A Surface Femtosecond Two-Photon Photoemission Spectrometer for Excited Electron Dynamics and Time-Dependent Photochemical Kinetics

搭建了一套研究金属和金属氧化物表面的超快激发态电子动力学和光化学动力学的飞秒双光子光电子能谱仪. 该装置将半球形电子能量分析仪和成像技术相结合,同时测量光电子的能量和角度分布.通过Mach-Zehnder干涉仪测量时间分辨的双光子光电子能谱获得超快激发电子态的动力学信息. 这一功能在Cu(111)上得到了证实. 另外还发展了一个通过实时测量双光子光电子能谱来研究表面光化学的方法,并成功应用到CH₃CH₂OH/TiO₂(110)体系. 研究表明,只有将两种方法结合起来才能正确地研究光诱导的表面激发共振的动力学.     

乙醇在TiO2(110)表面光催化解离的动力学和超快电子动态学 (cited: 2)

采用实时双光子光电子能谱和时间分辨双光子光电子能谱技术分别研究了乙醇在该表面光催化解离的动力学和超快电子动态学过程. 通过测量与乙醇光催化解离相关的电子激发态随时间的演化,发现这个反应满足分型动力学. 乙醇在还原性TiO₂(110)上的光催化解离比在氧化性表面快,这归结于缺陷的存在降低了反应能垒. 这样一个反应的加速过程很可能是与缺陷电子相关的. 通过干涉双脉冲相关的测量,得到了乙醇-TiO₂界面电子激发态的超快动态学. 与甲醇的情况类似,这个电子激发态的寿命为24 fs. 激发态的出现为TiO₂和它周围环境的电子转移提供了一个通道.     

甲醇在二氧化钛上的光化学性质与表面结构的相关性

通过双光子光电子的方法探测了TiO₂(011)-(2×1) 和TiO₂(110)-(1×1)表面的光催化氧化甲醇的性质. 在吸附了甲醇的二氧化钛(011)和(110)界面处探测到了一个费米能级以上2.5 eV的电子激发态,该电子激发态可作为测试二氧化钛界面还原性的探针使用. 利用此探针在甲醇/TiO₂(011)-(2×1)和甲醇/TiO₂(110)-(1×1)界面探测到了一个随光照时间的电子激发态信号变化,这一变化可以归于光催化生成的表面羟基对界面还原性的影响. 由此得出的光催化氧化甲醇的速率TiO₂(110)-(1×1)比TiO₂(011)-(2×1)快了大约11.4倍. 这可能由于表面原子结构排布的原因不同. 本工作不仅介绍了一个利用双光子光电子能谱探测到的甲醇/TiO₂界面电子结构的细节特征,还揭示了表面结构对二氧化钛光反应性质的重要影响.     

强阿秒XUV激光脉冲下H_2+分子光电子能谱角分布非对称性

利用2D平面模型,求解了描述定向H_{_2}⁺分子和阿秒XUV脉冲相互作用的薛定谔方程,并求得光电子的角度分布．在计算模型中,采用基态1sσ_g和第一激发态2pσ_u的等比例混合态作为初始态,而激光脉冲的光子能量大于电离势,强度为10¹⁴ W/cm²． 计算结果表明,光电子角分布的非对称性和脉冲的宽度密切相关．这种非对称性实际上是由于初始态的基态和激发态的相干振荡而导致的．当使用长脉冲时,这种相干振荡的周期效应就会被平均而消失,从而产生的光电子能谱会呈对称角分布．     

Br2+分子离子[1+1]双光子解离动力学

本文利用最近研制的低温离子阱-离子速度成像谱仪在冷离子束中研究了同位素质量分辨的⁷⁹Br₂⁺分子离子的[1+1]双光子激光解离动力学. 借助其1⁴Σ^-_u,3/2态为中间态使⁷⁹Br₂⁺共振吸收两个光子至4∽5 eV区域的高激发态并发生解离. 利用离子速度成像技术获得了光解产物⁷⁹Br⁺的二维速度分布和平动能释放谱. 通过平动能释放谱确定了不同解离能量处量子态分辨的解离产物通道分支比. 光碎片产物的角分布表明⁷⁹Br₂⁺分子离子的双光子解离是1⁴Σ^-_u,3/2态的ΔΩ=0平行跃迁至一个Ω=3/2高解离态发生的. 由于分子激发态中的强自旋-轨道耦合作用，高激发的四重态很可能参与到实验观测的光解过程.     

3-二甲氨基-2-甲基丙烯醛S2激发态结构动力学 (cited: 3)

采用共振拉曼光谱和完全活性自洽场理论计算研究了3-二甲氨基-2-甲基丙烯醛(DMAMP)光激发到S₂(ππ^*)态后的光物理性能．在B3LYP/6-311++G(d,p)水平计算确定了DMAMP与其三种异构体之间的基态异构化能垒,指认了振动光谱．采用涵盖紫外强吸收带的激光波长,获得了DMAMP在环己烷、乙腈和甲醇溶剂中的A-带共振拉曼光谱,含时密度泛函方法计算确定了该光谱中基频的相对强度,发现振动-电子耦合发生在S₂(ππ^*)态的Franck-Condon区域．CASSCF计算方法确定低单重和三重激发态、势能面锥形交叉点和系间窜跃点的激发能．共振拉曼光谱强度模式分析和CASSCF计算获得了DMAMP的A-带短时结构动力学和其后的衰变动力学表明,C1=O6和C2=C3之间的瞬时去共轭效应发生在S₂(ππ^*)态的Franck-Condon区域,激发态电荷重分布机制表明,C3和二甲氨基之间以及C1和C2之间的共轭增强效应发生在波包离开Franck-Condon区域后．C1=O6和C2=C3之间的去共轭效应使得-C3=N(CH₃)₂沿着C2-C3键旋转更加容易,C1-C2之间以及C3和N(CH₃)的共轭增强效应使得绕C1-C2和C3-N5旋转变得比较困难．这些表明DMAMP初始结构动力学沿着CI-1(S₂/S₀)交叉点展开,而沿CI-2(S₂/S₀)和CI-3(S₂/S₀)交叉点展开的几率可以忽略．提出了DMAMP分子受光激发从S_2,FC(ππ^*)经由各锥形交叉点和各系间窜跃点回到S₀或T_1,min的两个衰变通道.     

钾分子中里德堡态的紫外辐射研究

本文报道了在K-K₂系统中,利用双光子共振激发钾原子8s、6d态,观察到了K₂在(310～365nm)里德堡态的紫外辐射.钾分子里德堡态的布居,受到了多种激发过程影响.文中对其发射和激发过程进行了详细讨论.     

多光子过程和自发辐射对激光诱导自电离的影响

利用反作用算符的方法,研究了一个单模激光场将原子从束缚能级激发到自电离能级的过程中,双光子过程和自发辐射过程的影响。研究结果表明:自发辐射对这种光电离过程的影响一般情况下是很小的,但光子能谱特性十分明确地表明,辐射主要来源于由态共振加强的喇曼过程;并显示了激光诱导自电离过程中的缀饰(Dressed)原子图像;双光子过程将使光电子能谱和光诱导自电离过程变得非常复杂,单光子和双光子过程所产生的光电子能谱,都具有多极大值结构。     

N,N-二甲基硫代乙酰胺S3(ππ*)态的衰变动力学 (cited: 1)

采用共振拉曼光谱和完全活化空间自洽场方法研究了N,N-二甲基硫代乙酰胺在被激发至S₃(ππ^*)态后的衰减动力学. 指认了紫外吸收光谱和振动光谱. 获得了乙腈、甲醇和水溶剂中不同激发波长下的A带共振拉曼光谱,以探测Franck-Condon区域的结构动力学. 开展了CASSCF计算以确定低能单重激发态和锥形交叉点的电子激发能和优化几何结构. 通过共振拉曼强度分析和CASSCF计算获得了结构参数、A带结构动力学和S₃(ππ^*)态衰减机制. 提出了主要衰减通道为_3,FC(ππ^*)→S₃(ππ^*)/S₁(nπ^*)→_1(nπ^*).     

不同初始振动态下D2O在B带的态-态光解动力学

利用高精度的非绝热从头算势能面,采用量子实波包方法研究了D₂O在3个不同初始振动态(0,1,0)、(1,0,0)和(0,0,1)下光激发到~B态解离过程的态-态量子动力学．结果表明,由于初始振动波函数的形状不同,从不同振动态产生的吸收光谱、产物分布和产物分支比具有不同的动力学特性．弯曲振动态(010)的吸收谱呈现出两波瓣的形状且中间具有较浅的极小值,其基态产物OD( ~X)在高能量区域会出现轻微的振动态反转现象．所有初始振动态得到的激发态产物OD(_?)的振-转态分布都随能量的变化出现强烈的震荡,这是由于受到了激发态势能面上长寿命共振态的影响．反对称伸缩振动态(0,0,1)在高能量区域具有较大的产物分支比OD(_?)/OD( ~X),表明绝热通道在某些情形下是主要的解离通道．     

CH3OH/TiO2(110)界面的光电子能谱研究

Methanol/TiO₂(110) is a model system in the surface science study of photocatalysis where methanol is taken as a hole capture. However, the highest occupied molecular orbital of adsorbed methanol lies below the valence band maximum of TiO₂, preventing the hole transfer. To study the level alignment of this system, electronic structure of methanol covered TiO₂(110) surface has been measured by ultraviolet photoelectron spectroscopy and the molecular orbitals of adsorbed methanol have been clearly identified. The results indicate the weak interaction between methanol and TiO₂ substrate. The static electronic structure also suggests the mismatch of the energy levels. These static experiments have been performed without band gap excitation which is the prerequisite of a photocatalytic process. Future study of the transient electronic structure using time-resolved UPS has also been discussed.     

Dynamics of photoinduced electron transfer from adsorbed molecules into solids

Ultrafast interfacial electron transfer from the donor orbital of organic chromophores into empty electronic acceptor states of a semiconductor and of a metal was investigated by two-photon photoemission spectroscopy (2PPE). Experimental tools and procedures have been developed for carrying out wet-chemistry preparation of the molecule/solid interface. The organic chromophore perylene was investigated with several different bridge/anchor groups on TiO₂(110). One perylene compound was investigated for comparison on Ag(110). Angle and polarization dependent 2PPE measurements revealed the orientation of the perylene chromophore on the surface as controlled by the adsorption geometry of the respective anchor group on TiO₂. UPS measurements gave the position of the HOMO level of the chromophore with respect to the Fermi level of the solid. The donor level of each molecule was found high enough to fulfill the “wide band limit” of heterogeneous electron transfer dynamics. Time constants for heterogeneous electron transfer were extracted from 2PPE transients. A difference by a factor of four was found, 13 fs against 47 fs, when a conjugated bond was exchanged for a saturated bond in the otherwise identical bridge group. The two different contributions to the 2PPE transients arising firstly from the excited state of the chromophore and secondly from the injected electrons were separated by measuring the latter contribution separately in the case of instantaneous interfacial electron transfer realized with catechol as adsorbate. The time scales measured for the electron transfer step and for the subsequent electron escape process from the surface into the bulk of TiO₂ showed both good agreement with recent theoretical predictions of other groups for these systems. PACS 42.65.Ky; 79.60.BM; 78.47.+p; 73.20.-r; 79.60.Jv; 79.60.-i     

FePc与TiO2(110)及C60界面电子结构研究

基于C₆₀受体和有机分子给体的太阳能电池是目前非常重要的一个研究热点, 利用同步辐射真空紫外光电子能谱(SRUPS) 技术研究了酞菁铁(FePc)与TiO₂(110)及C₆₀的界面电子结构, 以及FePc与C₆₀分子混合薄膜的电子结构. SRUPS价带谱显示, FePc沉积在化学计量比与还原态两种不同的TiO₂(110)表面时, FePc分子的HOMO能级均随FePc厚度的变化发生了移动, 而在化学计量比的TiO₂(110)表面位移较大, 同时发生界面能带弯曲, 说明存在从有机层向衬底的电子转移. 在FePc/C₆₀和C₆₀/FePc界面形成过程中, FePc与C₆₀分子的最高占据分子轨道(HOMO)位移大小基本相同. 由界面能级排列发现, 在FePc与C₆₀的混合薄膜中, FePc分子的HOMO与C₆₀分子的最高占据分子轨道能级差较大, 这有利于提高器件开路电压, 改善器件性能.     

Photoemission study of the modification of the electronic structure of transition-metal overlayers on TiO2 surfaces II. Cr on TiO2(0 0 1)

The electronic states of the Cr overlayers on TiO₂(0 0 1) surfaces have been investigated using angle-resolved and resonant photoemission spectroscopy with synchrotron radiation. At lower coverages, Cr deposition on TiO₂(0 0 1) creates two well separated in-gap emissions due to the formation of surface Ti³⁺ (3d¹) ions and Cr³⁺ (3d³) ions. At higher coverages, the in-gap emission is developed into the 2-peak-structure emission of Cr 3d character. The corresponding state is considered to be of metallic nature from the viewpoint of the high ability of oxygen adsorption, but has no Fermi edge, indicating a possibility of forming small Cr clusters on TiO₂(0 0 1) at this stage.     

甲醇分子在金红石TiO2(110)和锐钛矿TiO2(001)表面吸附和反应的活性位点

通过高分辨的扫描隧道显微术研究并比较了金红石型TiO₂(110)-(1×1)和锐钛矿型TiO₂(001)-(1×4)两种表面的活性位点． 在金红石型TiO₂(110)-(1×1)表面, 观察到氧空位缺陷是O₂和CO₂分子的活性吸附位点,而五配位的Ti原子是水分子和甲醇分子的光催化反应活性位点．在锐钛矿型TiO₂(001)-(1×4)表面,观察到完全氧化的表面,Ti原子更可能是六配位的,H₂O和O₂分子均不易在这些Ti原子上吸附．经还原后表面出现富Ti的缺陷位点, 这些缺陷位点对H₂O和O₂分子表现出明显的活性． 锐钛矿型TiO₂(001)-(1×4)表面的吸附和反应活性并不具有很高的活性,某种程度上其表现出的活性似乎低于金红石型TiO₂(110)-(1×1)表面．     

Electronic exchanges between adsorbed Ni atoms and TiO2(1 1 0) surface evidenced by resonant photoemission

Nickel was deposited on stoichiometric TiO₂(1 1 0) surface in the 0.02–2.1 equivalent monolayer (eqML) range and analyzed by means of photoemission and resonant photoemission. In the case of very low coverage (lower than 0.1 eqML), deposited nickel reacts with the surface through an electronic transfer from nickel atoms towards titanium ions. This exchange caused the filling of unoccupied Ti3d states leading to the increase of a peak in the TiO₂ band gap. These states can be better characterized through resonant photoemission experiments at the Ti 3p → 3d absorption edge: for very low coverage, these states in the TiO₂ band gap have resonant behavior of Ti3d electrons rather than Ni3d ones, confirming the filling of Ti3d states and thus electron transfer between nickel and titanium. For coverage higher than 0.14 eqML, nickel peaks (both Ni3p core level and valence band) should be related to the presence of metallic nickel in small clusters.     

The interaction between hexahydroxytriphenylene and the rutile TiO2(110)-(1 × 1) surface at UHV conditions

The interaction between 2,3,6,7,10,11-hexahydroxytriphenylene (HHTP) and the rutile TiO₂(110)–(1 × 1) surface under ultrahigh vacuum (UHV) conditions was investigated using X-ray photoemission spectroscopy (XPS), ultraviolet photoemission spectroscopy (UPS), near-edge X-ray absorption fine structure (NEXAFS) spectroscopy, and density functional theory (DFT) calculations. The NEXAFS results showed that HHTP molecules formed a submonolayer and a monolayer that aligned along the [001]-direction with, respectively, a more or less flat downward orientation and a more upright orientation to the TiO₂ surface. The HHTP molecules that aligned along the [001]-direction were most likely grafted onto the TiO₂(110) surface by a bidentate bridge between each of the oxygen atoms of one of the catechol units within the HHTP molecule and two adjacent Ti(5f)⁴⁺ ions on the TiO₂(110) surface. The coordination is non-dissociative in the case of the submonolayer, but dissociative in the monolayer, according to the analysis of the C1s XPS, UPS, C1s NEXAFS data and complementary DFT calculations.