bring significant industrial benefits due to the fact that the raw material cost is 10 times lower, and the synthesis capacity is 20 times larger, which can significantly reduce the economic cost of production. Mass-produced metal-modified doped Sn:In₂O₃ nanocrystalline ink can be applied to a wide range of materials across industries. It can be applied to smart windows, photocatalysts, light sensors, and all other fields that require infrared blocking and visible light transparency. It can also be used as an infrared marker for security purposes, contributing to the advancement of application technology of infrared semiconductor NCs.