Energy & Materials · 人与自然 / 能源 / 材料科学 / 气候变化

给太阳能电池镀一层“特氟龙”:27% 的效率,如何换来 1200 小时不衰减

Perovskite solar cells are cheap, printable and nearly as efficient as silicon, yet the chemistry that makes them work also makes them fall apart. A film one molecule thick, modelled on the coating of a non-stick pan, now keeps one of them running at full power for 1,200 hours
来源 Nature Photonics(2025 年 11 月 7 日)作者 外刊精读专栏日期 2026-10-17难度 极难词数 约 530 词 / 9 段
户外太阳能电池板阵列的实景照片(配图)
配图:阳光下成排的太阳能电池板——决定一块电池能用多少年的,往往不是它吸收光的能力,而是层与层相接的那道缝(场景摄影配图,仅供教学使用)。

中文导读

过去十年,太阳能成了能源转型里最成功的故事之一:组件价格一路下探,新增装机年年刷新纪录。但这个故事有个物理上的“主角”始终没变——高纯度硅。它提纯耗能、质地刚硬、只能装在平整的屋顶和支架上。于是人们的目光转向了另一类材料:钙钛矿。这类晶态化合物可以用“墨水”印刷,能做成柔性薄膜,实验室效率从 2009 年的不足 4% 一路追到 27% 以上,几乎逼平硅电池。可它有个致命短板:短命。在光照和高温下,晶体里的离子会游走,表面缺陷不断增殖,上方各层还会与它反应,于是最好的一批早期器件在几百小时内就衰减了五分之一——而商用组件必须撑过二十五年。2025 年 11 月,《自然·光子学》(Nature Photonics)刊出一项研究,把问题定位在一个此前被相对忽视的位置:不是钙钛矿“体内”,而是它与上方电子传输层相接的那道界面。柏林亥姆霍兹中心(HZB)的 Antonio Abate 团队与第一作者李桂香(现为东南大学教授),联合中国、意大利、瑞士、德国的伙伴,用一种名为七氟丁酸钠的小分子在这道缝里铺了一层“近乎单分子”的膜。它的头部咬住表面裸露的铅原子,含氟的尾部朝外,化学性质类似不粘锅涂层——Abate 干脆称之为“特氟龙效应”。结果是效率 27.02%(认证 26.96%),最大功率点连续光照 1200 小时零衰减,85℃ 老化 1800 小时仍保留 92%。本文循“背景—瓶颈—定位—方案—机制—数据—意义—边界—收束”的链条,追问一个工程学上朴素的问题:一种材料的命运,究竟由它本身决定,还是由它的“接缝”决定?

一、原文精读

鼠标悬停在带 蓝色虚线下划线 的生词上,即可查看词性和释义。

🔊 英音朗读 · British English
由英音语音合成(en-GB)朗读全文原文,共 9 段。播放时原文会随进度逐句高亮,建议先盲听抓大意,再对照下方原文跟读,留意重音、连读与停顿。
【1】Solar power has become the cheapest source of new electricity, yet the material at the heart of almost every panel — highly purified silicon — remains expensive to refine, heavy to ship and too rigid to bend around a curve. That is why attention has turned to perovskites, crystalline compounds that can be printed from ink onto flexible film and whose efficiency has climbed from under 4 per cent in 2009 to 27 per cent today.
【2】The problem is that the chemistry which makes perovskites such efficient harvesters of light also makes them fragile. Under heat and illumination, ions wander through the crystal, defects multiply at its surfaces, and the layers above begin to react with it, so that the best early devices lost a fifth of their output in a few hundred hours.
【3】A study in Nature Photonics in November 2025 describes a way of closing that weakness with a film one molecule thick. It was led by Antonio Abate at the Helmholtz-Zentrum Berlin, with Guixiang Li, now a professor at Southeast University in Nanjing, as first author, and drew on partners in China, Italy, Switzerland and Germany.
【4】Their diagnosis was that the decisive damage happens not inside the perovskite but at its upper surface, where the crystal meets the electron-selective layer of carbon-60 molecules that carries charge to the electrode. That junction is studded with defects, and the C60 above it tends to clump when warm, leaving gaps through which ions escape and the metal contact can attack the crystal beneath.
【5】The team's remedy is a short molecule called sodium heptafluorobutyrate, whose head binds to exposed lead atoms while its fluorinated tail, much like a non-stick pan's coating, turns outward. Applied in solution, it settles into an almost continuous monomolecular layer. Abate describes the effect simply: it is, he says, rather like Teflon.
【6】The coating does more than repel water. It also raises the energy needed to create a defect, tunes the surface's work function so that charge leaves more readily, and induces the C60 above it to pack into a denser film that shuts off ion migration.
【7】The resulting cells reached a power-conversion efficiency of 27.02 per cent, certified at 26.96 per cent, with a one-square-centimetre device still delivering 25.95 per cent. More striking is their endurance: after 1,200 hours of continuous illumination at maximum power, they had lost none of their initial output, whereas unprotected cells shed 20 per cent in 300 hours. Aged at 85 °C for 1,800 hours they retained 92 per cent, and 94 per cent after 200 cycles between −40 °C and +85 °C.
【8】Those figures matter because the inverted architecture used here stacks naturally onto silicon, the route most manufacturers expect towards cells above 30 per cent. Yet 1,200 hours is roughly a year outdoors, not the quarter-century a commercial module must survive, and questions of lead content, scale-up and real-weather behaviour remain open.
【9】What the study really shows is that the fate of a promising material can hinge on the seams between its layers rather than on the material itself. Solar power's next leap may come less from discovering new crystals than from learning to seal the ones we already have.

原文出处:本文根据 Nature Photonics 论文《Stabilizing high-efficiency perovskite solar cells via strategic interfacial contact engineering》(Guixiang Li, Zuhong Zhang, Benjamin Agyei-Tuffour, Luyan Wu 等;通讯作者 Antonio Abate,Helmholtz-Zentrum Berlin;Nature Photonics, Vol. 20, 2026-01, pp. 55–62;线上发表 2025-11-07;DOI 10.1038/s41566-025-01791-1)及亥姆霍兹柏林中心(HZB)新闻稿公开资料改写整理,供学习使用。

三、长难句分析(三步拆解法)

Under heat and illumination, ions wander through the crystal, defects multiply at its surfaces, and the layers above begin to react with it, so that the best early devices lost a fifth of their output within a few hundred hours.
主干:三个并列分句 —— ions wander ...(离子游走), defects multiply ...(缺陷增殖), the layers begin to react ...(上层开始反应)
介词短语作状语:Under heat and illumination(在高温与光照下,交代三个动作发生的条件)
so that 引导结果状语从句:the best early devices lost a fifth of their output(“以至于”,说明前面三重退化的最终后果)
注意:so that 在此表“结果”而非“目的”,因为前文是已发生的事实而非意愿
考点:三项并列谓语;so that 结果状语从句;a fifth of 表分数;within + 时间段
译文:在高温与光照下,离子在晶体中游走,表面缺陷不断增殖,上方的各层也开始与之发生反应;以至于早期最好的器件在几百小时内就损失了五分之一的输出。
Their diagnosis was that the decisive damage happens not inside the perovskite but at its upper surface, where the crystal meets the electron-selective layer of carbon-60 molecules that carries charge towards the electrode.
主干:Their diagnosis was that ...(主语 + 系动词 + 表语从句)
not ... but ... 并列:not inside the perovskite but at its upper surface(否定前者、肯定后者,把“体内”与“表面”对立起来)
关系副词 where 引导定语从句:where the crystal meets the electron-selective layer(修饰 upper surface,说明“表面”具体是与谁相接之处)
that 引导定语从句:that carries charge towards the electrode(修饰 layer of carbon-60 molecules,注意谓语用单数 carries,先行词是 layer)
考点:表语从句;not ... but ...;where 定语从句;定语从句谓语与先行词一致
译文:他们的判断是:决定性的损伤并不发生在钙钛矿内部,而发生在它的上表面——也就是晶体与那层由碳六十分子构成、负责把电荷送往电极的电子选择层相接的地方。
It also raises the energy needed to create a defect, tunes the surface's electronic work function so that charge leaves more readily, and induces the C60 above it to pack into a denser film that shuts off ion migration.
主干:It raises ..., tunes ..., and induces ...(三个并列谓语,主语 It 指上文的 coating)
过去分词作后置定语:the energy needed to create a defect(“形成缺陷所需的能量”)
so that 引导目的/结果状语:so that charge leaves more readily(使电荷更容易离开)
induce sb/sth to do sth:induce the C60 to pack into a denser film(促使 C60 排列成更致密的膜)
that 引导定语从句:that shuts off ion migration(修饰 film,说明致密化的作用)
考点:三项并列谓语;分词作定语;so that 状语;induce ... to do;定语从句
译文:它还提高了形成缺陷所需的能量,调节了表面的电子功函数使电荷更容易逸出,并促使上方的 C60 排列成更致密的薄膜,从而阻断离子的迁移。
More striking is their endurance: after 1,200 hours of continuous illumination at maximum power, they had lost none of their initial output, whereas unprotected cells shed 20 per cent in 300 hours.
主干:More striking is their endurance(完全倒装:表语 + 系动词 + 主语,强调“更惊人的是耐久性”)
冒号后的解释:after 1,200 hours ..., they had lost none of their initial output(用具体数据支撑“耐久”)
过去完成时:had lost(以 1200 小时为参照点,回看此前的累计损失)
whereas 引导对比从句:unprotected cells shed 20 per cent in 300 hours(与前者形成强对比)
考点:完全倒装;冒号引出解释;过去完成时;whereas 对比;shed 的“脱落、损失”义
译文:更引人注目的是它们的耐久性:在最大功率点连续光照 1200 小时后,它们没有损失任何初始输出;而未加保护的电池在 300 小时内就衰减了 20%。
Yet 1,200 hours is roughly a year outdoors, not the quarter-century a commercial module must survive, and questions of lead content, scale-up and behaviour in real weather remain open.
主干:两个并列分句 —— 1,200 hours is ... a year outdoors(判断), questions ... remain open(结论)
同位语式否定补充:not the quarter-century ...(用否定短语修正前面的“一年”,把实验室时长与商用要求并置)
省略关系词的定语从句:the quarter-century (that) a commercial module must survive(survive 此处为及物动词,“经历……而存活”)
remain open:“仍悬而未决”,是学术写作中引出局限的常用表达
考点:并列句;省略关系词;survive 的及物用法;remain + 形容词
译文:然而 1200 小时大约只相当于户外一年,而不是商用组件必须撑过的二十五年;含铅问题、大面积放大以及在真实天气下的表现,依然是悬而未决的问题。

四、全文大意

过去十几年,太阳能是能源转型中最成功的故事之一:组件价格一路下探,新增装机年年刷新纪录。但这个故事里有个物理上的“主角”始终没换——高纯度硅。它提纯耗能、质地刚硬、只能装在平整的屋顶与支架上。于是人们的目光转向钙钛矿:这类晶态化合物可以用“墨水”印刷到柔性薄膜上,制备温度低、成本低,实验室效率从 2009 年的不足 4% 一路追到 27% 以上,几乎逼平硅电池。可它有个致命短板——短命。让钙钛矿高效捕获光的那套化学,同时也让它脆弱:在光照与高温下,离子在晶格中游走,表面缺陷不断增殖,上方各层还会与它发生反应;于是早期最好的一批器件在几百小时内就损失了五分之一的输出,而商用组件必须撑过二十五年。2025 年 11 月,《自然·光子学》刊出一项研究,把问题定位在一个此前被相对忽视的位置。研究由柏林亥姆霍兹中心(HZB)的 Antonio Abate 牵头,第一作者李桂香(Guixiang Li,现为南京东南大学教授),合作者来自中国、意大利、瑞士与德国,包括洛桑联邦理工学院(EPFL)与帝国理工学院。他们的诊断是:决定性的损伤并不发生在钙钛矿“体内”,而发生在它与上方电子选择层相接的那道界面——那层由碳六十(C60)分子构成,负责把电荷送往电极。这个界面布满缺陷,而 C60 在受热时容易结块,留下缝隙;离子从缝隙中逃逸,金属接触层也能穿过缝隙侵蚀下方的晶体。团队的补救办法是一种名为七氟丁酸钠的小分子:它的一端咬住表面裸露的铅原子,另一端是含氟的“尾巴”,化学性质类似不粘锅涂层,朝外伸展;以溶液形式涂布后,它会自行铺成一层近乎连续的单分子膜。Abate 把这个效果形容得很简单:就像特氟龙。这层膜的作用不止拒水——它还提高了形成缺陷所需的能量,调节表面电子功函数使电荷更易逸出,并促使上方的 C60 排列得更致密、更均匀,从而阻断离子迁移。数据相当亮眼:转换效率 27.02%(认证值 26.96%),一平方厘米器件仍有 25.95%;在最大功率点连续光照 1200 小时后零衰减,而未加保护的对照电池 300 小时就衰减 20%;85℃ 老化 1800 小时仍保留 92%,在 −40℃ 至 85℃ 之间循环 200 次后保留 94%。这些数字之所以重要,是因为这里采用的反式(p–i–n)结构可以自然地叠在硅电池之上——这正是多数厂商期望突破 30% 效率的现实路线。但作者也把边界说得很清楚:1200 小时大约只相当于户外一年,远不是商用组件要撑过的二十五年;含铅问题、大面积放大以及在真实天气(紫外、湿度、机械应力)下的表现,依然是悬而未决的课题。归根结底,这项研究真正说明的是:一种有前途的材料,其命运可能取决于层与层之间的“接缝”,而不只取决于材料本身。太阳能的下一次跃迁,或许更多来自学会把已有的晶体封好,而不是发现新的晶体。

五、思考与仿写

思考题:文中说,钙钛矿“让光转成电”的能力与它“容易坏”的脆弱性,来自同一套化学性质。请结合第 2 段与第 6 段,说明研究人员为什么不去改造钙钛矿本身,而选择在它与 C60 之间加一层膜?再想一步:这种“不换材料、只改界面”的思路,在生活或工程中还能找到哪些类似的例子(例如防锈涂层、隔热膜、防水透气面料)?它相对“换材料”有什么优势与代价?
仿写题:仿照第 8 段 "Yet 1,200 hours is roughly a year outdoors, not the quarter-century a commercial module must survive, and questions of X, Y and Z remain open."(Yet + 数据对比 + not + 更高标准 + and questions of ... remain open)的句式,为一项你熟悉的技术成果写一句“肯定成绩、同时点明局限”的评价。

六、文章结构图解

文章段落结构(共 9 段) 1 · 背景:硅贵、重、硬;钙钛矿可印刷、可柔性,效率已追至 27% 2 · 瓶颈:同一套化学既让它高效,也让它在光热下迅速衰变 3 · 研究:2025-11《自然·光子学》Abate(HZB)团队 + 李桂香 4 · 定位:损伤不在体内,而在钙钛矿与 C60 相接的上界面 5 · 方案:七氟丁酸钠铺一层“特氟龙式”单分子膜 6 · 机制:提高缺陷形成能、调节功函数、逼 C60 致密排列 7 · 数据:27.02% / 1200 h 零衰减 / 85℃ 1800 h 保留 92% 8 · 意义与边界:可叠硅破 30%,但 1200 h ≠ 商用的 25 年 9 · 收束:材料的命运取决于“接缝”,而非材料本身
全文共 9 段,按「背景 → 瓶颈 → 研究 → 定位 → 方案 → 机制 → 数据 → 意义与边界 → 收束」的递进逻辑层层展开。