武汉海山科技有限公司

突破35%效率天花板：钙钛矿/硅叠层，真正的胜负手在“界面”

不是简单把钙钛矿“贴”到硅上，而是把纹理化三维埋底界面当成一个耦合系统来设计

硅电池快“撞墙”了，叠层为什么是答案？ 
晶硅电池统治了全球光伏市场95%以上，但单结晶硅的效率正在逼近理论极限：实验室效率已超过27.9%，距离约29.4%的Auger极限越来越近。要继续降本增效，靠单结硅“修修补补”不够了。
钙钛矿/硅叠层太阳能电池（PSTSC）因此成为最受关注的路线：用宽带隙钙钛矿做顶电池吸收短波光子，用硅底电池吸收近红外透射光，理论效率可到约43%。认证效率已经突破35.2%，正式跨过35%里程碑。
图片1

Device architecture, working principle, evolutionary trajectory, and key challenges of PSTSCs. (a) Schematic illustration of a state-of- the-art monolithic 2T PSTSCs. The device integrates a double-side textured silicon heterojunction (HJT) bottom cell with a perovskite top cell, utilizing TCOs and SAMs for efficient electrical interconnection. The layer stack depicts the complete integration scheme, ranging from the rear silver electrode and silicon bottom cell to the intermediate ICL, wide-bandgap perovskite subcell, and antireflection coating. (b) Comparison of spectral utilization and energy loss mechanisms between single-junction and tandem solar cells. The diagram highlights thermalization losses of high-energy photons and transmission losses of low-energy photons in single-junction devices. In contrast, the tandem architecture achieves spectral splitting by stacking a wide- bandgap top cell with a narrow-bandgap bottom cell, effectively minimizing thermal losses and enhancing energy conversion efficiency. (c) Evolution of PCE for crystalline silicon (c-Si), single-junction perovskite, and PSTSCs. The trend illustrates the rapid progress of tandem technology in surpassing the single-junction Shockley–Queisser (S-Q) limit, with the current record efficiency reaching 35.2%. (d) The “Golden Triangle”of photovoltaic technologies. This schematic weighs the commercial potential of c-Si, single-junction perovskite, and PSTSCs across three dimensions: manufacturing cost (1/Cost), theoretical efficiency, and operational lifetime. (e) Detailed power loss breakdown analysis for high-efficiency PSTSCs. By comparing actual device performance against theoretical limits, losses are quantified into current transport losses, non-radiative recombination losses, and parasitic absorption and reflection losses, identifying critical pathways for further efficiency enhancement.

但问题也来了：35%之后，继续往上靠什么？

这篇综述给出的核心答案是：界面，尤其是纹理化硅表面上的三维埋底界面。 它同时决定电荷复合、光学管理、钙钛矿结晶和长期稳定性。换句话说，叠层电池不是“材料堆叠”，而是一个被同一个粗糙界面串起来的耦合系统。

01 先把账算清楚：35%到43%，损失在哪四类？

综述用一个“四类损失预算”框架，把当前效率与理论极限之间的差距拆开：
热化与透射损失：高能光子多余能量变热，低能光子穿透不发电，属于双结架构的本征上限；
非辐射复合损失：异质界面和互连层中的缺陷复合，是VOC损失的主要来源；
寄生光学吸收与反射损失：互连层、缓冲层、透明电极“吃掉”本该给底电池的近红外光子；
接触与串联电阻损失：接触电阻把光电流变成焦耳热，直接压FF。
这四类损失都对应到二端（2T）叠层器件的具体结构。2T架构少一组电极、寄生吸收低、能与现有硅产线结合，但代价是严格的电流匹配约束：
JSC,tandem=min⁡(JSC,topJSC,bot)。
顶底电池电流不平衡，性能就被“短板”锁死。
所以，2T叠层的两大瓶颈很明确：
互连层ICL：要电气连接好、近红外透明、还能贴合粗糙硅表面；
界面钝化与结晶控制：在微米级金字塔纹理上，钙钛矿要覆盖好、结晶好、缺陷少。

02 互连层进化史：从“隧穿”到“欧姆接触”

ICL是顶电池和底电池之间的“心脏”。它要同时满足三个互相打架的要求：低复合电流、低近红外吸收、低接触电阻。综述把ICL路线分成几代：
Gen 1：a-Si:H隧穿结
早期路线，工艺兼容性好，但非晶硅缺陷多，主要靠陷阱辅助隧穿（TAT），接触电阻高；为降电阻提高掺杂，又带来严重近红外自由载流子吸收。效率约13.7%时代。
Gen 2：nc-Si:H/nc-SiOx
用纳米晶硅的垂直渗流导电和nc-SiOx折射率梯度，部分解耦“导电—透明”矛盾，全纹理器件做到25.2%。
Gen 3：Poly-Si隧穿结
重掺杂多晶硅实现带间隧穿（BTBT），需要有效掺杂≥1020 cm⁻³，接触电阻显著降低；与TOPCon产线兼容，TOPCon叠层做到29.2%。但高温800°C以上工艺与HJT的≤200°C热预算不兼容，且多晶硅自身仍有近红外吸收。
Gen 4/4+：TCO/SAM混合接触
这是当前效率纪录的关键。高迁移率非晶IZO降低自由载流子吸收；自组装单分子层SAM通过分子偶极实现功函数调控和费米能级去钉扎，形成近欧姆接触。接触电阻可到sub-10 mΩ·cm²，FF超过83%，实验室器件效率达34.58%。
一句话：硅基隧穿结在“掺杂—吸收”之间左右为难；TCO/SAM用分子偶极把接触电阻和吸收解耦了。
图片2
Electrical contact optimization and light management strategies for the intermediate ICL and transparent electrodes. (a) Evolution of carrier transport mechanisms in Poly-Si tunneling recombination layers. As doping concentration increases, the dominant transport mode transitions from trap-assisted tunneling to BTBT. This transition is critical for maximizing the FF and overall device efficiency. (b) Comparison of specific contact resistivity for different silicon-based contact layers. Compared with amorphous silicon ( a-Si:H ), the use of nanocrystalline silicon ( nc − Si:H / nc − SiOx :H ) as carrier-selective contacts significantly reduces contact resistance with various TCOs (including indium-free TCOs), thereby improving the FF. (c) Optimization of band alignment at the contact interface between TCOs and SAMs. Compared with crystalline ITO or IZrO, amorphous IZO combined with SAMs exhibits a larger work function shift and a more uniform surface potential distribution, enhancing selective hole extraction. (d) Band alignment diagram of a novel gallium and titanium co-doped indium oxide (IO:GT) transparent electrode. IO:GT possesses a shallower work function than traditional ITO, which helps suppress the formation of Schottky barriers at the electron transport layer interface, thereby improving charge collection. (e) Comparison of optical transmittance spectra between IZO and ITO films. IZO demonstrates superior transmittance in the long-wavelength (near-infrared) region compared with ITO, which is beneficial for reducing parasitic absorption and increasing the photocurrent response of the silicon bottom cell. (f) Schematic of refractive index engineering for the intermediate ICL. Inserting a nc-SiOx layer with an intermediate refractive index between the perovskite top cell and the silicon bottom cell establishes a refractive index gradient, effectively reducing optical reflection losses at the internal interface.

图片3

Technological evolution and multidimensional performance evaluation of the ICL in PSTSCs. a, Device structure of early monolithic tandem cells based on a silicon homojunction tunnel layer. This design utilizes a heavily doped n ++ /p ++ silicon tunnel junction to achieve majority carrier recombination between the top and bottom cells. b, Microstructural characterization of a nc-Si:H recombination junction. Cross-sectional scanning transmission electron microscopy (STEM) and energy-dispersive X-ray spectroscopy (EDX) maps clearly display the interface and grain growth of the p + /n + nc − Si:H junction, a structure that effectively reduces parasitic light absorption in the intermediate layer. c, Doping distribution characteristics of a high-efficiency Poly-Si tunneling recombination layer. Secondary ion mass spectrometry (SIMS) profiles show minimal interdiffusion of dopants (boron and phosphorus) within the poly-Si(p + )/poly-Si(n + ) stack, forming a steep and high-quality tunnel junction interface. d, Comparison of surface potential uniformity for different TCOs ICL. Kelvin probe force microscopy (KPFM) images reveal that, compared with polycrystalline ITO or IZrO, amorphous IZO films exhibit superior surface potential uniformity after adsorbing SAMs, helping to reduce interfacial recombination losses. e, Multidimensional radar chart evaluation of ICL material performance. The chart analyzes the trade-offs among different interconnection technologies across five key dimensions: conductivity (measuring vertical conduction and contact resistance), optical transparency (measuring parasitic absorption), passivation capability (measuring the VOC contribution to the bottom cell), low-temperature processing compatibility (measuring the impact of fabrication temperature on underlying layers), and cost/resource abundance (measuring material cost and the usage of scarce elements like indium).

03 SAM不是“涂一层”这么简单：五步因果链

很多文章把SAM说成“界面修饰”，但这篇综述把它拆成五个顺序耦合的物理步骤：
锚定几何：磷酸基团在TCO表面形成P–O–金属键，决定分子倾斜角；
表面覆盖：TCO表面羟基密度、晶界、金字塔形貌决定SAM覆盖均匀性；
分子偶极：分子本征偶极的垂直分量产生偶极电势；
功函数偏移：偶极层改变TCO/SAM有效功函数；
费米能级去钉扎与欧姆接触：SAM化学钝化界面态、屏蔽金属诱导间隙态，使费米能级重新跟随功函数，形成真正欧姆接触。
这条链的关键洞察是：大偶极分子不一定好。 如果分子太“胖”，在金字塔尖覆盖不足，第二步失败，后面全白搭。非晶IZO比多晶ITO有更高表面羟基密度、没有晶界打断锚定，因此SAM覆盖更均匀。HTL201这类不对称分子设计，就是针对金字塔尖空间位阻做优化，把尖端覆盖和功函数均匀性一起改善。
KPFM图不是直接测“覆盖密度”，但在分子偶极固定时，表面电势均匀性可作为覆盖均匀性的代理指标；在纹理表面还要排除形貌对电容梯度的假信号。这个细节很重要，因为大面积均匀性正是TCO/SAM路线最大的未解问题之一。

04 纹理硅：机会和麻烦，来自同一个金字塔

硅表面金字塔纹理是叠层电池的“主变量”：
光学上：金字塔把光散射到朗伯极限，路径增强可达约520Lphys，让更薄的硅也能吸收近红外；
电学上：金字塔尖是应力集中点、SAM覆盖薄弱点、漏电通道起点；
结晶上：金字塔尖曲率凸、前驱体薄、成核密度高、晶粒小；谷部前驱体厚、成核少、晶粒大，但谷部 coalescence 形成高角晶界网络；
稳定性上：尖端同时叠加高晶界密度、CTE失配应变、局域光子通量，成为Hoke相分离的“自加速”位点。
所以，纹理不是可以忽略的衬底条件，而是编码了上层每一层质量的主动工程表面。 想同时提高效率、稳定性和大面积良率，必须做“界面协同设计”，不能只优化单一材料。

图片4
Perovskite film growth control on textured surfaces, interfacial molecular interactions, and energy band engineering. a, Cross-sectional scanning electron microscopy (SEM) image of a perovskite film deposited on a silicon bottom cell with large-sized pyramid texturing. The image clearly reveals that the tips of the silicon pyramids puncture the perovskite layer (Sample D), leading to severe electrical short circuits. The highly non-uniform film thickness (ranging from approximately 2 μm in the valleys to 0 μm at the tips) highlights the limitations of conventional processes in achieving coverage on rough surfaces. b, Cross-sectional morphology of high-quality, micrometer-thick perovskite films on textured silicon substrates. By utilizing a chemically insoluble, ligand-bridged electron-selective contact (C60-anchored a-NbOx) and high-concentration precursors, this process achieves complete conformal coverage of the perovskite layer on micron-scale pyramid textures, eliminating interfacial voids and shunt paths. c, Schematic of the anchoring mechanism of SAM on the surface of amorphous IZO transparent electrodes. It illustrates the synergistic effect of coordinate covalent bonds formed between phosphonic acid groups and the metal oxide surface, alongside hydrogen bonding. d, Adsorption configurations of different hole-selective SAM molecules (Me-4PACz, MeO-4PACz, and the asymmetrically designed HTL201) on the perovskite surface calculated via density functional theory (DFT). The results indicate that the asymmetric molecular design minimizes steric hindrance and enhances binding energy, thereby optimizing interfacial contact. e, Band alignment and hole transport mechanism at the perovskite/ITO contact interface. A comparison of band bending for the unmodified interface, conventional SAM modification, and molecular doping (DMACPA) reveals the physical mechanism of efficient hole extraction achieved through Fermi-level pinning. f, Spatial photoluminescence (PL) imaging of perovskite films under different interface passivation strategies. Compared to untreated or single-treatment samples, films employing the bilayer interwoven passivation strategy (LiF/EDAI) exhibit the highest luminescence intensity and spatial uniformity, confirming the effective suppression of non-radiative recombination.

05 稳定性：真正的量化缺口是60–110 meV

综述把离子迁移和相分离稳定性放进Arrhenius框架：寿命T90与激活能Ea指数相关。当前高性能宽带隙钙钛矿的Ea最好约0.41 eV，而25年商业寿命目标约需0.47–0.52 eV，缺口60–110 meV。在85°C加速条件下，Ea提高67 meV，寿命可提升约8.5倍。
稳定性测试还必须讲清ISOS协议。ISOS-L-1I（25°C、惰性）和ISOS-L-2I（85°C、惰性）之间，仅温度差就能带来约14倍的表观寿命差异。把不同协议数据混在一张图里比较，会制造“材料突然更稳定”的假象。综述建议最低报告标准：ISOS层级与温度、器件面积、封装材料、Ea值。
另外，ICL自身也会退化：
硅基隧穿结：热致掺杂互扩散，BTBT退回TAT，表现为FF和VOC同步下降；
TCO/SAM：湿热水解导致SAM脱附，欧姆接触退回肖特基接触，典型 signature 是FF先降、VOC后降，KPFM出现与金字塔尖相关的电势不均。
两条路线的退化模式不同，测试协议也必须匹配。
图片5
图片6
Towards commercialization: Scalable manufacturing, module integration, and reliability assessment of PSTSCs photovoltaics. a, Schematic illustration of large-area scalable deposition techniques for perovskite films. This highlights the transition from laboratory-scale spin-coating to industrial-grade processes, including blade coating, slot-die coating, spray coating, and inkjet printing. These techniques are pivotal for realizing low- cost, high-throughput module manufacturing while maintaining high efficiency. b, Typical interconnection configurations for PSTSCs solar modules. The monolithic 2T architecture is regarded as the most commercially promising approach due to reduced parasitic absorption and simplified peripheral circuitry, though it imposes stringent requirements on current matching. c, Operational stability analysis of high-efficiency tandem devices under accelerated ageing conditions. Under simulated 1-sun illumination at 85◦C in an inert atmosphere (ISOS-L-2I protocol), devices employing the surface functionalization strategy (Target) exhibit significantly superior long-term stability compared to the control group, reaching T90 > 1400 h. d, Long-term performance evolution and environmental impact analysis of tandem devices in real-world outdoor environments. One-year outdoor field monitoring of encapsulated devices reveals the specific impacts of spectral variation and soiling (dust accumulation) on current matching and energy yield. The devices retained approximately 80% of their initial efficiency after undergoing seasonal cycles.

06 路线怎么选？不是“谁绝对赢”，而是“约束条件决定赢家”

综述给出一个很务实的“约束条件 verdict”：

图片7
注意：当前34.58%是实验室小面积纪录，不等于组件效率；TCO/SAM的铟资源、SAM大面积均匀性和长期湿热耐久性，仍是产业化前必须回答的问题。

07 距离40%还差什么？

从35.2%到40%，不是“再找一个新材料”那么简单，而是三个目标同时达成：
非辐射电压损失从约150 mV降到80 mV以下；
FF逼近85–86%；
顶底电池电流匹配在≥100 cm²面积上控制在约±0.2 mA/cm²。
这三个目标被同一个“主变量”耦合：SAM在完整金字塔形貌上的第二步覆盖均匀性。它同时影响接触电阻、界面缺陷密度和钙钛矿浸润/成核。大面积M10/G12硅片上金字塔数量超过10⁷个，覆盖分布的统计尾部就可能设定VOC下限。

未来研究方向可以概括为四条：
发展大面积、亚金字塔分辨的KPFM/EL/TEM关联表征；
在2T实际光注入条件下验证Poly-Si隧穿氧化层完整性；
探索满足“低ρc、低FCA、低铟、SAM非依赖”的四约束ICL；
建立SAM分子与钙钛矿添加剂的物理描述符数据库，做数据驱动筛选。
制造端还有三道硬约束：银耗（15–25 mg/W）、铟供应、铅管理；2T与4T也不是简单替代关系，4T可作为早期产业化互补路径，但不解决2T的ICL和电流匹配问题。

钙钛矿/硅叠层突破35%，靠的不是把两种材料简单叠在一起，而是把纹理化三维埋底界面当作“电荷输运—光学—结晶—稳定性”的共同变量来协同设计。下一阶段的胜负手，是谁能在大面积纹理硅上把SAM覆盖、ICL接触和钙钛矿结晶同时做到均匀、可靠、可制造。

35%只是起点。叠层电池真正的技术深水区，不在“材料名单”，而在那个被金字塔纹理放大的埋底界面。

4CzIPN 2,4,5,6-四(9-咔唑基)-间苯二腈 1416881-52-1
4CzIPN 2,4,5,6-四(9-咔唑基)-间苯二腈 1416881-52-1
4CzTPN 2,3,5,6-四(9-咔唑基)-对苯二腈 1416881-53-2
4CzTPN-Ph 2,3,5,6-四(3,6-二苯基-9-咔唑基)-对苯二腈 1416881-55-4
4CzTPN-Bu 2,3,5,6-四(3,6-二叔丁基-9-咔唑基)-对苯二腈
4CzPN-Ph 3,4,5,6-四(3,6-二苯基-9-咔唑基)-对苯二腈 1469707-47-8
4CzPN-Bu 3,4,5,6-四(3,6-二叔丁基-9-咔唑基)-对苯二腈 1469705-93-8
DMAC-DPS 双[4-(9,9-二甲基- 9,10-二氢吖啶)苯基]硫砜  1477512-32-5
DPEPO 二[2-((氧代)二苯基膦基)苯基]醚 808142-23-6
BCPO 双-4(N-咔唑基苯基)苯基氧化膦 1233407-28-7
AQ(PhDPA)2,b1 2,6-双[4-二苯胺基苯基]-9,10-蒽醌 1640978-33-1
Ac-VPN 4,5-双[4-(9,9-二甲基-9,10-二氢吖啶)苯基]-1,2-二氰基苯 1784766-38-6
Px-VPN 4,5-双[4-(N-吩噁嗪)苯基]-1,2-二氰基苯 1784766-39-7
Ac-CNP 5,6-双[4-(9,9-二甲基-9,10-二氢吖啶)苯基]-2,3-二氰基吡嗪 1883400-34-7
Px-CNP 5,6-双[4-(N-吩噁嗪)苯基]-2,3-二氰基吡嗪    1883400-36-9
DMAc-MPM 2-甲基-4,6-双[4-(9,9-二甲基-9,10-二氢吖啶)苯基]嘧啶 1870041-76-1
DMAc-PPM 2-苯基-4,6-双[4-(9,9-二甲基-9,10-二氢吖啶)苯基]嘧啶 1870041-75-0
DPAc-MPM 2-甲基-4,6-双[4-(9,9-二苯基-9,10-二氢吖啶)苯基]嘧啶  2019165-20-7
DACT-II 9-[4-(4,6-二苯基-1,3,5-三嗪-2-基)苯基]-N3,N3,N6,N6-四苯基-9氢-咔唑-3,6-二胺 1613074-59-1
PXZ-TRZ 10-(4-(4,6-二苯基l-1,3,5-三嗪-2-基)苯基)-10氢-吩噁嗪 1411910-25-2
PXZ-DPS 双[4-(N-吩噁嗪)苯基]硫砜 1477511-57-1
DMAc-TRZ  10-(4-(4,6-二苯基-1,3,5-三唑-2-基)苯基)-9,9-二甲基-9,10-二氢吖啶 1628752-98-6
Cz-DPS 双[4-(9-氢-咔唑基)苯基]硫砜 733038-89-6
BuCz-DPS(t-Cz-DPS)  双[4-(3,6-二叔丁基-9- 咔唑基)苯基]硫砜   1396165-20-0
武汉海山科技 销售电话 17720597715

参考文献：
Breaking the 35% Efficiency Milestone: Material Innovations and Interface Mechanisms in Monolithic Perovskite/Silicon Tandem Solar Cells
Advanced Functional Materials
https://doi.org/10.1002/adfm.79031


转载内容（视频、文章、广告等）只以信息传播为目的，仅供参考，不代表本公众号认同其观点和立场也不作为投资意见及建议。内容的真实性、准确性和合法性由原作者负责。如涉及侵权，请联系删除，此转载不作为商业用途。