New Applications of Electrochemically Produced Porous Semiconductors and Nanowire Arrays
© The Author(s) 2010
Received: 15 April 2010
Accepted: 7 June 2010
Published: 15 June 2010
The growing demand for electro mobility together with advancing concepts for renewable energy as primary power sources requires sophisticated methods of energy storage. In this work, we present a Li ion battery based on Si nanowires, which can be produced reliable and cheaply and which shows superior properties, such as a largely increased capacity and cycle stability. Sophisticated methods based on electrochemical pore etching allow to produce optimized regular arrays of nanowires, which can be stabilized by intrinsic cross-links, which serve to avoid unwanted stiction effects and allow easy processing.
The availability of cheap and reliable secondary batteries for energy storage is required in many fields of present and particular future technology. Demands include storage systems for renewable energy sources, batteries for portable devices like notebooks cell phones and the like, to affordable and powerful electric cars. In state-of-the-art battery concepts, a graphite anode is used together with a cathode based on Li-compounds like, e.g., LiCoO2. To enhance the capacity of the battery, Si is a promising material for the use as anode. Si can incorporate very large amounts of Li, leading to a nominal capacity of about 4,200 mAh/g, about a factor 11 larger than the current state-of-the-art graphite anode . The redox potential of this system is also suitable for the use as anode, i.e. only a small potential is needed to extract the Li from the Si. During the incorporation of Li, phases like Li12Si7, Li7Si3, Li13Si4, and Li22Si5 are formed, and the corresponding phase transitions invariably lead to an expansion of the bulk Si of up to a factor of 4. In bulk Si, the resulting stress is so large that it fractures, preventing the use in a battery. A solution to this problem has been demonstrated by Chan et al. . In this work, Si nanowires have been used as Li host material. The wires still have the superior Li incorporation properties of bulk Si, but also allow for the large volume expansion without fracture, since a free increase in their diameter is now possible. The nanowires in the work of Chan et al. have been grown by a vapor-liquid-solid (VLS) method, using Au droplets on a stainless steel substrate. It was possible to generate nanowires with a diameter around 90 nm, which were able to withstand 10 charging/discharging cycles.
Despite the success of this and follow-up  work, the production method of the nanowires has some drawbacks. Without substantial added process complexity, it is only possible to grow rather unordered arrays of nanowires in a small range of diameters. The optimal geometry of a nanowire array, and thus maximum capacity, might therefore not be achievable. The applicability of the VLS method to large-scale production has not yet been demonstrated and is doubtful. In this work, we therefore demonstrate a cheap and reliable method for the production of optimized nanowire arrays by easy available and fairly routine techniques.
Production of Nanowire Arrays
The production of nanowire array anodes has been facilitated in a three-step procedure. (1) Electrochemical etching of ordered macropore arrays into single-crystalline Si. (2) Chemical over-etching of the macropores until the pore diameters touch and only small wires remain between the pores. (3) Deposition of a Cu diffusion barrier and contact layer.
The third step is necessary to produce a working device. Since the production method yields nanowires that are directly attached to bulk Si, Li would be incorporated during the charging/discharging cycle not only into the nanowires, but also into the bulk Si, if no precautionary measures are taken. Hence, a Cu layer, acting as diffusion barrier, is galvanically deposited at the bottom of the nanowire, i.e. at the nanowire-bulk Si interface. Figure 1c shows this layer schematically.
Electrochemical Macropore Etching
The etching of macropores into n-type and p-type Si is by now an established technique that allows to produce highly ordered arrays of macropores with diameters in the range between 200 nm and 10 μm, and as deep as 500 μm. For detailed reviews of the technique, refer to [4–6].
In this work, macropores have been etched into (100)-oriented p-type Si single-crystalline wafers with specific resistivities in the range of (15–24) Ωcm. As electrolyte, 5 wt% HF diluted in DMF has been used. All experiments have been carried out at T = 20°C, the current density was j = (3–23) mA/cm2, following an optimized profile over time. The total etching time was in the range of 2 h, yielding pore depths of about 150 μm.
Chemical Over-Etching Yielding Nanowires
Increasing the pore diameter by chemical over-etching has been performed by using an acidic etchant consisting of a mixture of HF/HNO3/HAc. Even though this generally isotropic etchant is well studied for “flat” structures, cf. e.g. , it is difficult to over-etch deep macropores homogeneously, since the concentration of the necessary species at the pore tips will always be different from the one at the top. It is only possible to etch the pores homogeneously, if the reactions are not diffusion controlled, necessitating very dilute acids. The drawback then is long etching times, which may take up to 12 h in our case. This might look like a severe obstacle for mass production on a first view, but the problem might easily be overcome by a batch process, i.e. many wafers are etched simultaneously in one (cheap) etching station, cutting down the effective production time.
Isolation by Galvanic Cu Deposition
The galvanic deposition of Cu onto Si is a standard process in IC technology. Nevertheless, as in the case of the over-etching of the pores, deposition of Cu on the bottom of a dense forest of nanowires is not easily achieved, but can be done, as has been demonstrated in [8–10]. As an electrolyte, a mixture of 300 ml H2O, 70 ml H2SO4, 5 g CuSO4, 0.1 g DTAC (1-dodecyl-trimethylammoniumchloride), 0.1 g SPS (Bis-3-sodiumsulfopropyldisulfide, and 0.1 g PEG (Polyethylenglycol) has been used. Experiments have been carried out at T = 20°C under a constant potential of −0.5 V.
Cycling Test of the Anode
Initially the efficiency is low, which is due to conditioning the Si (the initiation of the first phase transformation from Si to Si–Li) and also to the formation of the so-called silicon–electrolyte interface (SEI). The irreversible losses are about 18.8%.
The cycle stability of the anode is very good and always close to 100%. This indicates that the anode is mechanically stable and no nanowires are detached during the cycling. The cell has also been demounted after cycling and inspected by SEM. The structural investigation into the anode (not shown here) validates this statement; the anode is mechanically still intact.
The results are similar to the results obtained for the half-cell set-up. The irreversible losses are small in comparison with other Si anodes, and a good cycle stability is obtained.
The last feature shown in Fig. 5 is the global wedge, which makes the pore slightly bigger with increasing pore depth. This is done to counter the aforementioned diffusion limitation of the chemical over-etching speed at different pore depths since now less material needs to be removed deep down in the pores. Higher concentrations of the etchant are now possible, significantly reducing the etching time and thus production costs.
It has been demonstrated that optimized Si nanowire anodes can be produced by cheap and reliable standard techniques. The resulting structures are suitable for the use as anodes in Li ion batteries of the future. First tests have shown a substantially increased capacity and full cycle stability.
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
- Boukamp BA, Lesh GC, Huggins RA: J. Electrochem. Soc.. 1981, 128: 725. COI number [1:CAS:528:DyaL3MXhs1Cmu7o%3D] COI number [1:CAS:528:DyaL3MXhs1Cmu7o%3D] 10.1149/1.2127495View ArticleGoogle Scholar
- Chan CK, Peng H, Liu G, McIlwrath K, Zhang XF, Huggins RA, Cui Y: Nat. Nanotechnol.. 2008, 3: 1–31. COI number [1:CAS:528:DC%2BD1cXmvVGh]; Bibcode number [2008NatNa...3...31C] COI number [1:CAS:528:DC%2BD1cXmvVGh]; Bibcode number [2008NatNa...3...31C] 10.1038/nnano.2007.411View ArticleGoogle Scholar
- Föll H, Hartz H, Ossei-Wusu EK, Carstensen J, Riemenschneider O: Phys. Stat. Sol. RRL. 2010,4(1):4. COI number [1:CAS:528:DC%2BC3cXivVSjsLc%3D] COI number [1:CAS:528:DC%2BC3cXivVSjsLc%3D] 10.1002/pssr.200903344View ArticleGoogle Scholar
- Kochergin V, Föll H: Porous Semiconductors: Optical Properties and Applications. Springer, London; 2009.View ArticleGoogle Scholar
- Fll H, Leisner M, Cojocaru A, Carstensen J: Materials accepted. 2010.Google Scholar
- Lehmann V: Electrochemistry of Silicon. Wiley-VCH, Weinheim; 2002. 10.1002/3527600272View ArticleGoogle Scholar
- Zhang XG: Electrochemistry of silicon and its oxide. Kluwer Academic—Plenum Publishers, New York; 2001.Google Scholar
- Harraz FA, Kamada K, Sasano J, Izuo S, Sakka T, Ogata YH: Phys. Stat. Sol. (a). 2005,202(8):1683. COI number [1:CAS:528:DC%2BD2MXmtVent7c%3D]; Bibcode number [2005PSSAR.202.1683H] COI number [1:CAS:528:DC%2BD2MXmtVent7c%3D]; Bibcode number [2005PSSAR.202.1683H] 10.1002/pssa.200461226View ArticleGoogle Scholar
- Fang C, Foca E, Xu S, Carstensen J, Föll H: J. Electrochem. Soc. 2007,154(1):D45-D49. COI number [1:CAS:528:DC%2BD2sXht1yntw%3D%3D] COI number [1:CAS:528:DC%2BD2sXht1yntw%3D%3D] 10.1149/1.2393090View ArticleGoogle Scholar
- Fukami K, Tanaka Y, Chourou ML, Sakka T, Ogata YH: Electrochim. Acta. 2009, 54: 2197. COI number [1:CAS:528:DC%2BD1MXhvFajtr4%3D] COI number [1:CAS:528:DC%2BD1MXhvFajtr4%3D] 10.1016/j.electacta.2008.10.024View ArticleGoogle Scholar