p-Type InP Nanopillar Photocathodes for Efficient Solar-Driven Hydrogen Production
p-Type InP Nanopillar Photocathodes for Efficient Solar-Driven Hydrogen Production
复制标题
DOI：
10.1002/anie.201203174
复制
发表时间：
2012-01-01
影响因子：
16.6
通讯作者：
Javey, Ali
中科院分区：
文献类型：
--
作者：
Lee, Min Hyung;Takei, Kuniharu;Javey, Ali
利用太阳光分解水来生产氢，产生可储存的产品，可用作燃料。[1，2]对H2生成，即使用半导体光电阴极在水溶液中将质子还原成H2，存在相当多的研究。[3，4]为了最大限度地提高光电化学（PEC）性能，应仔细考虑活性材料和器件配置的选择。首先，短路电流密度（Jsc）应该通过选择具有高光吸收系数和低载流子复合率的材料来最大化，[5]在体和表面都是如此。应通过使用表面纳米纹理化来最小化反射率以进一步改善光吸收。[6-8]PEC装置的起始电位（Eos）相对于可逆H +/H2氧化还原电位应最大化。最后，需要控制表面能以使气泡在光电极表面上的积聚最小化。具有在1.1 - 1.7eV范围内的带隙的光吸收剂提供了与陆地太阳光谱的良好匹配和分解水所需的1.23eV自由能的显著部分。水溶液中的（溶剂化的）质子应该通过用助催化剂修饰半导体来改善载流子从半导体到电解质的传输而最小化，调谐带边缘和减小接触电阻。p型Si作为光化学制氢的光电阴极已被广泛研究。平面Si在AM1下具有相对低的短路电流密度。5G照明，约10mAcmT2（参考文献[9]），与pn结太阳能电池（> 35mAcmT2）中可以实现的相比。[10]纳米结构化和纳入的助催化剂已被用来提高短路电流密度超过30 mA cm2。[11最近使用n + p Si径向结微丝的研究报道了0.54V的Eos值和15mA的Jsc值，导致接近6%的总效率。[13]到目前为止观察到的p-Si光电阴极的起始电位小于总水分解所需值（1.23 V）的一半。这种低起始电位限制了串联或"Z方案"方法的性能，其将在没有外部偏压的情况下起作用，因为它限制了自发水裂解所需的电位重叠。[14在没有偏压的情况下用于太阳能驱动的氢生产系统中的理想光电阴极应具有高电流密度和相对于可逆H +/H2氧化还原对的有利开路电位两者。在此，我们采用纳米织构的p-InP光电阴极结合的TiO 2钝化层和Ru助催化剂，以增加在H2析出条件下的JSC和EOS值。InP作为光电阴极具有许多吸引人的属性：1）其1.3 eV的带隙与太阳光谱很好地匹配; InP基太阳能电池已实现AM 1。5G效率高达22%。[16]2）InP的导带边略高于水还原电位，有利于电子转移。3)未经处理的InP的表面复合速度低（约1000 nm）。n型为104 cm sector1，p型为105 cm sector1），[17]这对于具有高表面积的非平面器件特别重要，例如本研究中探索的那些。由于这些原因，InP先前已被研究为用于水分解和CO2还原的光电阴极。[18 - 20]具体来说，Heller和Vadimsky报告了具有吸引力的PEC性能，其电流密度高达28 mAcm3/2，InP光电阴极的转换效率约为12%。[19]在这些激励下…
Water splitting by using sunlight for the production of hydrogen yields a storable product, which can be used as a fuel.[1, 2] There is considerable research into H2 generation, namely the reduction of protons to H2 in aqueous solution using semiconductor photocathodes.[3, 4] To maximize the photoelectrochemical (PEC) performance, the selection of the active materials and device configurations should be carefully considered. First, the short-circuit current density (Jsc) should be maximized by choosing materials with high optical absorption coefficients and low carrier recombination rates,[5] both in the bulk and at the surface. The reflectance should be minimized by using surface nanotexturing to further improve light absorption.[6–8] The onset potential (Eos) of the PEC device versus the reversible H+/H2 redox potential should be maximized. Finally, the surface energy needs to be controlled to minimize the accumulation of gas bubbles on the surface of the photoelectrode. Light absorbers with band gaps in the range of 1.1–1.7 eV provide both a good match to the terrestrial solar spectrum and a significant fraction of the 1.23 eV free energy required to split water.Overpotentials associated with the electron transfer to (solvated) protons in aqueous solution should be minimized by improving carrier transport from semiconductor to electrolyte by decorating the semiconductor with cocatalysts, tuning band edges, and decreasing contact resistance. p-Type Si has been extensively investigated as a photocathode for photochemical hydrogen production. Planar Si has relatively low short-circuit current densities under AM1. 5G illumination, approximately 10mAcmÀ2(reference [9]), compared to what can be achieved in a pn junction solar cell (> 35 mA cmÀ2).[10] Nanostructuring and incorporation of cocatalysts have been used to raise the short-circuit current density to over 30 mA cmÀ2.[11, 12] A recent study using n+p Si radial junction microwires reported an Eos value of 0.54 V and an Jsc value of 15mA, leading to an overall efficiency near 6%.[13] The onset potential observed to date for p-Si photocathodes is less than half of the value required for overall water splitting (1.23 V). This low onset potential limits the performance of tandem or “Z-scheme” approaches, which would function without external bias, as it limits the potential overlap required for spontaneous water splitting.[14, 15] An ideal photocathode for use in a solar-driven hydrogen production system without bias should have both a high current density and a favorable open-circuit potential versus the reversible H+/H2 redox couple. Herein, we employ nanotextured p-InP photocathodes in conjunction with a TiO2 passivation layer and a Ru cocatalyst to increase both Jsc and Eos values under H2 evolution conditions. InP has a number of attractive attributes as a photocathode: 1) Its band gap of 1.3 eV is well-matched to the solar spectrum; InP-based solar cells have achieved AM1. 5G efficiencies of up to 22%.[16] 2) The conduction band edge of InP is slightly above the water reduction potential, thus electron transfer is favorable in this system. 3) The surface-recombination velocity of untreated InP is low (ca. 104 cm sÀ1 for n-type and 105 cm sÀ1 for p-type),[17] which is particularly important for nonplanar devices with high surface areas, such as those explored in this study. For these reasons, InP has been studied previously as a photocathode for both water splitting and CO2 reduction.[18–20] Specifically, Heller and Vadimsky reported attractive PEC performances with current densities up to 28 mAcmÀ2 and conversion efficiencies of approximately 12% in InP photocathodes.[19] Motivated by these …