不用强酸强碱！原位生成三碘化硼，室温实现芳香醛酮的选择性还原脱氧

阅读这篇文章之前，建议先看一篇老文章：

羰基，绝对算得上是有机合成里的 “万能枢纽”。加成、缩合、羰基α‑位官能化，几乎所有合成工作都离不开羰基的参与。

但很多时候，我们其实只想把这个氧原子彻底摘掉而已，也就是羰基还原脱氧。

但传统脱氧手段，一直存在暴力暴躁的短板。还原体系需要强酸、强碱、严苛高温。比如最经典的两个还原法：Clemmensen 还原（克莱门森）：锌汞齐 + 浓盐酸，强酸性、加热回流。怕酸的底物直接阵亡，汞试剂还有重金属污染风险。Wolff‑Kishner‑黄鸣龙还原：水合肼 + 强碱，180‑200 ℃高温。对碱敏感基团不能兼容，高沸点长时间加热，很多官能团扛不住高温强碱环境。

工业上制备二芳基甲烷类药物骨架（例如降糖药恩格列净、达格列净关键中间体），主流工艺常采用路易斯酸 + 氢硅烷体系。其中三(五氟苯基)硼烷（BCF）用得最多。BCF 很强，但这类硼路易斯酸对水氧极度敏感，需要手套箱操作；硅烷试剂也存在成本、安全管控的现实问题。

因此，寻找一种可室温操作、普通试剂参与、不用极端酸碱，还能选择性做羰基脱氧的方案是非常重要的。

而这篇文献报道正是基于此。

近期，美国爱荷华大学化学系 Williams 课题组 在JOC上撰文，报道了一种使用硼氢化钾 (KBH₄)+ 碘 (I₂)，在体系中原位生成三碘化硼 BI₃，室温条件完成芳香醛、酮的还原脱氧的可替代方法，还能实现醛和酮的产物分叉选择：醛出苄碘，酮直接脱氧成亚甲基！

注：纯的三碘化硼 BI₃熔点只有49.9 ℃ 左右，本身价格昂贵，极易水解，需要手套箱保存，很难直接拿来做日常合成。所以本篇原位生成该试剂是亮点。

反应优化过程中，研究人员发现：

• I₂ 和 KBH₄ 缺一不可：单独用 KBH₄ 只得到 6% 的醇（KBH₄ 在环己烷中溶解度太低），单独用 I₂ 完全不反应。

• 预生成 BI₃ 的烷烃溶剂沸点需 ≥70 ℃；DCE、甲苯可以作为共溶剂（也可以不用），但用甲苯时有少量 Friedel–Crafts 副产物。

• 纯 BI₃ 也能做同样的反应，但必须配化学计量的水。（醛：不加水 15%，加 1.5 当量水 77%；酮：加 3 当量水 93%）。

• BI₃ 过量会过度还原，醛会一路被拉到甲基（entry 10：5 当量 KBH₄/10 当量 I₂ 时出现 18% 的甲基产物）。

底物兼容性上，醛的底物范围相当宽：邻/间/对位取代、1,3,5-三甲基苯甲醛都能做；卤素（含邻位卤代，需延长时间）和 三氟甲基都能耐受（后者值得单独提一句，因为此前该课题组研究过硼三卤化物引发的 C–F 卤素交换，CF₃ 在这个体系里能"活下来"并不理所当然）。萘醛只需 30 分钟（时间一长就过度还原成甲基产物，作者归因于扩展芳香体系更能稳定自由基/阳离子中间体）。腈、酰胺、硫醚这类路易斯碱性基团会配位络合"吃掉"BI₃（），因此需要额外补加，但 没有观察到腈被还原、也没有硫醚被氧化或裂解。反应还能做到 克级规模。

酮的范围里，二苯甲酮类、芳基烷基酮都可以；邻位位阻（如邻溴）不成问题；芴酮则给出 12% 分离收率的 芴二聚体（途径自由基中间体的一个提示）。芳基烷基酮分离收率偏低（51–77%），主要原因是产物本身易挥发，而不是反应做得不好。支链酮（如异丙基苯基酮）只走到 45%，其中 31% 停在苄基碘，说明酮的"最后一步还原"对位阻敏感。

底物局限性：

• 4-硝基苯甲醛：硝基先被还原成苯胺，后者与亲电的苄基碘反应，导致聚合；

• 茴香醛（对甲氧基苯甲醛）：伴随脱甲基，得到 33% 的 4-羟基苄基碘，还有寡聚分解产物；

• 脂肪醛：主要发生羟醛缩合（自缩合），拿不到目标产物；

• 强吸电子底物：结论部分明确写道，强吸电子底物只能部分还原、停在醇这一步——酮类中的三氟苯乙酮（3l）和吡啶底物（3m）都只给出醇，且对照实验显示单用 KBH₄ 时两者几乎不转化；

• 烷基醚：条件不耐受，会裂解（这正是该课题组 2018 年专门研究过的 BI₃ 醚裂解反应）；

• 脂肪酮（二烷基酮） ：发生羟醛反应，不可用。

反应机理：

• BI₃和体系微量水发生作用，原位产生 HI；羰基被硼物种或者 HI 活化；

• 氧被硼结合，碘离子进攻，形成碘代醇中间体；

• 苄位 C‑I 键发生少量均裂，产生碘自由基，开启自由基链传递；苄自由基从 HI 攫取氢，完成还原；

• 醛生成一级苄基自由基，稳定性弱，链终止停留在苄碘；酮生成更稳定二级苄自由基，可继续链式还原，直接得到亚甲基产物。

反应操作：

KBH4 (43.2 mg, 0.800 mmol, 2.00 equiv.), I2 (406.1 mg, 1.600 mmol, 4.000 equiv.) and cyclohexane (4 mL) were added to a 25 mL single neck round bottom flask equipped with a stir bar. The round bottom flask was capped with a rubber septum cap, and the septum was tightened with a copper wire. The flask was placed in a 70 °C oil bath and allowed to stir until BI3 generation was complete. Completion of BI3 generation can occur anywhere from 10 to 90 minutes and can be indicated by a change in color. Colors can range from completely clear to magenta. The reaction was allowed to cool to room temperature and subsequently punctured with a needle, a nitrogen line was attached (note: a nitrogen line was not added during the BI3 generation), and diluted with an additional 4 mL cyclohexane. Aldehyde (0.4 mmol) was weighed out in a 20 mL scintillation vial and dissolved in DCE (2 mL). The resulting aldehyde solution was injected into the round bottom flask containing in situ generated BI3, after which a color change to dark purple was almost immediately observed. The vial was rinsed with another 2 mL of DCE and injected into the reaction mixture.

After 1 hour, the reaction was cooled to 0 °C and quenched with a 5% Na2S2O3 solution, 10% ascorbic acid solution, or 10% sodium ascorbate solution (10 mL) while stirring. The contents were added to a separatory funnel, and the flask was washed with 10 mL EtOAc. The separatory funnel was shaken until all the iodine was consumed, as determined by loss of color, and the layers were separated. Aqueous phase was washed with EtOAc (2 x 10 mL). The organic fraction was then washed with brine (1 x 10mL), dried over Na2SO4 and concentrated in vacuo. The crude product was analyzed and quantified by 1H-NMR spectroscopy using 1,3,5 trimethoxybenzene as an internal standard. Column chromatography with silica gel was used to

KBH4 (43.2 mg, 0.800 mmol, 4.00 equiv.) and I2 (406.1 mg, 1.600 mmol, 8.00 equiv.) and cyclohexane (4 mL) were added to a single neck 50 mL round bottom flask equipped with a stir bar. The flask was then capped with a rubber septum and tightened with a copper wire. The flask was placed in a 70 °C oil bath and allowed to stir until BI3 generation was complete. Completion of BI3 generation can occur anywhere from 10 to 90 minutes and can be indicated by a change in color. Colors can range from completely clear to magenta. The reaction was allowed to cool to room temperature and subsequently punctured with a needle, a nitrogen line was attached (note: a nitrogen line was not added during the BI3 generation), and diluted with an additional 2 mL cyclohexane. Ketone (0.2 mmol) was weighed out in a 20 mL scintillation vial and dissolved in DCE (2 mL). The resulting ketone solution was injected into the round bottom flask containing in situ generated BI3, after which a color change to dark purple was almost immediately observed. The vial was rinsed with another 2 mL of DCE and injected into the reaction mixture.

After 2 hours, the reaction was cooled to 0 °C and quenched with a 5% Na2S2O3 solution, 10% ascorbic acid solution, or 10% sodium ascorbate solution (10 mL) while stirring. The contents were added to a separatory funnel, and the flask was washed with 10 mL EtOAc. The separatory funnel was shaken until all the iodine was consumed, as determined by loss of color, and the layers were separated. Aqueous phase was washed with EtOAc (2 x 10 mL). The organic fraction was then washed with brine (1 x 10mL), dried over Na2SO4 and concentrated in vacuo. The crude product was analyzed and quantified by 1H-NMR spectroscopy using 1,3,5 trimethoxybenzene as an internal standard. Column chromatography was used to isolate the final products.

Note: For isolation, the reaction was run on a 0.4 mmol scale.

公众号观点：本文的主要卖点有两个，一个是原位生成BI3，一个是温和的反应条件，但是遗憾的是碘太贵，且在烷基醛酮上并没有突破，这才是目前所重点需要的。

参考文献：Chiarella T M, Furrule A P, Williams F J. Reductive Deoxygenation of Aromatic Aldehydes and Ketones with Boron Triiodide[J]. J. Org. Chem., 2026, 91, 11704‑11711.