单锚定推拉配体把电荷推向界面#

染料敏化太阳能电池和染料敏化光电合成电池靠光敏剂把激发态电子注入TiO₂。常用Dcb配体有两个羧酸锚定，但不少光敏剂注入产率不理想；铁光敏剂更受低注入限制。已有配体设计多关注表面稳定性，少直接增强注入。作者把单羧酸不对称“推-拉”联吡啶接到Ru(dtb)₂上，想在不牺牲水氧化所需Ru(III/II)电位的前提下，把金属到配体电荷转移（MLCT）激发态电荷推向锚定端，提高电子注入量子产率φ_inj。

Fig.1 (Left) Elementary steps for excited state injection emphasizing that injection formally occurs from the surface anchored ligand to TiO2 which motivates the use of asymmetric ligands that polarize charge for enhanced excited state injection, ϕinj. (Right) The family of photosensitizers and their abbreviations used in this study.

甲基版0.90，七成在亚皮秒#

作者合成R为H、CF₃、CH₃的三个推拉光敏剂，与双羧酸RuDcb对照，锚定介孔TiO₂，在0.1 M LiClO₄/乙腈中测瞬态吸收和电化学。φ_inj为RuMcb 0.90、RuDcb 0.80、RuHcb 0.6、RuFcb 0.5，吸电子的CF₃反而最低；Ru(III/II)电位保持在1.41–1.49 V vs NHE。RuMcb总注入中约70%发生在亚皮秒，RuDcb仅约20%；皮秒尺度动力学也显示推拉配体更快。380 nm的MLCT吸收衰减同时漂白增强，支持是注入而非快速复合。计算显示基态HOMO仍在Ru，LUMO和³MLCT在不对称bpy上极化；RuMcb把近24%多余电子放在单一锚定基团，RuDcb两个羧酸合计约23%。单锚定吸附模型给出更短Ti–O、更长Ru–Ti，作者据此认为电子耦合增强；驱动力差约40 mV不足以解释近两倍产率，重组能预期也小。但这些是计算解释，电子耦合未直接测；φ_inj误差±0.1，亚皮秒部分无法时间分辨，靠光谱拆分。

Fig.2 Transient absorption difference spectra measured at the indicated delay times after pulsed 480 nm light excitation of (A) RuMcb|ZrO2 and (B) RuMcb|TiO2 (right) in a 0.1 M LiClO4/CH3CN solution. Fig.3 (A) Absorption difference spectra measured 0.5 to 3000 ps after pulsed 460 nm excitation of RuDcb|TiO2 submerged in 0.1 M LiClO4 acetonitrile. The break in the wavelength axis was to remove the 460 nm excitation. The dashed vertical line at 380 nm shows where the MLCT excited state absorbs light strongly. (B) Absorption difference spectra for the indicated photosensitizers measured 1 ps after 460 nm excitation normalized to the bleach at 450 nm. The inset shows the absorption change monitored at 380 nm with overlaid fits to the Kohlrausch–Williams–Watts model.

设计参数有了，器件还远#

理论上，这把推拉配体几何写成增强界面电子耦合和超快注入的设计参数，并保留水氧化热力学窗口。做染料敏化光电合成电池、染料敏化太阳能电池或铁光敏剂的人，可以试甲基推电子、单锚定bpy，用超快瞬态吸收筛亚皮秒注入。边界是只验证Ru体系、乙腈电解质羧酸盐形式，单锚定长期稳定性未证，原文未提供直接器件转化证据；铁体系和水分解电池是作者提出的未来方向。