Patent Application
20130298984
The present disclosure relates to silicon surface passivation. More particularly, the present disclosure relates to electronic and optoelectronic devices with passivated silicon surfaces or interfaces.
The performance of many semiconductor electronic and opto-electronic devices is often inversely related to the density of surface states, which are caused by dangling bonds and defects at a surface or interfacial layer. Thermal oxide layers grown at high temperature (>900° C.) are known to cause passivation of such surface states, reducing their density and thereby improving device performance.
The passivation of crystalline silicon (cSi) surfaces by high temperature thermal processing was initially reported by Atalla et al. [1]. The first successful insulated-gate field-effect transistor (FET) utilizing an oxide passivation scheme was fabricated by Kahng et al. [2]. The reduced defect density at the interface between the silicon and its oxide in a sandwich including layers of metal, oxide, and semiconductor (e.g. silicon) (MOS) has made the MOS field transistor (MOSFET), also known as MOS, a widely used semiconductor device. Thermal oxide is also used as an insulator layer for Flash memory semiconductor device which is a MOSFET structure with an additional float gate [26].
Unfortunately, conventional high process temperatures that are employed for the passivation of semiconductor surfaces are incompatible with the production of some high density devices. In particular, as semiconductor devices are scaled down to smaller dimensions, the induced variation in device parameters due to high temperature processing can severely affect the device performance. For example, high-temperature oxidation changes the impurity profiles previously formed in the substrate. Moreover, high stress in oxide films caused by the difference in the thermal expansion of Si and other films causes wafer bending, film cracking, and defect formation in the underlying Si. Accordingly, low-temperature semiconductor device fabrication techniques are needed to prevent such problems.
The formation of silicon gate oxide films using processes which depart from the ‘standard thermal oxidation’ process has been reported. Non-standard thermal oxidation processes, applied on wafers with standard thicknesses, include using a variety of methods such as sputter deposition [3, 4], silicon oxidation in an oxygen plasma [5], thermal oxidation [6], chemical vapor deposition [7], and high-precision-controlled ion assisted process [8]. These techniques, while resulting in interfaces of practical quality, invariably require the use of high temperature treatments or/and are complicated processes.
Surface passivation of cSi is also important for photovoltaic (PV) solar cells in which silicon is used as an absorbing layer. Thermal oxide grown using high temperature processing has been used to fabricate PV solar cells yielding the highest efficiency [9]. Alternative low temperature passivation schemes such as deposition of hydrogenated amorphous silicon (aSi:H) [10, 11], plasma enhanced chemical vapour deposition (PECVD) of silicon oxide (SiOx) [12, 13], PECVD silicon nitride (SiNx) [12, 13, 14, 15, 16] and PECVD silicon carbide (SiCx) [17], have been reported as potential passivation processes for crystalline silicon (cSi) surface.
Among these passivation schemes, PECVD SiNx has been studied extensively. Stoichiometric SiNx is deemed more suitable owing to its lower absorption in the UV region compared to silicon rich SiNx. Further, stoichiometric SiNx has better chemical etching selectivity against cSi than silicon rich SiNx [15], which is desirable where selective etching of SiNx is necessary for device fabrication. High values of fixed positive charge density, Qs, at the cSi-SiNx interface leads to field effect passivation in the case of nearly stoichiometric SiNx. Charge densities of more than 1012 cm−2 have been reported [18, 19]. Unfortunately, an induced inversion layer due to the high trapped charge density can lead to parasitic shunting which reduces the short circuit current [20].
Studies have reported the inclusion of a layer of SiO2 having a thickness on the order of ten nanometers which has been provided between the cSi surface and the SiNx layer, wherein the silicon oxide layer is thermally-grown using high-temperature, [21, 22, 23, 24], in order to reduce or remove the aforementioned parasitic effect. Passivation using thermal oxide (10 nm thick) and SiNx has been reported, with a surface recombination velocity (SRV) of 10 cm/s [21, 22, 23]. Moreover, an SRV of 6 cm/s [19] has been reported for the PECVD oxide (50 nm thick) and SiNx passivation scheme.
Methods, structures and devices are provided in which a crystalline silicon surface is passivated by an ultra-thin silicon oxide layer and an outer passivating dielectric layer, where the ultra-thin silicon oxide layer has a thickness on an Angstrom scale. In some embodiments, both layers are formed by low temperature processes. The outer passivating layer may be formed according to a PECVD process that employs hydrogen-containing precursor gases, such that hydrogen is incorporated into one or both of the silicon oxide layer and the passivating dielectric layer. The present methods may be employed for the passivation of a wide variety of structures and devices, including photovoltaic cells, MOSFET devices, flash memory devices, and thin-film silicon substrates that may contain such devices.
Accordingly, in a first aspect, there is provided a method of passivating a crystalline silicon surface, the method comprising: cleaning the silicon surface and removing a pre-existing native oxide layer; performing a low temperature oxide growth process to form an ultra-thin silicon oxide layer on the silicon surface; and performing a low temperature depositing process to deposit a passivating dielectric layer on the ultra-thin silicon oxide layer.
In another aspect, there is provided a semiconductor structure having a passivated silicon surface, semiconductor structure comprising: a substrate including a crystalline silicon layer; a silicon oxide layer provided on at least a portion of a surface of the silicon layer, wherein the silicon oxide layer has a thickness less than approximately 15 Angstroms; and a passivating dielectric layer provided on the silicon oxide layer.
A further understanding of the functional and advantageous aspects of the disclosure can be realized by reference to the following detailed description and drawings.
Embodiments will now be described, by way of example only, with reference to the drawings, in which:
a) provides an illustration of an example solar cell including a passivation layer including an ultra-thin silicon oxide layer and a passivating dielectric layer.
b) shows an illustration of a unifacial solar cell structure employing an ultra-thin silicon oxide and an outer passivating dielectric according to a dual layer passivation scheme.
c) shows an illustration of a bifacial solar cell structure employing an ultra-thin silicon oxide and an outer passivating dielectric according to a dual layer passivation scheme.
Various embodiments and aspects of the disclosure will be described with reference to details discussed below. The following description and drawings are illustrative of the disclosure and are not to be construed as limiting the disclosure. Numerous specific details are described to provide a thorough understanding of various embodiments of the present disclosure. However, in certain instances, well-known or conventional details are not described in order to provide a concise discussion of embodiments of the present disclosure. It should be understood that the order of the steps of the methods disclosed herein is immaterial so long as the methods remain operable. Moreover, two or more steps may be conducted simultaneously or in a different order than recited herein unless otherwise specified.
As used herein, the terms, “comprises” and “comprising” are to be construed as being inclusive and open ended, and not exclusive. Specifically, when used in the specification and claims, the terms, “comprises” and “comprising” and variations thereof mean the specified features, steps or components are included. These terms are not to be interpreted to exclude the presence of other features, steps or components.
As used herein, the term “exemplary” means “serving as an example, instance, or illustration,” and should not be construed as preferred or advantageous over other configurations disclosed herein.
As used herein, the terms “about” and “approximately”, when used in conjunction with ranges of dimensions of particles, compositions of mixtures or other physical properties or characteristics, are meant to cover slight variations that may exist in the upper and lower limits of the ranges of dimensions so as to not exclude embodiments where on average most of the dimensions are satisfied but where statistically dimensions may exist outside this region. It is not the intention to exclude embodiments such as these from the present disclosure.
As used herein, the term “on the order of”, when used in conjunction with the quantity or extent of a given parameter, refers to a quantity of the parameter ranging between approximately one and ten times the stated parameter.
As used herein, the phrase “low temperature process” refers to a process performed at temperatures substantially below those typically employed for thermal growth of silicon oxide on silicon. According to some embodiments, a “low temperature process” is performed below a temperature of approximately 700° C., while according to other embodiments, a “low temperature process” is performed below a temperature of approximately 500° C. In some non-limiting examples provided below, low temperature processing is performed by processing a passivating substrate, layer or device near or below approximately 400° C.
As used herein, the phrase “ultra-thin” refers to a layer having a thickness on an Angstrom scale.
The phrase “Angstrom scale”, as used herein, according to some embodiments, is within a range of about 0.5 to 10 Angstroms, according to other embodiments is within a range of about 0.5 to 15Angstroms, according to other embodiments is within a range of about 0.5 to 20 Angstroms, and according to yet other embodiments, is within a range of about 0.5 to 100 Angstroms.
The present disclosure provides methods of passivating a surface of a crystalline silicon layer or substrate, in which an ultra-thin layer of silicon oxide, having a thickness of an Angstrom scale, is formed on a crystalline silicon surface, followed by the deposition of a passivating dielectric layer. In some embodiments, the passivating dielectric layer is formed by hydrogen-based precursors, such that during the second deposition step, hydrogen may be incorporated into one or both of the deposited layers. Structures and devices incorporating such a dual layer passivation composition are also disclosed.
As will be described in further detail below, such a process or structure provides a dual layer passivated silicon surface suitable for a wide variety of electronic and opto-electronic applications and devices. In contrary to the usual practice of etching the native amorphous silicon oxide (aSiOx) layer using hydrofluoric acid prior to the deposition of a passivation layer such as SiNx, the presence of a well-developed aSiOx ultra-thin layer reduces the interface dangling bond defect density, thereby enhancing device performance in selected embodiments.
Ultra-thin layer 110, while described herein as a “silicon oxide” layer, is composed substantially of silicon oxide, and need not possess the same stoichiometric relative concentrations of silicon and oxygen as in SiO2. Ultra-thin silicon oxide layer 110 may also include one or more additional substances, the presence of which may contribute to passivation of the silicon surface. For example, in some embodiments, silicon oxide layer 110 includes hydrogen.
The thickness of the ultra-thin silicon oxide layer may be controlled and/or selected by pre-calibrating the deposition rate, and its dependence on one or more process parameters, such as time, temperature, and/or humidity, and/or oxygen partial pressure, and/or UV light radiant intensity, and/or any other suitable process parameter influencing the silicon oxide layer thickness.
The thickness of the passivating dielectric layer may range, in some embodiments, from 0.5 nm to 10 nm, and in other embodiments from 10 nm to 100 nm, and in other embodiments from 100 nm to a few microns. In some embodiments, the passivating dielectric layer is formed by a plasma enhanced chemical vapour deposition (PECVD) process. For example, as described in the examples below, the passivating dielectric layer may be silicon nitride (SiNx), which may be provided by a low-temperature PECVD process involving the hydrogen-based precursor gases NH3 and SiH4. Other suitable, yet non-limiting, dielectric materials include PECVD silicon nitride alloys, PECVD silicon oxynitride alloys, PECVD silicon carbide alloys, PECVD hydrogenated amorphous silicon alloys. Still other example dielectric materials for depositing the dielectric passivating layer include an oxide silicon-based layer SiOx; SiOxNy (silicon oxynitride), and an oxide non-silicon based layer (for example, hafnium oxide). It is to be understood that the dielectric layer may consist of one uniform composition or varying composition or varying sublayers. It is also to be understood that the one or more additional layers may be provided in addition to the passivating dielectric layer. In one example embodiment, the passivating dielectric layer may be formed via a process (for example, PECVD) performed at a temperature ranging from approximately ambient temperature to 400 degrees C. In another example embodiment, the passivating dielectric layer may be formed via a process (for example, PECVD) performed at a temperature ranging from approximately 400 degrees C. to 800 degrees C. In another example embodiment, the passivating dielectric layer may be formed via a process (for example, PECVD) performed at a temperature greater than approximately 800 degrees C.
The silicon surface may be prepared prior to the formation of the ultra-thin silicon oxide layer, where the preparation may include the removal of a pre-existing native oxide layer. Suitable preparation steps include RCA cleaning, dilute HF dip, deionized water rinse, and spin drying. Silicon surfaces can include well-oriented surfaces such as (100) and (111), regular pyramidal surfaces with (111) facets, regular patterned or textured surfaces, random pyramidal surfaces, and random patterned or textured surfaces, as further described below. Furthermore, the silicon surface may be formed from single crystal silicon or polycrystalline (multicrystalline) silicon.
According to some embodiments, the methods disclosed herein provide surface passivation of crystalline silicon wafers through the controlled deposition of an ultra-thin layer of native silicon oxide, which is grown at or near ambient temperature, optionally with controlled humidity, onto which a layer of passivating dielectric is deposited via a low-temperature process, such as the deposition of SiNx by PECVD. In some embodiments, the native oxide layer may be formed over a time duration greater than 50K minutes, while according to other embodiments, the time duration may exceed 100K minutes, and according to other embodiments, the time duration may exceed 150K minutes. The humidity may be adjusted to obtain a given rate of silicon oxide growth. In one example, the humidity is maintained near approximately 50%.The process parameters of the PECVD deposition may be varied in order to obtain a SiNx layer of suitable thickness and composition. For example, in the examples provided below, a SiNx layer is obtained using operating conditions at the low plasma power of 25 W and chamber pressure of 1 Torr. It is to be understood that these parameters are merely provided for illustrative purposes, and that other process parameters may alternatively be selected in order to obtain a SiNx layer (or layer of another suitable passivating dielectric) with a desired thickness.
In other embodiments, the layer of silicon oxide is thermally grown in a low-temperature thermal environment prior to the deposition of the passivating dielectric layer. As in the case of a native oxide layer, the humidity may be selected to obtain a given rate of silicon oxide growth. In one example, the humidity is maintained near approximately 50%.
Although the examples provided below illustrate the formation of ultra-thin silicon oxide layers by native oxide growth in ambient conditions, it is to be understood that the ultra-thin silicon oxide layer may be provided according to a wide range of low-temperature deposition methods, provided that the deposited layer has a thickness on an Angstrom scale.
For example, in an alternative embodiment, the intermediate ultra-thin silicon oxide layer may be formed according to other processes such as, but not-limited to, silicon oxide grown in ozone and other oxidizing chemicals (such as, for example, a solution of H2SO4 and H2O2, a solution of HNO3, or a solution of H2O2). Passivation using an ultra-thin silicon oxide layer grown using ozone and PECVD SiNx may be performed to provide comparable passivation quality. This alternative method is illustrated in the examples below, where it is shown that an SRV of approximately 15 cm/sec was obtained using ozone for silicon oxide growth. Alternatives to these would include growth in an environment where there is catalytic dissociation of oxygen, hot-wire dissociation of oxygen, and provision for oxygen ions and neutrals using other processes. Other oxidation agents, such as HNO3 acid, and a sulfuric acid and hydrogen per-oxide mixture may alternatively be employed to oxidize the surface. Yet another oxidation environment would be liquid water solutions, pure or with appropriate electrolytes or admixture of ozone, at ambient or boiling temperatures or temperatures there between.
In one example implementation, described in further detail in the examples below, PECVD SiNx is deposited at approximately 400° C. on the native aSiOx and grown for various time periods. In one example, the growth of the native oxide layer is performed over four months. Using surface recombination velocity (SRV) as the parameter to characterize the interfacial passivation quality, an SRV of less than 8 cm/s at excess carrier density (ECD) of 1015 cm−3 (7 cm/s at ECD 5×1014 cm−3) is achieved for the passivation scheme. The passivation quality in terms of SRV is similar to the lowest reported values for other passivation schemes. Using the dangling bond interface recombination model [26, 27] and ECD-dependent lifetime measurements, it is shown that the inferred interface defect density NS is as low as 7.2×109 cm−2 and inferred interface surface charge density QS is 4.3×1011 cm−2.
The low-temperature passivation methods, as disclosed in selected embodiments herein, are compatible with a wide range of semiconductor device manufacturing protocols. For example, selected methods disclosed herein are compatible with manufacturing of semiconductor devices, including, but not limited to, metal oxide semiconductor field effect transistor (MOSFET), flash memory (E2PROM), photovoltaic (PV) solar cell and allied microelectronic devices. In particular, MOSFET and flash memory devices are two common electronic semiconductor device types for which the role of a low-temperature grown oxide based dielectric that effectively passivates the cSi surface can be crucial.
In one example embodiment, local layers of doped semiconductor materials may be formed on the dual passivating layer in order to form one or more semiconductor devices. For example, a layer of a first doped material may be deposited over a first region of the passivating dielectric layer, and a layer of a second doped material may be deposited over a second region of the passivating dielectric layer, and a suitable firing process may be performed in order to form a p-type hetero-/homo-junction with the crystalline silicon surface over the first region and an n-type hetero-/homo-junction with the crystalline silicon surface over the second region. One or more of the first doped material and the second doped material may be formed according to a simple printing process, such as ink jet printing or silk screen printing. Example processing conditions for the firing step include thermal processing at a temperature of approximately 750° C. for a time duration of approximately 60 s-120 s. Suitable electrical contacts may then be formed on or augmented on the first region and the second region. The first region and the second region may be suitably spaced to form a semiconductor device, such as a photovoltaic cell, a MOSFET device, a flash memory device, or other semiconductor devices.
In some example embodiments, the passivation methods disclosed herein may be employed to passivate the surface of a wide variety of crystalline silicon-based solar cells, including standard front contact solar cell, back contact solar cells, selective and/or passivated emitter front contact solar cells. Furthermore, the low-temperature synthesized ultra-thin oxide-silicon nitride passivation embodiments provided herein can be employed to fabricate ultra-thin silicon PV cells based on the MIS (metal-insulator-semiconductor) cell configuration.
Referring now to
On the front, direct light-facing side of the wafer, the dual layer passivation scheme provides for both chemical passivation through dangling bond termination and field passivation through the presence of charge in the layer(s), as well as providing a front surface field which serves as a minority carrier mirror. Further, the angstrom thin oxide layer is of negligible thickness to affect optical impedance matching; in this case, the progression of the change in index while progressing from silicon to air is from high to low, consistent with providing for antireflection.
On the rear side of the wafer, an interdigitated pattern of pN and nN junctions and contacts are formed by appropriately opening windows through the dual passivation layer. Specifically, pN junctions 130 are formed by the deposition of p-type hydrogenated amorphous silicon directly onto N-type crystalline silicon and nN junctions 140 are formed by the deposition of n-type hydrogenated amorphous silicon directly onto N-type crystalline silicon. Electrodes 150 are formed by depositing metal contacts directly onto the pN and nN junctions. As noted above, the preceding embodiments, as applied to a silicon solar cell, provide a suitable composition for exhibiting anti-reflection properties. This can be appreciated by comparing an example embodiment of the present disclosure, involving an ultra-thin silicon oxide layer sandwiched between a silicon substrate and a SiNx layer, to known structures in which a thicker silicon oxide layer, having thickness of several nanometers to tens of nanometers or more, is sandwiched between a silicon substrate and a SiNx layer. In the former case, corresponding to an embodiment of the present disclosure, the ultra-thin silicon oxide layer reduces the effective dielectric constant at the interface, thereby enhancing the gradient index effect provided by the transition from the thicker SiNx layer to the silicon substrate.
As described further in the examples below, a cell efficiency of 16.7% was obtained for native oxide/SiNx passivated cells with VOC of 641 mV, JSC of 33.7 mA/cm2 and fill-factor of 0.77 for a 1 cm2 untextured cell (all measurements having been performed under AM 1.5 global spectrum illumination).
Although the preceding embodiments disclose methods and structures in which the ultra-thin silicon oxide layer and the passivating dielectric layer are applied to planar crystalline substrates, it is to be understood that the present disclosure is not intended to be limited to planar substrates, and that the dual passivation layers may be applied to a wide variety of surface geometries.
For example, a common design feature of solar cell substrates is a textured surface, where the texturing is configured to enhance light trapping. An example of such a light trapping surface texturing is the formation of random pyramids, or inverted pyramid structures, which can be formed in silicon structures by appropriate alkaline etching techniques, or appropriate mask-writing and alkaline etching techniques, respectively. Accordingly, in one example implementation, the preceding embodiments may be adapted to be provided on such a textured surface by following appropriate surface preparation steps similar to that of a plane wafer described above.
In other embodiments, the dual layer passivation layer scheme described above may be employed to form a cell device using an amorphous silicon-crystalline silicon heterojunction. According to the present example embodiment, and as further described below, openings may be formed in the dual passivation layer on the front and back surfaces of the substrate. Doped amorphous silicon layers may be formed within the openings, and metal contacts (e.g. metal fingers) may be formed on the doped amorphous silicon layers. Accordingly, devices formed according to such an embodiment include a laterally and inherently thin layer of doped amorphous silicon, and may be referred to as “laterally inherently thin” (LIT) cells.
c) illustrates an example embodiment of such a structure in a unifacial configuration. A crystalline silicon wafer 100, the surfaces of which can be appropriately patterned/textured, is passivated by ultra-thin silicon oxide layer 110, on which is deposited silicon nitride 120, forming the dual passivation layer described above.
On both sides of the wafer, closely spaced windows with small areas and are appropriately opened through the dual passivation layer. In some embodiments, the spacing of the windows may range from approximately a few microns or less, tens of microns or less, or hundreds of microns or less, where the spacing may be determined by the transport properties of the device. For example, the windows may be formed with an 80 um spacing, and a line-width 4 um. In one example implementation, the total fractional area of the openings may be approximately a few percent or less (e.g. approximately 3-7%) so as to minimize shadowing losses. The windows can be employed, for example, to form narrow gridlines and busbar patterns. In the example embodiment shown in the figure, pN junctions 130 and nN junctions 140 are formed by appropriately depositing p-type and n-type hydrogenated amorphous silicon directly onto N-type crystalline silicon, respectively. An appropriate thin intrinsic hydrogenated amorphous silicon layer can optionally be included between the doped amorphous layer and crystalline silicon, for example, with a thickness of approximately 1-10 nm. Metal contact 150 is selectively deposited onto pN junction 130 on the front of the device, and a metal contact 160 is provided on the back surface of the device, thus forming a unifacial solar cell device.
The preceding example embodiment may be beneficial in maintaining high quality surface passivation for regions with no doped layers, while avoiding light induced degradation of the doped amorphous silicon layers. Moreover, there is little or no parasitic optical absorption loss over the AM1.5 solar spectrum for the example device shown in
In
Referring now to
As shown in the Figure, the p-doped regions 130 may be vertically depressed to an appropriate depth in the silicon wafer so as to improve the contact between the p-channel, which is immediately below the oxide layer, and the drain and source. Vertical depression into the silicon wafer can be attained by appropriate etching of silicon; for example, by alkaline etching using KOH. Suitable metal contacts 150 are deposited on top of the doped amorphous silicon regions forming the electrode contacts. Metal contact 150 deposited on the dielectric layer between the drain and the source serves as the gate electrode of the device. Examples of metals for forming metal contacts 150 include one or more of aluminum, nickel, chromium, silver, palladium, platinum, copper, alloys thereof, and layers thereof where specific functionalities are required such as an adhesive/bonding layer or a barrier layer.
In the preceding embodiments, the devices are synthesized using doped variations of hydrogenated amorphous silicon. These devices can also be fabricated using alternative materials, alloys, and combinations thereof. Example alternatives include n-doped and p-doped microcrystalline silicon, hydrogenated amorphous silicon alloys, metals, metal alloys, and other doped dielectric alloys.
In the preceding embodiments, the passivating dielectric layer can further extend to include doped dielectrics where the dopant material would alter its properties. For example, silicon nitride could be doped with an appropriate metallic dopant. In yet another example, the passivating dielectric layer may be a semiconductor such as amorphous silicon or hydrogenated amorphous silicon, which can act as dielectrics. Such semiconductors may also be doped.
The preceding embodiments may also be adapted to provide silicon-based flash memory devices, in which one or more silicon surfaces are passivated according to the dual layer passivation embodiments disclosed herein. It is to be understood that the thickness of the passivating dielectric layer (e.g. PECVD SiNx) may be adjusted according to the requirements of a given device or design.
This passivation scheme disclosed above may be employed for the passivation of devices fabricated in/on thin and flexible semiconductor films, where a dielectric layer providing high quality interfacial passivation grown at low temperatures is needed for such thin silicon wafers. For example, in the context of silicon wafer production, significant progress has recently been made to produce thin silicon wafers with thicknesses of 50 microns, 20 microns, 5 microns, and lower, using kerf-free wafering techniques.
Accordingly, in some embodiments, silicon surfaces of flexible, high-density and low-cost silicon-based devices may be passivated according to methods disclosed herein. In particular, selected embodiments disclosed herein may enable the fabrication of low-cost solar cells due to reduced material costs, at low thermal budget, and/or high photovoltaic conversion efficiency. The integration of high-quality interfacial passivation prepared at low-temperature may be employed to provide efficient light-weight flexible ultra-thin silicon photovoltaic solar cells.
Referring now to
The integration of high-quality interfacial passivation prepared at low-temperature may also be employed in the fabrication of micron- and sub-micron scaled MOS-type and associated devices. Such methods may be compatible with existing synthesis equipment, and may provide low-cost high-density light-weight, flexible, and/or thin silicon microelectronic circuitry/devices.
Referring now to
Gap 420 may be employed for cooling purposes (e.g. active and/or passive cooling). For example, forced air cooling may be employed to provide active cooling of the heat generating microelectronic circuitry/devices residing on the sheets. Gap 420 can equivalently be filled with a suitable thermally conductive material, in solid or liquid form. Non-limiting examples of suitable thermally conductive materials include copper foils, deposited copper films, conducting gels, liquids, and other thermally conductive media. The thermally conductive material provided within gap 420 may be in thermal communication with an external heat sink.
Accordingly, selected embodiments provided herein may be employed to provide high-density electronic, computing and/or optoelectronic devices with high spatial density in three dimensions and suitable passivation of silicon surfaces.
The following examples are presented to enable those skilled in the art to understand and to practice embodiments of the present disclosure. They should not be considered as a limitation on the scope of the present embodiments, but merely as being illustrative and representative thereof.
Double-side polished n-type (100) FZ wafers with resistivity of 1 to 5Ω-cm were used in this passivation study. The wafers were cleaned using standard RCA cleaning steps [28]. The wafers were subsequently dipped in 5% HF for 2 minutes to etch the native oxide and then kept in a clean-room environment, with ambient average humidity of 50% and temperature of 25° C., for different durations; during this period the wafers were stored within a typical 25 wafer carrier box.
The duration for the growth of native oxide layers prior to the deposition of SiNx was varied from 20 min to 40,000 min. The 100 mm circular wafers were cut into four quadrants prior to the deposition of SiNx deposition.
Silicon nitride deposition was carried out using the Oxford PlasmaLab 100 direct rf PECVD system. During a series of preliminary experiments it was determined that higher SiNx deposition temperatures resulted in better passivation, and accordingly all the SiNx depositions were carried out at 400° C. While the use of low chamber pressure (200 mTorr) and high plasma power (100 W) have been reported for SiNx passivation studies using a similar system [15], it was found that such process parameters resulted in significant non-uniformity in film quality.
Instead it was found that operating at a higher chamber pressure of 1 Torr and lower plasma power of 25 W yielded uniform and reproducible films. Further, the use of higher chamber pressure allowed better convective heat transfer from the substrate holder to the wafer and hence required a lower preheating period compared to that at a lower chamber pressure. Operating at higher plasma power increased the deposition rate of SiNx albeit with no significant change in the passivation quality. Accordingly, all SiNx depositions were carried out at the low plasma power of 25 W and chamber pressure of 1 Torr.
Precursor gases for silicon nitride depositions consisted of NH3, which was held constant at a flow rate of 50 sccm, and 5% silane in nitrogen at flow rates ranging from 200 sccm to 550 sccm. The precursor gas ratio (GR), as used herein, is defined as the quotient of (SiH4+N2) to NH3 flow rates.
The effect of varying the GR from 4 to 7 was explored for samples where the native oxide growth time (OGT) ranged from 20 min to 2K min. A small enhancement in passivation was observed for these relatively short OGTs. On the other hand, excellent passivation enhancement in quality was observed for native oxide growth times of the order of 40K min. For these samples, the GR was varied from 4 to 11. For all samples the SiNx deposition time was held constant at 7 min; further, SiNx was deposited on both sides of the wafer to yield a symmetric set of passivation layers. Thicknesses and refractive indices of the SiNx films were measured using SOPRA GES 5E spectroscopic ellipsometer.
The excess carrier density (ECD) dependent effective minority carrier lifetime, τeff, was measured using a Sinton Silicon Lifetime Tester WCT-120 system. Transient and Quasi-Steady-State Photo-Conductance (QSSPC) methods were used to measure injection-dependent τeff of the sample [29]. Spatial distributions of the lifetime for different samples were measured using Semilab's WT-2000 PVN Microwave Photo Conductance Decay (p-PCD) instrument.
In fitting the data obtained from SE measurements, we included an ultra-thin native oxide layer between the cSi substrate and SiNx layer. Silicon nitride layer thickness of ˜93 nm thickness was inferred from the SE measurements which corresponds to a growth rate of 13.3 nm/min. Refractive index (at 633 nm wavelength) increased from 1.9 to 2.06 with increasing silane precursor concentration, i.e., the GR. The increase in the refractive index is consistent with a transition from a sub-stoichiometric to a silicon-rich silicon nitride film [15].
τeff values at an excess carrier density of 1015 cm−3 (unless stated otherwise) were used to calculate the effective surface recombination velocity (SRV), Seff,
S
eff
=W/(2τeff) (1)
where W is the thickness of the sample and the bulk lifetime is assumed to be infinite.
b) shows the passivation quality as a function of native oxide thickness. Native oxide thicknesses are measure using Angled Resolved x-ray photoelectron spectroscopy. Thickness values are of same of order of magnitudes in comparison to the thickness measured in Morita et al.'s investigation of native oxide growth [30] on n-type cSi wafer with doping concentration similar to that used in this study. As seen from the figure increase in the native oxide thickness corresponds to an improvement in the passivation quality. The limiting oxide thickness reflected in
The surface charge density, QS, and defect density, NS, are inferred from the ECD dependent Seff and simple closed form (SCF) dangling bond interface recombination model [26, 27]. The extracted QS and NS values are shown in
A sample with a native oxide growth time of 160K min was also prepared using identical SiNx deposition conditions stated above and at a GR of 7.
Given the observed importance of the native amorphous oxide layer, we also explored the potential of chemically grown oxide playing an equivalent role. Using a 62% nitric acid solution at 60° C., amorphous silicon oxide was grown on a quadrant of a cSi wafer by submerging it in the bath for 60 min. The spatial distribution of the passivation attained, for a process that has yet to be optimized, is shown in
Ozone can also oxidize the silicon at low temperature. Passivation qualities of ultra-thin oxide layers grown in ozone rich environment for different times and PECVD SiNx deposited using the above optimized condition are also studied. Atmospheric oxygen is used for ozone generation using a UV light source.
The passivation scheme is utilized to fabricate back amorphous-crystalline silicon heterojunction [32] PV solar cell. Double side polished n-type FZ wafers with 1-5Ω-cm resistivity were used for the cell fabrication. Doped amorphous silicon layers are deposited at 170° C. using DC saddle field electrode configuration to create the emitter and collector layers. Around 75 nm thick SiNx layers are deposited on the front light-facing side to provide anti-reflective properties. Interdigitated chromium-silver metal layers are deposited on the back-side using e-beam thermal evaporation process. The dark and AM1.5 light IV curves of the back amorphous-crystalline silicon heterojunction cell using native oxide and PECVD SiNx are shown in
The passivating scheme is also implemented on wafers with inverted pyramid textured front surface.
With reference to
As shown in
The specific embodiments described above have been shown by way of example, and it should be understood that these embodiments may be susceptible to various modifications and alternative forms. It should be further understood that the claims are not intended to be limited to the particular forms disclosed, but rather to cover all modifications, equivalents, and alternatives falling within the spirit and scope of this disclosure.
This application claims priority to U.S. Provisional Application No. 61/646,085, titled “PASSIVATION OF SILICON SURFACES USING INTERMEDIATE ULTRA-THIN SILICON OXIDE LAYER AND OUTER PASSIVATING DIELECTRIC LAYER” and filed on May 11, 2012, the entire contents of which is incorporated herein by reference.
| Number | Date | Country | |
|---|---|---|---|
| 61646085 | May 2012 | US |
