Mask-free fabrication and chemical vapor deposition synthesis of ultrathin zinc oxide microribbons on Si/SiO₂ and 2D substrates

ZnO nanostructures have unique properties which are rarely found in other materials. Besides its semiconducting properties, ZnO exhibits a strong piezoelectric effect due to its wurtzite structure, which makes it suitable for different types of applications.^1–4 ZnO has high exciton binding energy of 60 meV at room temperature which can be additionally enhanced, up to 100 meV, via doping.⁵ These characteristics make it useful and effective materials for many optoelectronic applications, including tunable UV photodetectors and light-emitting diodes.^6,7 Moreover, the transparency, wide bandgap energy (≈3.37 eV), and low toxicity further diversify ZnO nanomaterials potential applications. For example, ZnO thin films can be used as conducting windows for silicon solar cells,⁸ high-performance piezoelectric layers for piezoelectric actuators and integrated MEMS,^2,9 and versatile platforms for cancers imaging and treatment.¹⁰ 

Several methods have been utilized to fabricate ZnO thin films. Frequently used techniques are sol–gel methods,¹¹ pulsed laser deposition,¹² physical vapor deposition,¹³ molecular beam epitaxy,¹⁴ magnetron sputtering,¹⁵ atomic layer deposition (ALD),¹⁶ and chemical vapor deposition (CVD).¹⁷ Most of these methods can produce ZnO thin films with the thicknesses in the range of 100 nm to a few microns.^{11,12,18–23} Recently, some studies have shown the ability to produce ZnO thin films with controlled thicknesses. For example, the sol–gel and hydrolysis-condensation methods were used to produce 30 nm thick ZnO thin films by spin-coating of the sol–gel precursor at 3000 rpm for 40 s.²⁴ In another study, ALD method was utilized to control the production of ZnO thin film from 5 to 70 nm thick with the growth ratio of ≈0.16 nm/cycle at 200 °C.²⁵ 

However, in order to produce patterned ZnO thin films with controlled shapes and/or predesign architectures, usually, a lithographic processing is required. Most commonly used methods are conventional lithography techniques, such as photo- or e-beam lithography which involves polymer-resists, or/and hard masks and multistep fabrication procedures.²⁶ Attempts have been made to use nonconventional methods (such as microcontact and inkjet printing techniques) to demonstrate patterned ZnO thin films on substrates.^27,28 Although these techniques have succeeded in producing ZnO patterns, they have certain limitations. For example, patterns produced with an inkjet printing method, usually, were larger in lateral dimensions ranging from 80 to 400 μm, where edges of the patterns, their uniformity, and the thicknesses of the patterned structures were not easily controlled.^27,29

In contrast, the patterns of ZnO thin films (with lateral dimensions of 20 and 170 μm, and with the thickness of ∼7 nm) were achieved with polydimethylsiloxane (PDMS) stamps using the micro-contact patterning method.²⁸ However, in this study the individual PDMS masks were required for each new pattern design, which is, in turn, a time consuming and laborious fabrication process.

Producing ZnO thin films, selectively, with desired shapes and design patterns in a controlled way has proven challenging. Recently, 20 nm thick ZnO patterns were fabricated on Si/SiO₂.³⁰ Shadow-mask was used on spin-coated samples with the thin film polymer precursor, zinc diacrylate, to define specific patterns on the substrates. Then, a series of steps including the UV-light exposure, etching, developing, and annealing were required to produce desired ZnO patterns. More recently, a laser-induced microbubbles technique has been employed to produce spatial patterns of thin ZnO nanostructures on indium-tin-oxide substrates using the Zn(NH₃)₄ solution.³¹ Although ZnO patterns (∼70 nm thick) were successfully produced, it was reported that in order to produce simultaneous patterns, it was critical to use transparent substrates with low thermal conductivity, and ability to absorb the IR laser wavelength (λ = 1064 nm). These conditions would limit the substrate selections. Thus, it is important to develop simple and versatile methods to produce ZnO thin films patterned in desired shapes in a simple and flexible fashion, with the resolution of the patterns comparable to those produced via photo- or e-beam lithography.

In this study, we employed a two-step fabrication process which combines (a) direct-write patterning (DWP) technique and (b) CVD method, to produce controlled patterns of thin ZnO nanostructures with micrometer width. An aqueous Fe containing solution was used as a catalytic ink, which was deposited directly on the substrates via mask-free DWP. In this DWP method, multipen atomic force microscope (AFM) cantilevers were employed to deposit the prepared catalytic ink directly and precisely on preselected locations of the substrates. This technique makes it possible to control the geometry of the catalytic precursor patterns (designed on the computer software), thus eliminating the need for “masks” and simplifying the fabrication process by reducing the number of steps in the protocol. Most important, however, is that our DWP could be performed on a variety of substrates, and produce heterostructures and/or multilayered architectures with high precision.

In this work, ZnO microribbons, with varying thicknesses, were produced on Si/SiO₂ substrates by controlling the key parameter, “speed of writing,” during the deposition of Fe precursor patterns prior to the growth. Additionally, ZnO microribbons were also prepared on other surfaces, such as graphene (Gr) and boron nitride (BN) on Si/SiO₂, effectively, creating ZnO/2D heterostructures. Creating ZnO directly on 2D materials and producing such interfaces/heterostacks could be promising for future nanoscale applications.

Iron (III) nitride nonahydrate (Fe(NO₃)₃·9H₂O), (Sigma-Aldrich, 99.99% purity), 2.6 mg was dissolved in 15 ml of deionized (DI) water (18.2 MΩ cm) and sonicated for 10 min to prepare the stock solution “master ink.” The master ink is further diluted in the DI water with the ratio of (1:1) to be utilized as the Fe-catalytic precursor for the direct-write patterning. ZnO powder (Alfa Aesar, 99.999% purity) and Graphite (Alfa Aesar, 99.999% purity) were mixed together in equal amounts of (35 mg) each and used as ZnO/C source precursors during the growth process. Ammonia borane (NH₃-BH₃) (Sigma-Aldrich, 97% purity) 200 mg powder was used as a BN-source precursor. Copper foils (Alfa Aesar, 99.8% purity and thickness of 0.025 mm) were used as substrates for Gr and BN thin film syntheses.

Different substrates: Si/SiO₂, Gr/SiO₂/Si, and BN/SiO₂/Si substrates were employed for patterning process. These substrates were prepared and cleaned before the patterning. P-doped Si substrates with 300 nm thermal oxide, <0.005 Ω cm resistivities, and <100> orientation were cut into pieces of 5 × 5 mm² and cleaned in DI water, and subsequently in acetone and isopropanol using an ultrasonic bath sonicator, for 10 min in each solvent. Next, samples (chips) were treated in a UV-ozone system to render their surfaces hydrophilic. Gr and BN surfaces were also utilized as substrates. Both Gr and BN 2D films were prepared in a home-built CVD system using Cu foils as catalytic growth substrates (see the supplementary material⁵⁶ for details on CVD growth process and etching). First, Cu foils were etched with a 1M FeCl₃ solution diluted in DI water [volume ratio of (1:3)], and subsequently cleaned with 50% HCl acid diluted in DI water [volume ratio of (1:1)] following the common wet etching procedure.³² Finally, Gr and BN films were carefully transferred onto the Si/SiO₂ substrates and cleaned with acetone in order to remove the polymer sacrificial layer which was used to aid the transfer process (see the supplementary material and Fig. S1 for more details on Gr and BN preparation methods⁵⁶).

Additionally, Si/SiO₂ (∼100 nm thick SiO₂) substrates with prefabricated electrodes (Pt/Ti, 60/6 nm) were used as a device platform for the direct fabrication of ZnO microribbons. Metal electrodes were prepared using conventional photolithography and lift-off process. These Si/SiO₂ chips with Pt/Ti electrodes were thermally treated at 450 °C with 400/100 sccm of Ar/H₂ gas mixture to remove any remaining contaminants from the surface.

A custom-made DWP system is equipped with three piezo-driven nano-positioning stages for xyz positioning, resolution of the piezo-stages is ∼100 nm. The system’s sample stage is also equipped with the tilt correction capabilities. This tool, which is computer-controlled, has a removable AFM-cantilever holder which can accommodate any AFM-cantilever chip. Silicon nitride AFM cantilevers with 12 pens (∼60 μm pitch, 2.6 N/m fabrication k, 107 and 22 μm fabrication length and width, respectively, and nominal tip radius of 15 nm), were used in all experiments. Matching inkwells were also purchased from the Advanced Creative Solutions Technology LLC. In this work, multipen AFM-cantilever chips were used to deposit the Fe-catalytic ink onto the substrates. The ink transfer takes place when ink coated AFM tips are brought in contact with the substrate. Specific environment conditions and substrate properties should be considered to ensure successful patterning; such as temperature and the relative humidity (RH), cleanliness, and hydrophilicity of the substrates/surfaces. The optimal ink composition/concentration and the speed of writing are additional critical parameters that ensure reproducibility. In this study, the RH was adjusted to be in the range of 50–55% at room temperature. Also, the substrates and AFM cantilevers were treated with UV-ozone to render their surfaces sufficiently hydrophilic, hence, to facilitate the ink transfer from cantilever tips to the substrate during the patterning process. The prepared iron containing ink “master ink” was diluted in DI water with the volume ratio (1:1) to serve as the catalyst. Patterning on the substrates was performed with different writing speeds; 10, 20, and 50 μm/s. With our patterning tool, we could control writing speed that proved to be an important parameter in determining the thickness of the resulting microribbon structures. Based on simple analyses, the average amount of Fe material per unit area left on the surface during the heat treatment (before the growth stage) was estimated for the different patterned line-shaped microribbons. Calculations were made by approximating volume concentration of Fe catalyst based on geometric footprint (half cylinder) shaped liquid ink deposits. Interestingly, we found that the average ratio of the Fe material deposited in these microribbons (∼3–6 μm in width, 50 μm in length) via DWP technique on the substrates at a low speed (10 μm/s) is ∼1.5–2 times more as compared to the high patterning speed (50 μm/s). The average surface roughness of the patterned Fe-line/microribbon deposits (prior to ZnO growth) was determined from AFM topography images to be in the range of 1–3 and 2–4 nm, respectively for highest and lowest speeds, as shown in the supplementary material and Fig. S2.⁵⁶ 

A home-built CVD system was employed to synthesize ZnO microribbons on different substrates. This system is equipped with a three-zone furnace and controller (Thermo Scientific Lindberg/Blue M), 5 ft long and 1 in. diameter quartz tube, and the digital box to control four individual gas-flow meters (Sierra Instrument). Furnace controller is interfaced with the computer via laboratory virtual instrument engineering workbench (labview) based code that ensures automated repeatability and monitoring of the CVD process. A mixture of ZnO/C powder with the weight ratio of 1:1 was loaded into the high-temperature zone (930 °C) of the CVD system, where ZnO source precursor vapor was produced. Ultrahigh-purity Argon (Ar) gas (70 sccm) carried the produced precursor vapor downstream to the low-temperature zone (795 °C), where samples were situated. In addition to the patterned samples (prepared via DWP), substrate pieces prepared by dip-coating (in Fe-catalytic ink) were also inserted into the low-temperature zone, to serve as reference samples. The optimal growth time was 80 min, after which the furnace was allowed to cool down naturally (see Table S1 in the supplementary material⁵⁶ for more details related to the growth of ZnO microribbons).

An AFM (Park Systems, NX10) was used to acquire topography images, measure the thickness of ZnO microribbons, and determine the surface roughness of the ZnO films. Renishaw InVia Raman spectroscopy system was used to characterize Raman modes of the ZnO microribbons with a laser wavelength of 532 nm and 100× objective. X-ray photoelectron spectroscopy (XPS) (PHI 5000 versa probe-II) was utilized to determine the elemental compositions of ZnO microribbons and their chemical states. Rigaku-miniFlux x-ray diffraction (XRD) system was used to identify the crystal structure of ZnO microribbons with Cu Kα radiation (λ = 0.154 nm), 40 kV, and 15 mA source. JEOL JSM-7001 LVF Field Emission scanning electron microscope system was employed for energy-dispersive x-ray (EDX) analyses. Keithley electrometers (6514 system electrometer) and current/voltage source were utilized for I–V measurements.

Here, we report the capability of producing arrays of ultrathin ZnO microribbons, with controlled thicknesses and lateral geometries, on the Si/SiO₂ substrates using the mask-free patterning approach and CVD method. The iron aqueous solution is employed here as the liquid catalyst to facilitate the formation of the ZnO microribbons. We also demonstrate the ability to fabricate ZnO microribbons directly at precise locations on the two-dimensional materials, such as Gr and BN, effectively creating heterostructures interfaces. Figure 1 schematically illustrates the process of controlled production of ZnO microribbons on the substrates. Figure 1(a) depicts the patterning process where multipen AFM cantilever coated with the Fe-catalytic ink is used to create arrays of linelike patterns at precise locations on the substrates. The substrate can be Si/SiO₂, or Gr and BN prepared on Si/SiO₂. Figure 1(b) is a cartoon representation of the interaction between the vapor (precursor) molecules and the catalyst nucleation sites on the substrate during the CVD synthesis. Figure 1(c) schematically shows a patterned sample after the growth and the inset represents an actual AFM topography image of a ZnO microribbon patterned on Si/SiO₂ substrate. (More details about the ink preparation, patterning process, and CVD growth conditions can be found in the experimental part and Table S1 in the supplementary material⁵⁶).

Briefly, iron aqueous “master ink” was diluted in the DI water to achieve optimal concentration for the growth. Next, the working ink was used to create arrays of linelike patterns of the catalytic deposits on the substrates. ZnO and graphite powders, mashed together, were used as the source precursor placed inside the boat and situated at high-temperature zone (930 °C) of the CVD. The patterned samples were placed 18 cm away from the source location, at the lower-temperature zone (795 °C). Ultrahigh-purity Ar gas carried the precursor vapor from the source to the sample zones, in the CVD furnace, facilitating the vapor-solid growth of the ZnO nanostructures at the patterned locations. The optimal time for the growth was 80 min. Additional samples prepared via dip-coating, using the same ink, were processed simultaneously in the CVD furnace to serve as reference samples.

Using these procedures, thin patterned arrays of ZnO microribbons with controlled geometries were successfully synthesized on Si/SiO₂ substrates. Our approach also allows for the rapid and convenient measurements of their topographical characteristics.

Figure 2(a) shows AFM image of an array of ZnO microribbons produced on Si/SiO₂ substrate. The average width of these microribbons is approximately 3 μm, as shown in Fig. 2(b). Higher magnification images of the same area of the microribbon are shown in Figs. 2(b) and 2(c), where ∼2 nm roughness of the ZnO surface can be seen in Fig. 2(c), inset. Our results demonstrate the ability to produce desired ZnO patterned thin structures with the uniform surfaces, well-defined edges, and most importantly, with the lateral resolution/size comparable to the conventional lithographic methods. Previously, high-quality vertically oriented ZnO nanowires have also been produced in various controlled geometric architectures using similar procedures in our laboratory and reported elsewhere.³³ In general with the DWP technique and the CVD method, a variety of patterned nanomaterials have been successfully prepared in a controlled and scalable manner.^34,35 Using our two-step (first, patterning and second, growth) fabrication approach, arrays of ZnO microribbons were also synthesized on different substrates and will be discussed later in this paper.

In order to evaluate and determine the quality, compositions, and crystalline structure of the synthesized ZnO microribbons, different characterization techniques have been employed. Figure 3 shows spectroscopy data acquired from the prepared arrays of ZnO microribbons on Si/SiO₂ substrates. Resonant Raman spectroscopy was used to examine the main characteristic vibrational modes of ZnO nanostructures at the locations of the patterned ZnO microribbons arrays, presented in Fig. 3(a). The Raman spectral plot shows the main active Raman mode (E₂) of the ZnO, which is split into two modes, represented by the E₂^low and E₂^high, as shown in Fig. 3(b). These two modes correspond to the specific peaks that are located at ≈100 and ≈437.6 cm⁻¹, respectively. The presence of these modes in the Raman spectral signals indicated that our ZnO microribbons produced from the Fe-catalytic ink have a wurtzite lattice structure.³⁶ The FWHM of the E₂^high mode was measured (∼13.9 cm⁻¹) and indicated that the as-grown ZnO microribbons have high crystallinity. Also, the absence of the E₁(TO) mode in our Raman measurements refers to the low defects/vacancies, thus indicating the high quality of the produced ZnO microribbons. The Raman data were collected from the probed area marked by the laser spot (bright dot) in the middle of the patterned microribbon, Fig. 3(a). Typical diameter probed by the laser (λ = 532 nm and 100× objective) is ∼0.8 μm. See the supplementary material (Fig. S3)⁵⁶ for additional examples of the Raman measurements.

For further investigation, XPS system was utilized to identify the elemental compositions and chemical states of the patterned ZnO microribbons samples. Figure S4(a) (in the supplementary material⁵⁶) represents the XPS survey spectra of the grown ZnO microribbons on Si/SiO₂ substrates, where the typical presence of the Zn and O can be seen. Figure S4(b) shows the carbon peak C1s (located at ∼282.4 eV) that is used for the calibration and binding energies normalization of the measured data. From the survey XPS analysis, it is obvious that only Zn and O are the predominant elements in our patterned ZnO microribbons, which were produced using Fe-catalytic ink. Our results are comparable to other reported studies where alternative catalysts were used for the growth of the ZnO nanomaterials.^37,38

Next, we focused on examining the Zn2p binding energy peaks, which are located at the positions ≈1020 and ≈1043 eV, and which represent Zn2p_3/2 and Zn2p_1/2 chemical states, respectively, as shown in Fig. 3(c). Also, the energy difference between the spin–orbit interaction of the Zn2p state [Fig. 3(c)] was calculated to be ∼23 eV, which relates to the Zn2p state in the ZnO nanostructures.^37,39 Figure 3(d) shows the binding energy of the oxygen O1s state, where the peak is located at ≈530 eV of the XPS spectrum, which was ascribed to the oxygen ions participation in the ZnO crystal lattice. The Gaussian fit was performed on the oxygen curve and clearly showed a single oxygen peak, which indicates the absence of the oxygen vacancies/defects and confirms the quality/purity of our as-grown ZnO microribbons.^38,39

To explore the structural properties of the ZnO microribbons, XRD measurement was performed. We found that measuring the XRD signals on the patterned samples was difficult due to the small amount of the materials in the patterned areas of ZnO micro/nanostructures; XRD signal was dominated by Si signature peak. Therefore, instead, we performed the XRD analysis on the reference samples (prepared via dip-coating) with the ZnO thin films, which had significantly more surface area covered with the synthesized ZnO material.

Figure 4(a) depicts a representative AFM topography image (30 × 30 μm) of the ZnO thin film on Si/SiO₂ substrates (the dip-coated sample), and Fig. 4(b) shows a representative XRD measurement of the ZnO thin film prepared via dip-coating (5 × 5 mm sample).

Here, XRD analysis demonstrates one dominant peak which is related to the (111) crystal orientation. In contrast, the patterned samples with ZnO microribbons are examined locally by the EDX spectroscopy to determine the elemental composition of the synthesized material. Figures 4(c) and 4(d) show the EDX spectrum and the corresponding SEM image of the microribbon area where the EDX analysis was performed. The EDX plot shows main dominant peaks related to the Zn (contributing at 35.9%) and O (contributing at 61.1%), respectively. Additionally, a prominent Si peak is also present which is expected for the samples fabricated on Si/SiO₂ substrates. These results confirmed the structural quality and elemental content of the produced ZnO microribbons.

In spite of numerous methods which successfully demonstrated the fabrication of ZnO thin films, the preparation of these films with specific and controlled thicknesses has proved challenging. Several studies have shown the impact of the thickness on the properties of the prepared ZnO thin films. For example, the band gap, average transmittance, absorption, and other electrical, optical, and piezoelectric properties of ZnO thin films are highly dependent on the thickness of the films.^40–43 

Therefore, controlling the thickness of the produced ZnO nanostructures is essential for targeting specific applications. Here, we show the ability to control three different thicknesses of synthesized ZnO microribbons on Si/SiO₂ substrates by controlling the speed of writing, which is a critical parameter in our DWP technique. Figures 5(a), 5(d), and 5(g) show representative AFM topography images of three different arrays of the ZnO microribbons on Si/SiO₂ substrates. Three different speeds of writing were used in each case during the patterning process, in order to create catalytic ink deposits on the substrate, prior to the growth. Figures 5(b), 5(e), and 5(h) represent corresponding higher magnification topography images of the same ZnO microribbons at the areas marked with the dotted squares in Figs. 5(a), 5(d), and 5(g). Here, the average width of the produced ZnO microribbons ranged between ∼3 and 6 μm. Figures 5(c), 5(f), and 5(i) show the line profiles of the ZnO microribbons arrays prepared on Si/SiO₂ with three different controlled average thicknesses (∼2, 4, and 8 nm) corresponding to three different speeds of writing (50, 20, and 10 μm/s), respectively. AFM height profile data acquired along the lines are shown in Figs. 5(a), 5(d), and 5(g). We found that the speed of patterning “writing” played an important role in determining the resulting thickness of the produced nanostructures. For example, using relatively low speed (10 μm/s) of writing allowed us to continuously deposit more catalysts per line-patterns, while the cantilever tips were in contact with the surface of the substrates.

These patterned catalysts tend to agglomerate during the CVD synthesis and form denser nucleation and therefore thicker ZnO microribbons. On the other hand, faster “writing” speeds (20 μm/s and greater) deposit lesser amount of the ink-catalysts, per line-patterns, which resulted in thinner microribbons formations.^33,37 Table S2 (in the supplementary material⁵⁶) describes the relationship between the optimal “writing” speeds which were used during the patterning and their respective thicknesses of the ZnO microribbons. These results imply that our mask-free DWP technique combined with the CVD growth method can produce ultrathin ZnO microribbons in a controlled fashion precisely on preselected locations of the Si/SiO₂ substrates. Controlling the thickness of the prepared microribbons would add a significant advantage in the future strategies for the development of other patterned ZnO nano-architectures using our bottom-up fabrication approach.

We have also employed the DWP technique and CVD method to produce arrays of controlled ZnO microribbons on other substrates. In this work, Gr and BN thin films were tested as substrates for the direct fabrication of heterostructures interfaces of ZnO/2D materials. Such heterostructures could be promising for future nanoscale applications, such as supercapacitors, UV-photodetector, biosensors, and other.^44–46 

ZnO thin films on Gr and BN in heterostructures have been demonstrated with various fabrication and growth methods, including the metal-organic vapor-phase epitaxy, solution-based spin-coating of ZnO thin films, and ALD, to advance the electronic and optoelectronic applications.^47–51 Some of these methods produced ZnO thin films on the 2D materials with thicknesses ranging from 100 nm to few microns. However, in those studies in order to produce desired patterns of ZnO nanostructures on Gr or BN, some type of conventional lithographic technique was usually employed for patterning; photo- or e-beam lithography. Fewer studies have shown precisely controlled thin film architectures of ZnO/2D materials due to the complexity associated with the multistep procedures in the conventional fabrication processes.⁵⁰ 

In our study, ZnO microribbon arrays were patterned directly on Gr and BN thin films using the same two-step protocol (DWP patterning and CVD growth). Examples of ZnO microribbons arrays on Gr and BN substrates are shown in Figs. S5(a) and S5(b) (supplementary material⁵⁶). Figures 6(a) and 6(d) represent AFM topography images of the ZnO microribbons which were prepared on Gr and BN sheets, respectively. Widths and thicknesses of the ZnO microribbons on Gr and BN were also measured using AFM as shown in the insets. Figures 6(b) and 6(e) are zoomed-in AFM topography images, generated from the areas outlined with dotted squares in Figs. 6(a) and 6(d). Raman measurements were also performed on these prepared heterostructures of ZnO/Gr and ZnO/BN as shown in Figs. 6(c) and 6(f), respectively. These Raman measurements show the presence of the E₂^low and E₂^high modes of the ZnO nanostructures, in addition to the strong Raman signals of Gr and BN. These results demonstrate the ability to prepare good-quality patterns of ZnO microribbons on preselected locations of the 2D materials, which could be further incorporated into devices for flexible and transparent electronics, and for energy harvesting applications.^52–54 

Although we showed the ability to use the DWP technique to write arrays of patterns on the 2D surfaces, there were certain challenges associated with ZnO/2D heterostructures fabrication, which limited our control over the resulting ZnO microribbons. As it was mentioned earlier in the experiment section, the UV-ozone treatment of the substrates renders surfaces hydrophilic for optimal patterning. However, treating the 2D surfaces for more than few minutes could result in etching of the 2D films and compromise their quality. Because hydrophilic properties of the 2D surfaces could not be tuned more accurately, it prevented the fine control over the thickness of the resulting ZnO microribbons. Hence, with less hydrophilic surfaces, slower speeds (10 μm/s) of writing were used in order to produce continuous patterns of catalytic ink on the 2D thin films. Less hydrophilic surfaces and slower speed of writing, therefore, resulted in the thicker patterns. This is because in those conditions larger volumes of the catalytic inks were deposited per unit area of the surfaces (i.e., higher catalyst density); hence, thicker microribbons were formed.^33,55 Despite some present limitations in patterning on hydrophobic surfaces, we have demonstrated that the Fe-based ink is an effective catalyst which can be patterned via the DWP technique on different surfaces regardless of the physical nature of the substrate.

Lastly, we tested ZnO microribbons fabricated directly on metal electrodes following the same two-step protocol. After the CVD synthesis, ZnO microribbons formed a direct interface with the metal electrodes of the devices. This could simplify and simultaneously advance the device fabrication, by eliminating a number of steps in the fabrication process, as compared to the standard conventional and/or unconventional lithography methods which have been employed and reported in other studies.^26–28 Figure 7(a) shows a representative optical image of ZnO microribbon in contact with four metal electrodes fabricated on Si/SiO₂, thus creating a simple device architecture. The distance between the electrodes was 5 μm, and the average width of the ZnO microribbon was ∼5 μm. Figure 7(b) shows a representative I–V measurement of such device.

The I–V characteristics were measured under vacuum at the room temperature on several devices, and most of them showed a similar response. Although the maximum sweeping currents which we were able to apply were limited to −1 to 1 nA, and above 5 nA devices generally did not survive, we were still able to record measurable resistances of the patterned ZnO microribbon in the low current regime. We speculate that the cause of such breaking down could be explained by weak/bad contacts between Pt/Ti metal electrodes and the ZnO microribbon because of the migration of the Fe-catalyst particles from the Pt metal surface towards the edges of the electrodes during the CVD synthesis.

In conclusion, we have demonstrated a unique and scalable technique to fabricate ultrathin ZnO microribbons precisely at selected locations on the Si/SiO₂ substrates using mask-free DWP approach and CVD method. The produced ZnO microribbons were synthesized using Fe aqueous catalytic ink deposited on the substrates in the line-shaped patterns and subsequently processed in the CVD furnace to form ZnO nanostructures. Patterned ultrathin ZnO microribbons on Si/SiO₂ in three different thicknesses, 2, 4, and 8 nm, were achieved. Those thicknesses corresponded to three speeds of writing, 50, 20, and 10 μm/s, that were used during the patterning process. The average lateral dimension of the patterned ZnO microribbons ranged from 3 to 6 μm.

In this study, we also demonstrated the ability to pattern directly on Gr and BN films. The average thickness of ZnO microribbons on Gr and BN surfaces was 50 and 400 nm, respectively. Although it was more challenging to achieve a fine control over the thickness of the ZnO microribbons on less hydrophilic Gr and BN surfaces, these proof-of-concept experiments demonstrated the compatibility of DWP (using aqueous Fe-catalytic ink) with Gr and BN, and the resilience of these 2D materials to high temperatures (∼800 °C) required for the CVD process. Moreover, we showed the ability to pattern directly on prefabricated metal electrodes; however, we obtained measurable resistance only at very low currents. We believe that if an alternative 2D material, such as graphene, is used as electrodes, it could improve the metal–ZnO interface. Controlled production of thin ZnO nanostructures on Si/SiO₂ substrates and directly on 2D materials, in heterostructures and/or multilayer assemblies, can open-up new avenues for integration of patterned ZnO thin film into different types of architectures and into devices for improved and more efficient future transparent electronics, optoelectronics, piezotronics, and other applications.

The Center for Nanoscale Materials-Argonne National Laboratory was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC02 06CH11357. I.K. acknowledges the support of NSF MRI program (Award No. 1338021) and the Saint Louis University seed funds. D.A. acknowledges the financial support of the Iraqi Ministry of Higher Education and Scientific Research (MOHESR) and University of Misan, College of Science, Misan, Iraq.