Methods For Dual-Scale Surface Texturing

Abstract

Methods for preparing a substrate surface are provided, for purposes including manufacturing a low reflectivity surface. In some aspects, the methods include providing a material comprising an etching mask on a substrate, subjecting the material to a first isotropic etching phase, and subjecting the material to a first anisotropic etching phase, thereby forming a textured surface on the material, wherein the textured surface comprises structures with dimensions in a sub-micron range.

Description

BACKGROUND OF THE INVENTION

The field of the invention relates to a process and method for preparing at least one surface of a substrate. In one embodiment, the invention relates to a dual-scale surface texturing method involving isotropic and anisotropic etching to provide nanometer-scale and micrometer-scale features capable of determining optical properties of one or more substrate surfaces.

Surface texturing is one of several methods utilized across multiple industries in fabrication processes associated with electronic and optical devices, biochips, biosensors, and so forth. For example, surface texturing approaches have found use in the manufacturing of photovoltaic (solar) cells. In many optical device applications, and other applications, the geometrical aspects, as well as the optical properties, of materials employed are used as a way to control the optical behavior. Specifically, particular textures or structures are utilized to dictate light absorption and reflection characteristics of devices. For instance, micrometer- and nanometer-scale features associated with particular solar cell layers can improve energy capture by reducing losses due to reflected light from cell surfaces, and increasing absorption. In this manner, the efficiency of solar cells can be increased. In addition, improved capture of incident radiation, by way of a smaller active layer volume required for a desired efficiency, results in a reduced amount of materials needed in the fabrication process.

Current surface texturing technologies are limited since they create texture structures with large sizes, on the order of 5-10 micrometers, and high aspect ratios. For solar cell applications, such features produce significant light reflection, and hence reduce cell efficiency. Hence, there is a need for improved device patterning techniques related to producing structures at the nanometer (i.e. sub-micrometer) range.

SUMMARY OF THE INVENTION

The present invention overcomes the aforementioned drawbacks by providing a methodology for producing textured substrate surfaces. Specifically, a method is provided for preparing an etching mask on a substrate, and utilizing the etching mask to produce textured surfaces. The method also includes performing a dual-scale etching process, using a prepared etching mask, to achieve a texturing of the one or more substrate surfaces.

In one aspect of the present invention, a method of preparing a substrate surface is provided. The method includes providing a material comprising an etching mask on a substrate. The method also includes subjecting the material to a first isotropic etching phase, and subjecting the material to a first anisotropic etching phase, thereby forming a textured surface on the material, wherein the textured surface comprises structures with dimensions in a sub-micron range.

In another aspect of the present invention, a method of preparing a substrate surface is provided. The method includes dispersing a plurality of particles in a suspending medium to form a suspension, and spin-coating the suspension on a substrate comprising a material to form an etching mask on the substrate. The method also includes subjecting the material to a first isotropic etching phase, using the etching mask, forming a modified etching mask, and subjecting the material to a first anisotropic etching phase, using the modified etching mask, thereby forming a textured surface on the material, wherein the textured surface comprises structures with dimensions in a sub-micron range.

The foregoing and other aspects and advantages of the invention will appear from the following description. In the description, reference is made to the accompanying drawings which form a part hereof, and in which there is shown by way of illustration a preferred embodiment of the invention. Such embodiment does not necessarily represent the full scope of the invention, however, and reference is made therefore to the claims and herein for interpreting the scope of the invention.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 shows a silica micro-sphere (SMS) dispersibility comparison between N,N-dimethyl-formamide (DMF) and water, wherein (a) calculated and measured absorbance is plotted versus wavelength for a single 310 nm SMS in each solvent; and scanning electron microscopy (SEM) images of (b), and (c) show SMS assemblies using DMF and water, respectively.

FIG. 2 shows a SEM image illustrative of the SMS cluster effect producing non-uniform SMS distribution during spin-coating.

FIG. 3 shows: a contact angle measurement of (a) water and (b) DMF; a comparison of surface coverage with 300 ul (100 mg/ml) for (c) SMS_water and (b) SMS_DMF solution droplets on 2-inch Si substrate; and surface images of (e) SMS_water, and (f) SMS_DMF.

FIG. 4 shows SEM images comparing the coverage difference between (a) SMS_DMF and (b) SMS_water.

FIG. 5 is a schematic illustration of two spheres partially immersed in a fluid layer for capillary attraction, F_cap.

FIG. 6 is a schematic illustration of (a) fast solvent evaporation, and (b) slow solvent evaporation during spin-coating, producing (c) localized SMS assembly with discontinuous F_cap, and (d) long-range SMS assembly after expanded F_cap.

FIG. 7 shows SEM images for SMS coverage from various concentration of SMS_DMF, which are (a) 50 mg/ml, (b) 100 mg/ml, and (c) 150 mg/ml.

FIG. 8 shows SEM images illustrative of the acceleration rate effect for uniform SMS assembly layer formation; (a) 20 rpm/s, (b) 50 rpm/s, and (c) 80 rpm/s.

FIG. 9 shows a surface image of (a) SMS deposited on a 2-inch round Si substrate, and the corresponding SEM images at different magnifications of (b) 250×, (c) 2000×, and (d) 25000×, respectively.

FIG. 10 shows a surface image of (a) SMS deposited on a 4-inch round Si substrate, and the corresponding SEM image at magnification (b) 100×; SEM images (c) and (d) represent regions of low and high SMS monolayer coverage, respectively.

FIG. 11 shows a schematic illustration of RIE process to etch solid surface.

FIG. 12 shows a schematic illustration and SEM images to fabricate various shape of silicon nanostructure with an RIE process.

FIG. 13 shows reflectance measurements for various RIE processed silicon surfaces.

DETAILED DESCRIPTION OF THE INVENTION

Production of modern-era devices, with increased complexity and ever-decreasing dimensions, has required overcoming many technical challenges associated with approaching fundamental physical limits. In looking to push limits further, there are remaining challenges for producing devices structures with dimensions in a micrometer and nanometer range at lower-cost and with high-throughput.

The present disclosure describes an approach for treating or processing substrate surfaces and materials therein for purposes of producing structures or features with dimensions in range including a sub-micron range. As will be described, in some aspects, a substrate can be prepared with single or multiple layers of particles, dispersed and/or self-assembled as desired on a substrate surface. The particles, in addition to other materials and compositions, may then serve as a mask or protective layer in a surface treatment process for shaping, texturing, or profiling a substrate surface or multiple substrate surfaces.

Among other applications, the present invention can be utilized in the fabrication process of photovoltaic cells. As such, the scale of colloidal particles have found important uses in device fabrication, and particularly in photovoltaic cell manufacture, where the need for increased cell efficiency and reduced cost has pushed the technology toward reduced dimensionality and enhanced energy absorption.

In accordance with one embodiment of the present invention, an approach for achieving a desired layer uniformity using a spin-coating process is described. Particularly when colloidal particles are utilized, uniformity over large substrate areas may be achieved by using a medium for suspending the particles with properties configured for efficient self-assembly during a spin-coating process. For example, as will be described, by utilizing a suspending medium with a relatively slow rate of evaporation, as compared to say water, a more consistent convective flux of the particles and medium, the particles dispersed or suspended therein, can be achieved during a spin-coating process. Similarly, by utilizing a suspending medium in which the particles are evenly or uniformly dispersed before a spin-coating process begins, a uniformity of a resultant layer may be further improved. For example, an aprotic medium for suspending colloidal particles may be used, since such medium can possesses characteristics that improve the effectiveness of the spin-coating process.

Particles for use in accordance with aspects of the present disclosure, may be generally sized to an average dimension in the range of 10 nanometers to 10 micrometers, and shaped to be spherical, ellipsoidal, and the like, although other shapes and sizes may be possible. Particularly, the average dimension typically refers to the longest longitudinal length common to the plurality of particles. The particles may be manufactured using suitable materials and methods, in accordance with a desired application. By way of non-limiting example, the particles may comprise metal oxides, or silica, or alumina, or zirconia, or titania, or carbon, or any combination thereof. The particles may be dispersed in a suspending medium, using suitable techniques, such as sonication, to form a suspension with properties according to a desired concentration, or other criteria. For example, the concentration of the plurality of particles in the suspending medium can be in the range of 0.1 to 10.0 wt %, although other values are possible.

By way of non-limiting example, a suspending medium, for use in accordance with the present disclosure, can be an amide solvent, or a hydrophilic solvent, or a suspending medium selected from a group consisting of tetrahydrofuran, ethyl acetate, acetone, dimethylformamide, acetonitrile, dimethyl sulfoxide, and mixtures thereof. In some aspects of the invention, the suspending medium does not include a surfactant. In other aspects of the invention, the suspending medium may be dimethyl-formamide. In addition, the suspending medium can have a boiling temperature in the range of 100 degrees Celsius to 200 degrees Celsius, or a viscosity in the range of to 0.1 mPa sec to 1.0 mPa sec at 25 degrees Celsius, or a surface tension in the range of 20 mN/m to 40 mN/m at 25 degrees Celsius, or combinations thereof, although other values are possible. As will become apparent, suspending media, with properties as described, facilitate dispersibility, wettability, and evaporation requirements for obtaining an enhanced uniformity of particles on a substrate.

The suspension, or solution containing the particles dispersed in the suspending medium, can then be distributed onto a substrate in any manner using, for example, a spin-coating process. In some aspects, the suspension can undergo a rotational acceleration to a rotational target speed between 1000 and 5000 rotations per minute, wherein the rotational acceleration can be in the range of 50 rotations per minute per second to 100 rotations per minute per second, although other rotational acceleration and speed values, along with multiple rotational stages and durations, may be possible.

In addition, the substrate may be of any type, and can include materials such as doped or undoped silicon, although it may be appreciated that other substrate types and substrate materials may also be possible. By way of example, the substrate may have a surface area of about 1 cm² to about 1000 cm², although other values are possible. Moreover, in certain aspects, the dispersed particles form a two-dimensional (“2D”) self-assembled layer on a substrate surface. In other aspects, a monolayer coverage of the particles on the substrate can be in the range of 80% to 100% of the surface of the substrate, although other values may be possible.

In accordance with another embodiment of the present invention, a method for preparing a substrate surface is provided. As will be described, multiple surface treatments can be utilized to form and shape structures or features with dimensions in a micrometer and nanometer scale. In some aspects, described treatments may be applied for purposes of generating textured surfaces.

Producing structures or features on a substrate surface with dimensions in a range including a sub-micron range may involve additive and/or subtractive surface treatments using a number of masking layers. For instance, a surface treatment can include an etching process, such as a plasma etching, a reactive ion etching, an ion milling, or any other physical or chemical etching process, or material removal process. Additionally, forming desirable shapes and profiles for particular structures or features may involve use of isotropic and/or anisotropic treatment techniques. Particularly, an isotropic etching process would be capable of removing a material uniformly, while an anisotropic etching process would remove material preferentially along a specific direction.

Hence, in accordance with aspects of the present invention, preparing a substrate surface can include subjecting the substrate surface and/or materials therein or thereupon to a number of isotropic or anisotropic etching steps, or combinations thereof, for purposes including generating textured surfaces. For example, an isotropic etching can be achieved during a reactive ion etching phase via plasma generated, for example, using trifluoromethane, or a noble gas, or a sulfur halide or, more specifically, sulfur hexafluoride. The plasma may also include oxygen. In addition, an anisotropic etching can be achieved by way of plasma generated using an elemental halogen such as, for example, diatomic chlorine. By way of example, a substrate surface including an etching mask can be subjected to an isotropic etching phase and subsequently subjected to an anisotropic etching phase. However, it may be appreciated that other isotropic or anisotropic etching phases are possible during a surface treatment, and can further be combined or interleaved with other processes or treatments, including material deposition processes. In some aspects, etching processes can partially or entirely remove masking layers or materials comprising a substrate surface.

The size, self-assembly, and etch properties of colloidal particles lend themselves well to substrate surface shaping and profiling at the micrometer and nanometer scale. As such, colloidal particles, shaped, dimensioned, and assembled in accordance with methods described by the present disclosure, may be utilized as masking layers to create structures, features or textures on a substrate surface. For example, a monolayer of silica nanospheres may be applied to a substrate surface prior to a surface treatment process. In this manner, colloidal particle assemblies can be used as etching, or deposition, masks to generate structures, features or textures with sizes in a range between 10 nanometers and 10 micrometers, although other sizes may be possible. However, it may be appreciated that other assemblies, materials, layers, and so on, suitably shaped and dimensioned, may also be utilized as masks. Additionally, one or more masking layers may be applied to the surface of a substrate. Moreover, any of the above-mentioned masking layers can be applied to the entirety of a selected surface, or can be applied to only a portion of the surface. As described, in some aspects, dispersed colloidal particles forming a 2D self-assembled monolayer with coverage in the range of 80% to 100% of the surface of the substrate can be utilized.

It is contemplated that any combination of micro-scale and nano-scale textures may be produced using methods described, to include structures and features of any shape, size and pattern. By way of example, such structures and features can be selected from or include one or more of rods, cones, frustums, pyramids and needles. In some aspects, nano-scale structures may be disposed on micro-scale structures, to include one or more of rods, cones, frustums, pyramids and needles. For example, in some forms, a textured substrate surface may have spiked nano-rod structures. In addition, the textured surface may further include a substantially uniform array or a substantially ordered array. The textured surface may further include silicon nano-structures.

With respect to solar cell applications, the above methods can be applied to produce particular features or structures advantageous in generating a low reflectivity surface. For example, a textured surface may be formed to exhibit a weighted reflectance of less than 5% over a wavelength range of about 300 nanometers to about 1100 nanometers. In this manner, such textured surface can provide both light trapping for enhanced cell efficiency and anti-reflection coating effects. Using a dual-scale texturing process, large scale (micron-scale), and small scale (sub-micron) features may be created in a substrate surface, whereby the micro-scale textures can increase the light path by increasing the number of surface reflections, while the sub-micron scale textures provide a gradient refractive-index at the interface producing an anti-reflection coating effect.

Specific examples are provided below. These example are offered for illustrative purposes only, and are not intended to limit the scope of the present invention in any way. Indeed, various modifications of the invention in addition to those shown and described herein will become apparent to those skilled in the art from the foregoing description and the following example and fall within the scope of the appended claims.

EXAMPLE I

Polished n-type Si (100) round substrates, 2-inch and 4-inch in diameter with 280 μm and 460 μm thickness, respectively, were used to demonstrate the dispersibility of particles on a substrate. As a preparation step, the substrates were cleaned in piranha solution [H₂SO₄ (96%): H₂O₂ (30%)=4:1] for 15-min to form hydrophilic Si surface followed by 10-min de-ionized water (DI-water) rinse.

Next, silica micro-sphere (SMS) solutions for spin-coating were prepared by adding 310 nm-diameter silica microsphere powder to separate solvents, namely N,N-dimethyl-formamide (DMF) and de-ionized water. The solutions were subjected to a sonication for 5 hours to agitate the particles in order to produce complete dispersion of SMS in the solution. The SMS dispersion for each solvent was then characterized by measuring absorbance using ultraviolet-visible (UV-VIS) spectrophotometer. Subsequently, the solutions were spin-coated onto the Si substrates, using a Delta80BN spinner (SUSS MicroTec), under ambient laboratory conditions (21˜23° C. temperature and 25˜30% humidity). A single step of spin-coating recipe was implemented using various acceleration rates (20-80 rpm/s) for 2000 rpm target speed. The SMS spin-coated substrates were subsequently examined by scanning electron microscopy (SEM, JEOL XL-30), and coverage of SMS was calculated using image analysis software, “Image J” (National Institutes of Health, USA) after the direct counting of the SMS area. The wettability of solvent on Si wafer was assessed by measuring contact angle with EasyDrop contact angle measurement system (KRUSS).

To demonstrate the excellence of DMF for 2D colloidal particle spin-coating on large-scale surface area, one of the most widely used solvent, water, has been selected to compare the quality of SMS monolayer assembly layer, and the coverage after spin-coating under ambient conditions.

Dispersibility in Solution

Before performing a spin-coating of particles dispensed in a dispersing medium for use in, for example a lithographical process, a uniform dispersion of particles in the medium may be advantageous to avoid clustered deposition on a substrate surface. Therefore the choice of medium, such as N,N-dimethyl-formamide (DMF), helps to ensure a high dispersibility.

Turning to FIG. 1(a) a plot of absorbance versus wavelength is shown for 0.5 wt % 310 nm SMS dispersed in DMF (denoted as SMS_DMF) and water (denoted as SMS_water). Using a UV-VIS spectrophotometer, normalized absorbance measurements (solid lines) were performed to compare SMS dispersibility for each solvent. Also, extinction cross-sections calculations for a single SMS of the same size (dashed lines) were fitted to the experimental measurements (solid lines). As shown, DMF produces a high level of SMS dispersibility since the measured absorbance (denoted as DMF_UV-VIS) of SMS in DMF has a very well-matched trend with the calculated extinction cross-section (denoted as DMF_calc), indicating SMS_DMF was nearly completely dispersed. By contrast, for SMS_water the experimental absorbance (denoted as WATER_UV-VIS) was substantially broadened compared to calculated extinction cross-section (denoted as WATER_calc). This spectral broadening of absorbance suggests that particles larger than 310 nm are dispersed in the water solution, which appears to result from aggregated SMS. FIG. 1(b) and FIG. 1(c) highlight the dispersivity difference between spin-coated solutions of SMS_DMF and SMS_water, wherein FIG. 1(b) shows the appearance of clusters on the substrate surface.

The existence of a large number of SMS clusters in a dispensed solution helps to produce uniform monolayer deposition during spin-coating. Due to a heavier weight, and hence higher surface friction, clusters can anchor to the substrate surface and consequently act as a flow barrier, preventing uniform distribution of the SMS. This is illustrated in FIG. 2, in which poor SMS dispersibility in a solvent is shown to produce a large, SMS-free area a radial direction induced by the spinning process, and thus significantly affecting the coverage of SMS monolayer. Therefore, a high level of SMS dispersibility in solution is desired to produce a high uniformity and coverage of SMS monolayer following spin-coating.

Wettability on Silicon Substrates

Turning to FIG. 3, the contact angles (a) of DMF and water on Si surface were measured to compare the degree of wettability of each solvent. Comparing FIGS. 3(a) and 3(b), DMF is shown to offer an outstanding wettability [α_DMF≈0 in FIG. 3(b)] compared to water [α_water=26.9° in FIG. 3(a)] on the piranha cleaned Si surface. The importance of solvent wettability is a crucial factor to obtain highly uniform monolayer coverage, the reason being that a high wettability can provide fast, uniform, and omni-directional spread-out during (or even before) the spin-coating step due to a low surface tension. For a wettability comparison between DMF and water, 300 ul volume solutions of 100 mg/ml SMS_water and SMS_DMF were been dispensed on piranha-cleaned 2-inch Si substrates. From FIGS. 3(c) and 3(d), it is clear that the SMS_DMF solution shows a complete wetting layer on the surface once the solution was dropped, whereas the same solution volume of SMS_water produced only a partially-covered surface.

The significance of wettability for spin-coating is illustrated in FIGS. 3(e) and 3(f), where the surface uniformity of SMS_water and SMS_DMF after spin-coating is illustrated. The images clearly illustrate that, unlike water-based solution, a DMF-based solution can produce outstanding uniformity from the center of the substrate all the way to the edge. The significantly improved uniformity can be explained by the low surface tension (γ) of DMF (γ_DMF=25 mN/m) as compared that of water (γ_water=73 mN/m). For high γ solvent (e.g. water), a strong centrifugal force improves solution distribution. However, a strong centrifugal force is achieved by a high rotation speed, which in turn will also cause fast solvent evaporation. A fast evaporation can then remove a significant solvent volume before a uniform SMS distribution can made. As such, formation of non-uniform SMS layer is unavoidable, and thus a low γ solvent is more desirable for uniform SMS assembly.

In addition, a low γ is also beneficial to producing a large SMS coverage. This is because a lower γ solvent, utilizing slower spinning speeds, produces only small amounts of SMS loss in the spinning process. By contrast, a high γ solvent produces a large loss of SMS due to the strong centrifugal force to distribute the solution during spin-coating. This difference is illustrated in FIG. 4, wherein SEM images reveal a very noticeable coverage improvement of SMS_DMF over SMS_water. For this example, the images were obtained at the center of each substrate to exclude possible secondary SMS delivery during solution spreading. Therefore, it is clear that the great wettability of DMF can offer excellent uniformity in the assembled SMS layer, with outstanding coverage and small loss of SMS.

Solvent Evaporation Rate

For highly uniform and closely-packed SMS monolayer formation using a spin-coating approach, two processes should be efficiently taken into account, namely the (1) capillary assembly, and (2) convective flux. Turning to FIG. 5, capillary assembly, which uses capillary forces to organize particles in a suspension, is dominant at the short range because the magnitude of capillary force is inversely proportional to the inter-particle distance, as follows:

F _cap=2πγr_c²(sin²Ψ_c)(1/L)  (1)

where γ is the surface tension of a liquid, r_c is the radius of the three-phase contact line at the particle surface, Ψ_C is the mean meniscus slope angle at the contact line, and L is the distance between the centers of the particles as is illustrated in FIG. 5

To generate a close-packed SMS monolayer over an expanded surface area, a sufficient particle flux provides for uninterrupted growth of SMS assembly monolayer. Therefore, an effective SMS flux should follow initial nucleation of SMS monolayer caused by an initial capillary assembly. For a spin-coating process, this SMS flux may be convective, which originates from the different hydrodynamic forces induced by the variation in wetting layer thickness from substrate center to edge while spinning.

The evaporation of solvent induces a gradual decrease of the wetting layer thickness with time, and until the wetting layer is thicker than the SMS diameter, it decreases evenly over wetting area. Once ordered regions are formed, there would be solvent convective flux from thicker wetting (or disordered) region to thinner wetting (or ordered) region followed by SMS flux. During this convective flux, there is a different evaporation rate between ordered and disordered regions due to the slower solvent evaporation rate in the ordered region caused by hydrophilic property of the SMS. With water, however, its high vapor pressure (VP_water=17.54 Torr at 20° C.) leads a rapid evaporation rate at disordered region during spin-coating, and produces fast reduction of fluid level variation from ordered region to disordered region.

This process is illustrated in FIG. 6. For a solution with a fast evaporation rate, convective flux may only occur for short period of time. As such, an insufficient convective flux as illustrated in FIG. 6(a) leads to only a short-range SMS ordered region, formed through a localized F_cap, as shown in FIG. 6(c). Consequently, in order to achieve long-range SMS assembly, solvents with high vapor pressure (e.g. water, methanol) have required additional treatment involving surfactants or have necessitated systems for temperature and humidity control in order to provide delayed evaporation.

By contrast, DMF, has a slow evaporation rate caused by its low vapor pressure (VP_DMF=2.7 Torr at 20° C.), which can then lead to the long period convective flux needed to deliver sufficient amount of SMS from disordered region to ordered region, shown in FIG. 6(b). Consequently, a long range SMS assembled region can be achieved, with an expanded F_cap on the substrate surface, shown in FIG. 6(d). The slow evaporation rate of DMF, combined with excellent dispersibility and enhanced wetting properties have successfully produced well close-packed SMS monolayer assembly with an outstanding coverage on the surface, as will be shown in following section.

Large-Scale Area SMS Monolayer Spin-Coating with DMF

In addition to the choice of solvent, the solution concentration and spin-coating speeds are also effective parameters in assembling highly uniform SMS monolayers with great coverage on large-scale surface area. Turning to FIG. 7, the effect of SMS_DMF concentration on substrate coverage is illustrated. The figure shows SEM images of spin-coated substrates using 50 mg/ml, 100 mg/ml, and 150 mg/ml of SMS_DMF, without a spin-coating process optimization, namely 20 rpm/s acceleration, 2000 rpm for 150 sec. It can be seen that a higher SMS coverage is achieved with increased concentration, and at 150 mg/ml SMS_DMF, a complete surface coverage was formed after spin-coating. However, with increased concentrations of SMS_DMF a more severe non-uniformity of SMS assembly layer is observed, which may be due to the increased viscosity of solution at higher SMS concentrations. Therefore, a well-optimized spin-coating process is necessary for uniform distribution of SMS over the substrate surface.

Previously, for self-assembled microsphere (MS) monolayer deposition, conventional spin-coating processes involve two steps: (1) a dispersion step at slow speed rotation for uniform MS distribution on the surface, and (2) a drying step at high speed rotation for removing solvent residue and prevent further solvent interaction with MS after spin-coating. This two-step spin-coating process has been developed because the conventional solvents, like water, have a high γ that produce a large loss of MS for high spinning speeds. Moreover, fast evaporation rates at high speeds prevented the manufacture of uniform MS assembly. As such, a slow spin speed dispersion step is needed for increased uniformity with reduced MS loss.

However, with DMF only a one step spin-coating process may be sufficient due to the great wettability and slow evaporation rate. In the example process, target speed is fixed at 2000 rpm, and only acceleration rates are changed, for a total of 150 sec spin-coating duration. Turning to FIG. 8, the effect of acceleration on the formation of a SMS assembly layer is shown. The figure shows SEM figures of coverage obtained using accelerations of 20 rpm/s, 50 rpm/s, and 80 rpm/s. As the acceleration rate is increased, an enhanced surface morphology of SMS assembly layers is observed, whereby at 80 rpm/s acceleration a uniform SMS monolayer is achieved.

Turning to FIG. 9(a), the image of 2-inch substrate is shown after undergoing spin-coating with 80 rpm/s acceleration. Aside from insignificant surface defects, examining the SMS monolayer at high magnification in FIG. 9(d), the overall uniformity is excellent. Moreover, more than 95% of average coverage has been achieved, which is the highest SMS monolayer coverage on 2-inch substrates ever reported by spin-coating.

Turning now to FIG. 10, in order to explore the feasibility of DMF for SMS monolayer assembly on even larger-scale, 4-inch round Silicon substrates were prepared. An identical spin-coating recipe and SMS_DMF concentration was applied to examine the area dependence, and only the solution volume was adjusted to 800 μl to account for the increased surface area. The figure demonstrates a great overall coverage of SMS monolayer on the 4-inch substrate, with great uniformity from center to edge. Although FIG. 10(c) identifies a few areas with relatively low SMS coverage, consideration should be made that the process has not been optimized for 4-in area deposition. As such these issues may be resolved by further adjustment of spin-coating process or solution concentration. Nevertheless, this process provides more than 90% of average monolayer coverage, which is still a superb SMS coverage assembled by a spin-coating process. In addition, it should be highlighted that these results were achieved by spin-coating under common ambient laboratory conditions, without any surfactant mixture or additional treatment on the substrate and SMS. Therefore, the unique solvent properties of DMF are believed to offer a high tolerance to large-scale surface area SMS spin-coating accompanied by outstanding SMS monolayer uniformity and coverage.

In the present invention, we introduced a new organic solvent, N,N-dimethyl-formamide (DMF) for silica microsphere (SMS) monolayer spin-coating on Si surface, which has proven its great potential for high-throughput spin-coating process application leading large-scale area coverage of well close-packed SMS monolayer assembly without any surfactant mixture and environment control during spin-coating. We showed that the DMF can provide outstanding competence to replace conventional solvents, (e.g. water, and methanol) to enhance the uniformity, coverage, and packing of SMS monolayer even under the ambient laboratory spin-coating environment.

From a comparison with water, DMR was shown to offer enhanced properties for spin-coating applications. We demonstrated that DMF allows for well-dispersed SMS in the medium that is close the theoretical limit, which is an important property in producing a uniform SMS distribution on the surface of a substrate. Moreover, the outstanding wettability of DMF forming a thin wetting layer on the substrate surface provides superb coverage of SMS assembly layer compared to same volume and concentration of SMS in water. As such, we have successfully shown more than 90% of SMS monolayer assembly on 2-inch (˜95%) and 4-inch (˜90%) Si substrates without the need for additional surfactant additives nor special environment control (humidity and temperature) for spin-coating. Therefore, it is clear that DMF offers a great potential for high-throughput, easy, and low-cost spin-coating process to produce highly uniform 2D colloidal particle assembly on large-scale deposition area.

EXAMPLE II

Plasma-assisted RIE combines physical and chemical etching which can be done by ion sputtering, and chemical reaction of radicals with target materials. RIE uses a gas glow discharge to dissociate and ionize relatively stable molecules, thereby forming chemically reactive and ionic species. The etching chemistry is formulated such that the reactive and ionic species formed react with the solid surface to be etched to form volatile products.

The processes taking place during RIE process are described below and schematically illustrated in FIG. 11. The key terms used in describing RIE are as follows:

Generation—A glow discharge is used to generate the gas phase etching environment. The gas phase is generated from a suitable feed gas (e.g., SF₆ for silicon etching) by electron-impact dissociation/ionization. The resulting gas phase etching environment consists of neutrals, electrons, photons, radicals (e.g., F*) and positive (e.g., SF⁺⁵) and negative (e.g., F⁻) ions.

D.C. bias formation—A substrate, such as a silicon wafer is placed on a radio-frequency (RF) driven capacitatively-coupled electrode. The electron mobility is much greater than the ion mobility. Therefore, after ignition of the plasma, the electrode acquires a negative charge, which is the direct current (D.C.) self-bias voltage.

Diffusion/forced convection—The transport of reactive intermediates from the bulk of the plasma to the silicon surface occurs by diffusion. Positive ions from the glow region are forced to the substrate surface by way of the D.C. self-bias (negative) voltage and will assist in the etching reaction.

Adsorption—Reactive radicals adsorb on the silicon surface. This step can be strongly enhanced by concurrent ion bombardment which serves to produce active sites as it aids in the removal of, for example, the SiF_x layer which otherwise passivates the Si surface.

Reaction—The desired reaction occurs between the adsorbed species and the silicon. In the case of fluorine-based etching of silicon, chemical reactions between fluorine atoms and the surface spontaneously produces either volatile species, such as SiF₄, or their precursors, such as SiF_x, where x<4. However, in chlorine-based etching, atoms adsorb readily on silicon surfaces, although the spontaneous etch rate is very slow. Ion bombardment makes it possible for adsorbed chlorine atoms to attack the backbones of silicon more efficiently and form a volatile SiCl₄ molecule. This mechanism is called ion-induced RIE.

Desorption—The desorption of the reaction product into the gas phase entails that the reaction product is volatile. Thus, it should have a high vapor pressure at the substrate temperature. Additionally, there should be no deposited blocking film at the surface. The removal of these films can be greatly accelerated by ion bombardment via sputtering. This mechanism is known as ion-inhibitor RIE.

Exhaust—The desorbed species diffuse from the etching surface into the bulk of the plasma and should be pumped out. Otherwise, plasma-induced dissociation of product molecules will occur and re-deposition can take place.

In certain embodiments, prior to etching, a masking layer is applied to a surface of the substrate. This mask can be applied to the entirety of the selected surface, or it can be applied to only a portion of the surface. In one embodiment, silica nanospheres are applied to the surface. The application of the nanospheres can be accomplished using the solvent-controlled spin coating techniques described herein such as in Example I. Based on experimental results, a silica NS solution including DMF provided greater control over NS deposition due to the unique properties of DMF including extremely low contact resistance to the silicon surface and relatively high boiling temperature (b.p.=153° C.) as compared with a conventional solvent such as water.

Silicon surface etching was achieved with using chlorine- and fluorine-containing gases. The use of these gases in RIE provides more control over the surface texturing process. Two categories of RIE include isotropic and anisotropic etching. An isotropic etching process involves the non-directional removal of material from a substrate, whereas anisotropic etching is a more defined, directional etching process. For example, anisotropic etching can involve primarily vertical etching in one-dimension. By comparison, isotropic etching can involve non-specific etching from multiple directions. Fluorine-based plasmas offer high reaction probability with silicon, which is suitable for isotropic silicon etching. In contrast, low reaction probability of chlorine can offer a highly anisotropic etching environment producing surface structures with greater profile control. Therefore, by combination of fluorine and chlorine based RIE processes under various etching conditions (for example, gas flow rate, RF power, pressure) a diversity of nanostructures can be fabricated with relatively easy control.

Referring to FIG. 12, the RIE texturing process has been illustrated with actual SEM images provided for each step. An exemplary substrate (a) including a nanosphere monolayer on a surface of the substrate can be a starting substrate. The first etching step can use an etching gas, such as CHF₃/Ar, to reduce the size of the silicon microspheres (SMS) (SiO₂), thereby forming substrate (b). This step provides an etching path for the following RIE processes. Concomitantly, this step determines the lateral size of the nanostructures. After SMS size reduction, Cl₂ (10 sccm & 30 mTorr) etching is used to form nanopillar structures by intensified anisotropic etching with high RF power (100 watts) as shown for substrate (c). Under continuous etching, however, the increased interparticle distance between SMS on top of silicon nanopillar accelerates the etching rate of silicon from the top down. This results in weakened anisotropic etching, thereby producing a varied etching rate of silicon from top to bottom as for substrates (d) and (e). With extended etching, the SMS can be completely removed and the top of silicon anisotropically etched to produce dual-scale nanostructure as shown for substrate (h). If reduced RF power (50 watts) is applied under the same conditions, then the overall density and energy of the free electrons is decreased. This produces a less negative D.C. voltage causing further weakened anisotropicity and etching rate. Consequently, a super-sharp nano-tip structure can be fabricated as shown for substrate (f). However, a low-aspect ratio nanotip structure can be achieved as in substrate (g) by switching from, for example, a Cl₂ to an SF₆/O₂ based plasma in order to provide intensified isotropic etching. It is noteworthy that SF₆ etching can occur rapidly, so processes incorporating this chemistry can be operated for a shorter period to prevent over-etching and destruction of the desired texturing. Overall, RIE processes with fluorine and chlorine based plasmas can offer excellent control over silicon etching directionality and selectivity to fabricate various desired silicon nanostructures.

Referring now to FIG. 13, the spectral response measurement for reflectivity demonstrates the potential of RIE texturing for reduction of light reflection from a surface. From FIG. 13, the UV-VIS spectrophotometer results show that high aspect ratio (H.A.) nanotip structures result in extremely low light reflection from the surface, and the weighted reflectance (R_w) of H.A. nanotip structures is only 1.4% in the measured wavelength range (λ=300 nm-1100 nm). Table 1 below lists the calculated weighted reflectance based on measured reflectance from FIG. 13. From these results, H.A nano-tip structures, nano-pillars, and dual-scale surface structures are observed to provide improvements in reduction of light reflection from a surface.

TABLE 1
Wavelength
(λ = 0.3-1.1 μm) Rw (%)
Bare-Si          35.7
L.A. Nano-tip    21.4
Nano-pillar      3.0
Dual-scale       2.0
H.A. Nano-tip    1.4

Regarding the application of RIE texturing processes to the manufacture of solar cells, the degree of surface plasma damage can be estimated and a damage recovery treatment can be performed on the Si surface.

Surface passivation also has the potential to further reduce surface reflectance of RIE textured silicon surfaces. A Quinhydrone/Methanol (QHY/ME) passivation technique was applied textured samples in order to evaluate the textured surface for solar cell applications.

Three nominally identical p-type CZ with resistivity of 5-15 Ω-cm and thickness of 450 microns were tested. The samples were subjected to three different surface treatments. Sample A was treated with HF and exposed to air before the measurement of effective carrier lifetime (τ_eff). Sample B was treated with an organic passivation (QHY/ME) layer on both sides, while sample C was RIE textured on one side followed by treatment with an organic passivating layer on both sides. The organic passivating layer in other experiments (FZ, n-type <100>, 100 Ω-cm) exhibited a high effective lifetime (˜3.2 ms) on FZ wafers, implying that the τ_eff for sample B in this experiment is controlled by the bulk lifetime. A τ_eff of 44 μs was measured for sample B, and can be approximately considered equal to the bulk lifetime (τ_bulk) of the wafers. τ_eff for sample A was 4 μs, and this value approximately corresponds to wafers with infinite surface recombination velocity. τ_eff for sample C was 12 μs. Values for τ_eff were calculated using equation 2:

$\begin{array}{cc}\frac{1}{{\tau }_{\mathrm{eff}}}=\frac{1}{{\tau }_b}+\frac{S_{\mathrm{front}}}{W}+\frac{S_{\mathrm{rear}}}{W}& \left(2\right)\end{array}$

where S_rear=0 for a rear organic passivating layer, and S_front=2700 cm/s.

These results show that even the largest surface area of the types of texturing can be passivated using passivation approaches such as organic passivation. The number is likely to be an upper estimate of surface recombination using this the highly surface textured samples because the substrate was p-type, which has a weaker QHY/ME passivation effect.

In conclusion, RIE texturing with silica NS lithography can provide great flexibility on the choice of the shape of texturing on surfaces such as crystalline silicon. In one embodiment, by adjusting the parameters of RIE etching to control the etching properties on silica NS deposited on a silicon surface, diverse shapes and textures were produced in the nano-scale (and submicron) range, which can significantly reduce the loss of incident sun light by reflection. Moreover, with the described solvent controlled spin-coating method, silica NS deposition on silicon surface is achieved in a cost-effective manner. Therefore, the combination of RIE texturing with silica NS lithography provides a cost-effective RIE surface texturing method for producing myriad surface topographies.

For reactive ion etching (RIE) on SMS deposited Si surface, fluorine (F), and chlorine (Cl) based gas was used depending on the desired type of etching (for example, isotropic/anisotropic). During the etching process, gas flow rate, chamber pressure, RF power, and etching time were varied to fabricate the desired shape of nanostructures by controlling the etching selectivity between SMS and Si and the etching direction (i.e., anisotropic, isotropic etching). More specifically, silicon substrates with silica NS monolayers were transferred to the chamber of a reactive ion etcher (PlasmaLab 80+, RIE). The RIE process was set-up with a two or three step process depending on desired shape. However, processes involving less than two steps and more than three steps are anticipated for achieving the desired surface texturing. A CHF₃/Ar gas blend was used for size control of silica NS and for strip-off of the native oxide. Cl₂ gas was also used for anisotropic selective etching and SF₆/O₂ gas for isotropic selective etching between SiO₂ and silicon.

The etched surfaces were observed by scanning electron microscopy (SEM, JEOL XL-30), and surface reflectance was measured with a UV-VIS spectrophotometer. The Sinton lifetime tester, WCT-120, was used for effective carrier lifetime measurement after organic passivation on RIE textured surface.

Although the etching steps of Example II to create dual scale texturing with silica beads are described as being performed after the spin coating steps of Example I, it will be appreciated that the deposition according to Example I could be performed using techniques other than spin coating. Accordingly, the dual scale texturing technique may be performed on previously prepared substrates or substrates having silica beads (and/or other materials) deposited on the substrate by other means.

The present invention has been described in terms of one or more preferred embodiments, and it should be appreciated that many equivalents, alternatives, variations, and modifications, aside from those expressly stated, are possible and within the scope of the invention.

Claims

1. A method of preparing a substrate surface, the method comprising: (a) providing a material comprising an etching mask on a substrate;(b) subjecting the material to a first isotropic etching phase; and(c) subjecting the material to a first anisotropic etching phase,thereby forming a textured surface on the material, wherein the textured surface comprises structures with dimensions in a sub-micron range.

2. The method of claim 1, wherein at least one of the first isotropic etching phase and first anisotropic etching phase comprises reactive ion etching.

3. The method of claim 1, the method further comprising generating a plasma.

4. The method of claim 1, wherein the structures include nano-scale structures disposed on micro-scale structures.

5. The method of claim 1, further comprising subjecting the material to a second isotropic etching phase or a second anisotropic etching phase, or both.

6. The method of claim 1, wherein the textured surface further comprises silicon nano-structures.

7. The method of claim 1, wherein the textured surface further comprises spiked nano-rod structures.

8. The method of claim 1, wherein the textured surface further comprises a substantially uniform array or a substantially ordered array, or both.

9. The method of claim 1, wherein the structures are selected from the group consisting of rods, cones, frustums, pyramids and needles.

10. The method of claim 1, wherein the textured surface is formed to exhibit a weighted reflectance of less than 5% over a wavelength range of about 300 nanometers to about 1100 nanometers.

11. A method of preparing a substrate surface, the method comprising: (a) dispersing a plurality of particles in a suspending medium to form a suspension;(b) spin-coating the suspension on a substrate comprising a material to form an etching mask on the substrate;(c) subjecting the material to a first isotropic etching phase, using the etching mask, forming a modified etching mask; and(d) subjecting the material to a first anisotropic etching phase, using the modified etching mask,thereby forming a textured surface on the material, wherein the textured surface comprises structures with dimensions in a sub-micron range.

12. The method of claim 11, wherein the plurality of particles comprises colloidal particles.

13. The method of claim 11, wherein the suspending medium is selected from a group consisting of tetrahydrofuran, ethyl acetate, acetone, dimethylformamide, acetonitrile, dimethyl sulfoxide, and mixtures thereof.

14. The method of claim 11, the method further comprising generating a plasma.

15. The method of claim 11, wherein the structures include nano-scale structures disposed on micro-scale structures.

16. The method of claim 11, further comprising subjecting the material to a second isotropic etching phase or a second anisotropic etching phase, or both.

17. The method of claim 11, wherein the textured surface further comprises spiked nano-rod structures.

18. The method of claim 11, wherein the textured surface further comprises a substantially uniform array or a substantially ordered array, or both.

19. The method of claim 11, wherein the structures are selected from the group consisting of rods, cones, frustums, pyramids and needles.

20. The method of claim 11, wherein the textured surface is formed to exhibit a weighted reflectance of less than 5% over a wavelength range of about 300 nanometers to about 1100 nanometers.