MOF 氧化脱硫

MOF Oxidative Desulfurization · v0.2.0

登录后使用此模板 下载 CSV 模板
33
字段（mof 本体）
10
优化目标

字段说明

取值范围与选项来自模板 YAML
字段类型单位取值范围 / 选项示例值建模
activation_method
活化方式
 categorical thermal-vacuum, thermal-inert, solvent-exchange, supercritical-co2, none thermal-vacuum 连续描述符
activation_temp_c
活化温度
 numeric °C 80.0 – 300.0 120.0 可搜索
bandgap_ev
光学带隙
 numeric eV 1.0 – 5.0 3.656 仅模型输入
bet_m2_g
BET 比表面积
 numeric m2/g 0.0 – 8000.0 1422.5 仅模型输入
catalyst_loading_g_l
催化剂用量
 numeric g/L 0.1 – 20.0 1.0 可搜索
crystallite_size_nm
晶粒尺寸
 numeric nm 1.0 – 5000.0 34.8 仅模型输入
cycle_retention_percent
循环保持率
 numeric % 0.0 – 120.0 98.46 仅模型输入
extractant_used
是否使用萃取剂
 boolean 自由 True 可搜索
extraction_removal_percent
萃取段脱除率
 numeric % 0.0 – 100.0 仅模型输入
initial_sulfur_ppm
初始硫浓度
 numeric ppm 50.0 – 5000.0 500.0 可搜索
linker
配体
 categorical H2BDC, NH2-H2BDC, H2BPDC, NO2-H2BDC, OH2-H2BDC H2BDC 连续描述符
linker_to_metal_ratio
配体金属比
 numeric 0.2 – 6.0 1.08 可搜索
linker_vacancy_fraction
配体缺失分数
 numeric 0.0 – 0.6 0.0712 仅模型输入
metal_leaching_ppm
金属浸出
 numeric ppm 0.0 – 100.0 0.291 仅模型输入
modulator
调节剂
 categorical none, formic acid, acetic acid, benzoic acid, trifluoroacetic acid, hydrochloric acid formic acid 连续描述符
modulator_equiv
调节剂当量
 numeric equiv 0.0 – 500.0 60.0 可搜索
oil_type
模拟油基质
 categorical n-octane, decalin 可搜索
open_metal_site_mmol_g
开放金属位点密度
 numeric mmol/g 0.0 – 20.0 0.326 仅模型输入
oxidant
氧化剂
 categorical H2O2, TBHP, O2, air, cumene-hydroperoxide H2O2 连续描述符
oxidant_to_sulfur_ratio
氧硫比
 numeric 1.0 – 30.0 6.0 可搜索
pore_diameter_nm
平均孔径
 numeric nm 0.2 – 20.0 0.633 仅模型输入
pore_volume_cm3_g
孔容
 numeric cm3/g 0.0 – 6.0 0.556 仅模型输入
rate_constant_min
拟一级速率常数
 numeric 1/min 0.0 – 2.0 仅模型输入
reaction_duration_min
反应时长
 numeric min 10, 15, 30, 45, 60, 90, 120, 180 30.0 可搜索
reaction_temp_c
反应温度
 numeric °C 20.0 – 120.0 30.0 可搜索
solvent
溶剂
 categorical DMF, DEF, DMA, methanol, water DMF 连续描述符
substrate
硫底物
 categorical DBT, 4,6-DMDBT, BT, thiophene, model-oil-mixture DBT 连续描述符
sulfone_selectivity_percent
砜选择性
 numeric % 0.0 – 100.0 93.65 仅模型输入
sulfur_removal_percent
脱硫率
 numeric % 0.0 – 100.0 8.88 仅模型输入
synthesis_duration_h
合成时长
 numeric h 6, 12, 16, 24, 48, 72, 96, 120 24.0 可搜索
synthesis_temp_c
合成温度
 numeric °C 80, 100, 110, 120, 140, 150, 160, 180, 200 120.0 可搜索
ti_source
钛源
 categorical Ti(OiPr)4, Ti(OBu)4, TiCl4, TiF4, TiO2-anatase, Ti(OEt)4, Ti6O6(OiPr)6(abz)6, Ti6O6(OiPr)6(tbbz)6, Ti6O6(OiPr)6(tbbz)10 Ti(OiPr)4 连续描述符
turnover_frequency_h
转化频率
 numeric 1/h 0.0 – 5000.0 138.84 仅模型输入

优化目标与采集策略

acquisition 默认值按目标标定
目标方向单位代理模型采集函数标定来源
bet_m2_g 最大化 m2/g gp ucb β=1.5 · batch 6 NOT calibrated. Small-sample default (24 SYNTHETIC rows from scripts/generate_mof_data.py, 2026-08-07). beta=1.5 follows the small-sample guidance in mof-ods-optimization-design.md §5 stage 1 (beta 1.0-2.0 at 12-30 points), not a held-out experiment. Recalibrate once real synthesis data lands. Do not cite as calibrated.
pore_diameter_nm 最大化 nm gp ucb β=1.5 · batch 6 NOT calibrated. Small-sample default (24 SYNTHETIC rows from scripts/generate_mof_data.py, 2026-08-07). beta=1.5 follows the small-sample guidance in mof-ods-optimization-design.md §5 stage 1 (beta 1.0-2.0 at 12-30 points), not a held-out experiment. Recalibrate once real synthesis data lands. Do not cite as calibrated.
open_metal_site_mmol_g 最大化 mmol/g gp ucb β=1.5 · batch 6 NOT calibrated. Small-sample default (24 SYNTHETIC rows from scripts/generate_mof_data.py, 2026-08-07). beta=1.5 follows the small-sample guidance in mof-ods-optimization-design.md §5 stage 1 (beta 1.0-2.0 at 12-30 points), not a held-out experiment. Recalibrate once real synthesis data lands. Do not cite as calibrated.
linker_vacancy_fraction 最大化 gp ucb β=1.5 · batch 6 NOT calibrated. Small-sample default (24 SYNTHETIC rows from scripts/generate_mof_data.py, 2026-08-07). beta=1.5 follows the small-sample guidance in mof-ods-optimization-design.md §5 stage 1 (beta 1.0-2.0 at 12-30 points), not a held-out experiment. Recalibrate once real synthesis data lands. Do not cite as calibrated.
sulfur_removal_percent 最大化 % gp ucb β=1.5 · batch 6 NOT calibrated. Small-sample default (24 SYNTHETIC rows from scripts/generate_mof_data.py, 2026-08-07). beta=1.5 follows the small-sample guidance in mof-ods-optimization-design.md §5 stage 1 (beta 1.0-2.0 at 12-30 points), not a held-out experiment. Recalibrate once real synthesis data lands. Do not cite as calibrated.
sulfone_selectivity_percent 最大化 % gp ucb β=1.5 · batch 6 NOT calibrated. Small-sample default (24 SYNTHETIC rows from scripts/generate_mof_data.py, 2026-08-07). beta=1.5 follows the small-sample guidance in mof-ods-optimization-design.md §5 stage 1 (beta 1.0-2.0 at 12-30 points), not a held-out experiment. Recalibrate once real synthesis data lands. Do not cite as calibrated.
turnover_frequency_h 最大化 1/h gp ucb β=1.5 · batch 6 NOT calibrated. Small-sample default (24 SYNTHETIC rows from scripts/generate_mof_data.py, 2026-08-07). beta=1.5 follows the small-sample guidance in mof-ods-optimization-design.md §5 stage 1 (beta 1.0-2.0 at 12-30 points), not a held-out experiment. Recalibrate once real synthesis data lands. Do not cite as calibrated.
cycle_retention_percent 最大化 % gp ucb β=1.5 · batch 6 NOT calibrated. Small-sample default (24 SYNTHETIC rows from scripts/generate_mof_data.py, 2026-08-07). beta=1.5 follows the small-sample guidance in mof-ods-optimization-design.md §5 stage 1 (beta 1.0-2.0 at 12-30 points), not a held-out experiment. Recalibrate once real synthesis data lands. Do not cite as calibrated.
rate_constant_min 最大化 1/min gp ucb β=1.5 · batch 6 NOT calibrated. Small-sample default (24 SYNTHETIC rows from scripts/generate_mof_data.py, 2026-08-07). beta=1.5 follows the small-sample guidance in mof-ods-optimization-design.md §5 stage 1 (beta 1.0-2.0 at 12-30 points), not a held-out experiment. Recalibrate once real synthesis data lands. Do not cite as calibrated.
metal_leaching_ppm 最小化 ppm gp ucb β=0.8 · batch 6 NOT calibrated. Small-sample default (24 SYNTHETIC rows, 2026-08-07). beta lowered from the domain default of 1.5 as a judgement about asymmetric cost, not from measurement. Do not cite as calibrated.

示例数据

官方 mof-ods 示例前 5 行（共 49 列）
sample_id material_name framework_family node_metal batch_identifier ti_source ti_source_steric_bulk linker …
MOF-001 BDC-framework batch 001 MIL-125 Ti B807-01 Ti(OiPr)4 3.0 H2BDC +41 列
MOF-002 BDC-framework batch 002 MIL-125 Ti B807-02 Ti(OBu)4 4.0 H2BDC +41 列
MOF-003 N-BDC-framework batch 003 NH2-MIL-125 Ti B807-03 TiCl4 0.0 NH2-H2BDC +41 列
MOF-004 N-BDC-framework batch 004 NH2-MIL-125 Ti B807-04 Ti(OiPr)4 3.0 NH2-H2BDC +41 列
MOF-005 BPDC-framework batch 005 MIL-125-analogue Ti B807-05 Ti(OBu)4 4.0 H2BPDC +41 列

完整文件可从上方「下载 CSV 模板」获得（表头为模板字段名，含一行示例）。

建模思路：2 段级联

合成条件 → 结构描述符

输入：ti_source、linker、modulator、modulator_equiv、solvent、linker_to_metal_ratio、synthesis_temp_c、synthesis_duration_h、activation_method、activation_temp_c → 输出：bet_m2_g、pore_volume_cm3_g、pore_diameter_nm、crystallite_size_nm、linker_vacancy_fraction、open_metal_site_mmol_g、bandgap_ev

合成 + 结构 + 反应条件 → 脱硫性能

输入：ti_source、linker、modulator、modulator_equiv、solvent、linker_to_metal_ratio、synthesis_temp_c、synthesis_duration_h、activation_method、activation_temp_c、bet_m2_g、pore_diameter_nm、open_metal_site_mmol_g、linker_vacancy_fraction、bandgap_ev、substrate、oxidant、oxidant_to_sulfur_ratio、initial_sulfur_ppm、oil_type、catalyst_loading_g_l、reaction_temp_c、reaction_duration_min、extractant_used → 输出：sulfur_removal_percent、sulfone_selectivity_percent、turnover_frequency_h、metal_leaching_ppm、cycle_retention_percent、rate_constant_min、extraction_removal_percent