下载:
-
气态亚硝酸(HONO)和氮氧化物(NOx = NO + NO2)作为羟基(OH)自由基和臭氧(O3)的重要前体物, 是大气活性氮(Nr)循环的重要组成部分, 在对流层光化学过程中扮演着至关重要的角色[1,2]. HONO作为一种反应活性气体, 在大气中的浓度较低且寿命较短, 其具有易损耗、高溶解性的特点. 因此, 针对大气HONO的快速、准确、定量测量具有一定的挑战. 自1979年Perner和Platt [3]采用差分吸收光谱(DOAS)技术首次观测到HONO以来, 包括湿化学法和光谱法的多种测量技术被应用于大气HONO的测量. 湿化学法如长程吸光光度法(LOPAP)利用吸收液吸收气态HONO, 通过测量吸收液中亚硝酸根离子或衍生物的含量, 实现对于大气HONO的定量测量, 其具有较高的探测灵敏度(pptv量级, 1 pptv = 10–12). 然而, 该方法易受到大气中NOy物种的化学干扰[4], 且时间分辨率受限, 需要经常标定和更换化学溶液. 光谱法作为一种直接测量方法, 通过测量HONO在特定波段(紫外或红外)的特征吸收光谱定量HONO浓度, 不易受到化学干扰的影响, 主要包括DOAS技术[3]、可调谐激光吸收光谱(TDLAS)技术[5]、腔衰荡光谱(CRDS)技术[6]和宽带腔增强吸收光谱(BBCEAS)技术[7-10]. 其中BBCEAS是基于CRDS发展的高灵敏度探测技术, 通过增加气体在光学腔内的有效吸收光程, 实现对痕量气体的高灵敏度探测. 该技术具有结构简单、精度高、适用于外场测量和多种气体同时探测等优点, 已成功应用于多种痕量气体(NO2, N2O5, HONO, IO, CHOCHO, NO等)的高灵敏度探测研究[11-16].
目前, BBCEAS技术中的光学腔主要采用密封腔体的结构设计, 环境大气通过泵采样至光学腔内进行测量, 对于大气自由基(如NO3)或易损耗(如HONO)活性气体的测量, 需要标定采样损耗和腔壁吸附的影响[11,15]. 区别于前者, 开放光路宽带腔增强吸收光谱(OP-BBCEAS)技术可以避免采样损耗和腔壁吸附的影响, 其已经成功应用于大气模拟舱痕量气体如NO3和气溶胶的研究[17-20]. 此外, Wu等[21]报道了基于常规BBCEAS光谱反演算法的OP-BBCEAS系统应用于实验室清洁空气HONO和NO2的测量研究, 系统积分时间90 s下对HONO和NO2探测灵敏度(1σ)分别为0.43 × 10–9和1 × 10–9. 由于常规BBCEAS反演算法依赖于光强的绝对稳定, 观测中发现高气溶胶浓度会影响光谱拟合结果, 相较于无气溶胶影响时的光谱拟合, 其光谱拟合残差较大[22]. Horbanski等[23]首次提出了基于迭代算法的BBCEAS技术, 利用DOAS反演消除了宽带变化的影响. 模拟显示在有效吸收光程减少约80%的情况下, 该算法仍然能够准确反演被测痕量气体的浓度. Tang等[24]验证了该算法对光强波动的不敏感性, 并与常规BBCEAS反演结果进行了对比, 不同算法的反演结果显著性相关(HONO: R2 = 0.94. NO2: R2 = 0.99). 然而, 上述算法仅在采用封闭腔结构的BBCEAS系统上得到了验证, 并未开展过实际大气的应用研究.
本文介绍了基于迭代算法的大气HONO和NO2开放光路宽带腔增强吸收光谱定量测量技术, 开放光路的测量模式避免了光学腔和采样管表面损耗和二次生成的影响. 基于迭代反演算法, 通过多次迭代确定有效吸收光程, 然后利用DOAS反演HONO和NO2浓度, 消除了气溶胶颗粒Mie散射和光源波动的宽带变化影响. 通过对比不同算法的反演结果并开展基于迭代算法的OP-BBCEAS系统与常规封闭腔BBCEAS系统的HONO和NO2测量对比实验, 验证了迭代算法应用于OP-BBCEAS系统测量的可行性.
-
BBCEAS技术是基于痕量气体对光辐射的特征吸收实现对痕量气体的定性和定量测量, 其实质是通过光在有限长光学腔内的多次反射增加有效吸收光程, 提高对痕量气体的探测灵敏度. 由于腔增强吸收光谱技术的有效吸收光程很长, 常规封闭腔BBCEAS技术通常会采用聚四氟乙烯(PTFE)过滤膜过滤气溶胶等颗粒物, 消除气溶胶颗粒Mie散射消光的影响. 然而, 对于OP-BBCEAS技术而言, 不仅需要考虑气体分子吸收和瑞利(Rayleigh)散射的影响, 还需要考虑气溶胶颗粒Mie散射消光的影响. 常规BBCEAS反演算法依赖于光源的绝对稳定, 通过测量绝对光强的变化确定吸收系数
$ \alpha \left( \lambda \right) $ , 利用最小二乘算法拟合被测气体的吸收截面和测量的吸收系数$ \alpha \left( \lambda \right) $ 反演被测气体的浓度:$ \begin{split} \alpha \left(\lambda \right)=\;&{\displaystyle \sum _{i}{c}_{i}{\sigma }_{i}\left(\lambda \right)}+\text{Polynomial}\\ =\;&{c}_{\text{HONO}}{\sigma }_{\text{HONO}}\left(\lambda \right)+{c}_{{\text{NO}}_{\text{2}}}{\sigma }_{{\text{NO}}_{\text{2}}}\left(\lambda \right)\\ &+{c}_{{\text{O}}_{\text{2}}{\text{-O}}_{\text{2}}}{\sigma }_{{\text{O}}_{\text{2}}{\text{-O}}_{\text{2}}}\left(\lambda \right)+{a}_{1}{\lambda }^{2}+{a}_{2}\lambda +{a}_{3}\text{, }\end{split} $ 式中
$ {c_{{\text{HONO}}}} $ ,$ {c_{{\text{N}}{{\text{O}}_{\text{2}}}}} $ 和${c_{{{\text{O}}_{\text{2}}}{\text{-}}{{\text{O}}_{\text{2}}}}}$ 分别为HONO, NO2和O2-O2的浓度;$ {\sigma _{{\text{HONO}}}}\left( \lambda \right) $ ,$ {\sigma _{{\text{N}}{{\text{O}}_{\text{2}}}}}\left( \lambda \right) $ 和${\sigma _{{{\text{O}}_{\text{2}}}{\text{-}}{{\text{O}}_{\text{2}}}}}\left( \lambda \right)$ 分别为HONO[25], NO2[26]和O2-O2[27]高分辨率截面与仪器函数卷积后获得的吸收截面; 多项式项a1, a2和a3为背景基线中的宽带变化(气溶胶颗粒Mie散射等). 由于该算法利用绝对光强的变化反演被测气体的浓度, 对环境变化十分敏感, 环境空气中气溶胶消光随时间的变化和光源波动均会影响被测气体的浓度反演.基于迭代算法的OP-BBCEAS技术, 是BBCEAS技术和DOAS反演算法的结合[23,24]. 通过BBCEAS技术增加有效吸收光程提高对痕量气体的探测灵敏度, 利用DOAS反演算法不受宽带变化影响的特点, 使用窄带差分结构反演被测气体浓度, 消除了气溶胶颗粒Mie散射消光和光源波动宽带变化的影响, 其光学厚度(
${D_{{\rm{CE}}}}\left( \lambda \right)$ )的定义如下:$ \begin{split} \;& {D}_{\text{CE}}\left(\lambda \right)=\text{ln}\left(\frac{{I}_{\text{tot0}}}{{I}_{\text{tot}}}\right)\\ =\;&\overline{{L}_{\text{eff}}}\left(\lambda \right)\cdot \left({\displaystyle \sum _{i}{c}_{i}\cdot {\sigma }_{i}\left(\lambda \right)}+{\varepsilon }_{\text{R}}\left(\lambda \right)\right)\text{, } \end{split}$ 式中
$ {I_{{\text{tot}}}} $ 和$ {I_{{\text{tot}}0}} $ 分别为有气体吸收和无气体吸收时的透射光强,$\overline {{L_{{\text{eff}}}}} \left( \lambda \right)$ 为有效吸收光程,$ {c_i} $ 和$ {\sigma _i}\left( \lambda \right) $ 分别为第i种气体的浓度和吸收截面,${\varepsilon _{\text{R}}}\left( \lambda \right)$ 为Rayleigh散射和Mie散射的宽带消光. 当已知有效吸收光程$\overline {{L_{{\text{eff}}}}} \left( \lambda \right)$ 时, 就可以利用DOAS反演算法反演被测气体的浓度. 然而, 由于BBCEAS技术的有效吸收光程强烈依赖于波长, 其有效吸收光程$\overline {{L_{{\text{eff}}}}} \left( \lambda \right)$ 不等于“空腔”时的光程$\overline {{L_0}} \left( \lambda \right)$ :$ \overline {{L_{{\text{eff}}}}} \left( \lambda \right){\text{ = }}\frac{{{D_{{\text{CE}}}}\left( \lambda \right)}}{{\exp \left( {{D_{{\text{CE}}}}\left( \lambda \right)} \right){{ - }}1}}\overline {{L_0}} \left( \lambda \right). $ 因此, BBCEAS技术利用DOAS反演算法反演被测气体浓度时需要修正有效吸收光程. 采用Horbanski等[23]提出的迭代算法修正有效吸收光程, 通过多次迭代计算BBCEAS的有效吸收光程, 然后利用DOAS进行光谱拟合获取HONO和NO2的浓度信息. 该算法具有DOAS反演算法的优点, 可以通过高通滤波的方式消除气溶胶颗粒Mie散射消光和光源波动宽带变化的影响, 而不会影响窄带吸收结构的拟合, 实现了开放光路测量模式下HONO和NO2浓度的准确测量. Horbanski等[23]和Tang等[24]已经详细介绍了迭代算法计算有效吸收光程的过程, 有效吸收光程
$\overline {{L_{{\text{eff}}}}} \left( \lambda \right)$ 可表示为$ {\overline{L}}_{\text{eff}}(\lambda )=\frac{{D}_{\text{CE}}(\lambda )}{\sigma (\lambda )\cdot c}\text{, } $ $ {\bar L_{{\text{eff}}}}(\lambda ) = \frac{d}{{1 - R(\lambda ) + d \cdot {\varepsilon _{\text{R}}}(\lambda )}} \cdot \frac{{{D_{{\text{CE}}}}(\lambda )}}{{\exp \left( {{D_{{\text{CE}}}}\left( \lambda \right)} \right){{ - }}1}}. $ 由(4)式和(5)式可知, 若已知光学厚度, 则可以修正有效吸收光程. 基于迭代算法假定HONO和NO2的浓度已知, 根据(2)式可计算光学厚度[23,24]. 通过多次迭代确定真实的有效吸收光程, 从而获得最终的HONO和NO2反演浓度, 其迭代的停止条件为两次迭代反演的浓度差异小于光谱拟合的误差.
-
不同于常规封闭腔BBCEAS系统, OP-BBCEAS系统实际测量时无光学腔, 仅在仪器标定时将光学腔放置在两片高反镜之间, 实际测量时将其移除. 因此, OP-BBCEAS系统避免了因光学腔和采样造成的吸附损耗和二次生成的影响. 本文采用高反镜间距为33 cm的OP-BBCEAS系统来验证基于迭代算法的OP-BBCEAS技术应用于大气HONO和NO2测量的可行性.
OP-BBCEAS系统示意图如图1所示, 系统主要由LED光源、透镜、高反镜、离轴抛物面镜和光谱仪等组成. LED(LZ1-00 UV00)光源中心波长为365 nm, 半峰全宽13 nm, 光功率可达到1680 mW. 通过PID算法控制与LED芯片相连的半导体制冷片, 利用热敏电阻温度探头进行实时温度反馈, 实现LED光源恒温(20 ± 0.1) ℃控制. 光源发出的光通过消色差透镜(f = 50 mm)准直后, 耦合到由两片高反镜(d = 25 mm)组成的间距为33 cm的开放腔中. 开放腔测量状态下, 环境空气经PTFE过滤膜(0.2 μm)后由隔膜泵(KNF)抽取至两端高反镜前维持高反镜前端正压, 以降低环境空气中气溶胶颗粒物对高反镜造成的污染. 透射光由90°离轴抛物面镜(f = 25.4 mm)聚焦耦合进入光纤, 经光纤传输至光谱仪(QE65000, Ocean Optics)采集光谱信号, 通过电脑分析采集的光谱信号获得被测痕量气体的浓度信息.
图 1 OP-BBCEAS系统示意图
Figure 1. Schematic diagram of the OP-BBCEAS instrument setup.
-
由于光学腔内气体的吸收光程和高反镜镜片反射率相关, 需要标定镜片反射率
$ R\left( \lambda \right) $ 才能准确反演被测气体的浓度. 镜片反射率的标定可以通过在高反镜之间增加光学腔后, 测量已知浓度的气体吸收或散射引起的光强变化确定, 图2示意了采用高纯氮气(N2, 99.999%)和高纯氦气(He, 99.999%)的Rayleigh散射差异性标定高反镜镜片反射率的标定结果. 图中显示了高反镜镜片反射率随波长的变化曲线, 黑线和红线分别为腔内充满高纯氮气和高纯氦气时的光谱信号, 计算的HONO吸收峰368 nm处高反镜镜片反射率为0.99983, 空腔时的有效吸收光程为1.94 km.
图 2 氮气(N2)谱(黑色)、氦气(He)谱(红色)和镜面反射率曲线(蓝色)
Figure 2. Nitrogen (N2) spectrum (black line), helium (He) spectrum (red line) and the derived curve of mirror reflectivity (blue line).
不同于常规封闭腔BBCEAS系统, 开放光路的测量模式会受到环境空气中气溶胶颗粒Mie散射消光的影响, 如图3(a)所示. 图中显示了环境空气中气溶胶颗粒Mie散射消光对透射光强的影响, 在有和无气溶胶过滤装置时, 透射光谱强度变化明显, 其峰值光强下降了约21%, 估算368 nm处的气溶胶颗粒Mie散射消光系数约为 3 × 10–6 cm–1, 与污染城市环境中的气溶胶消光水平相当[28]. 考虑到气溶胶颗粒Mie散射消光会减少测量时的有效吸收光程, 根据有效吸收光程(
$ {L_{{\text{eff}}}}\left( \lambda \right) $ )表达式[29]:
图 3 环境空气中气溶胶颗粒Mie散射对透射光谱强度和有效吸收光程的影响 (a)气溶胶颗粒Mie散射对透射光谱强度的影响; (b)气溶胶颗粒Mie散射对有效吸收光程的影响
Figure 3. Influence of Mie scattering of aerosol particles in ambient air on transmission spectral intensity and effective absorption optical path: (a) Influence of Mie scattering of aerosol particles on transmission spectral intensity; (b) influence of Mie scattering of aerosol particles on effective absorption optical path.
$\begin{split} & {L_{{\text{eff}}}}\left( \lambda \right) = \\ & \frac{d}{{1 - R\left( \lambda \right) + {\alpha _{{\text{Ray}}}}\left( \lambda \right)d + {\alpha _{{\text{Mie}}}}\left( \lambda \right)d + \displaystyle\sum {{\sigma _i}\left( \lambda \right){c_i}d} }}, \end{split} $ 式中
$ {\alpha _{{\text{Ray}}}}\left( \lambda \right) $ 和$ {\alpha _{{\text{Mie}}}}\left( \lambda \right) $ 分别为Rayleigh散射和Mie散射消光系数; d为腔长;$\displaystyle\sum {{\sigma _i}\left( \lambda \right){c_i}}$ 为被测气体总的吸收(10–8 cm–1量级), 其与高反镜镜片反射率引起的消光$\dfrac{{1 - R\left( \lambda \right)}}{d}$ (10–6 cm–1量级)相比可忽略不计. 计算了空腔、腔内充满N2和实际大气测量时的有效吸收光程, 如图3(b)所示. 相较于空腔时的有效吸收光程, 气溶胶颗粒Mie散射消光显著减少了测量时的有效吸收光程, 368 nm处的最大有效吸收光程仅为1.33 km. -
由于BBCEAS技术采用宽带光源(LED)覆盖了HONO, NO2和O2-O2的特征吸收波段, 综合考虑LED光谱范围、高反镜的高反区域、HONO和NO2吸收峰等因素, 选择拟合波段为362.4—389 nm. 该拟合波段除HONO, NO2和O2-O2吸收外, 其他大气分子(CH2O, O3, BrO, IO和CHOCHO等)也有结构性吸收. 然而, 考虑到其吸收较弱, 实际大气浓度远低于系统的探测限, 在光谱拟合中可以忽略. 大气HONO和NO2的迭代反演实例如图4(a)所示, 拟合得到的HONO和NO2浓度分别为(1.81±0.11) × 10–9和(9.77±0.13) × 10–9. 光谱拟合残差的标准偏差为1.75 × 10–4, 拟合残差无明显结构, 光谱拟合效果较好.
图 4 基于迭代反演的OP-BBCEAS算法和常规BBCEAS反演算法存在气溶胶消光影响时的大气HONO和NO2反演实例 (a)基于迭代反演的OP-BBCEAS算法存在气溶胶消光影响时的光谱拟合结果, 拟合残差的标准偏差为1.75 × 10–4; (b)常规BBCEAS反演算法存在气溶胶消光影响时的光谱拟合结果, 拟合残差的标准偏差为7.10 × 10–9 cm–1
Figure 4. Examples of HONO and NO2 retrieval of iterative retrieval algorithm and conventional retrieval algorithm with the influence of aerosol extinction: (a) Spectral fitting results of iterative retrieval algorithm with the influence aerosol extinction, the standard deviation of fit residual is 1.75 ×10–4; (b) spectral fitting results of conventional retrieval algorithm with the influence aerosol extinction, the standard deviation of fit residual is 7.10 × 10–9 cm–1.
由于常规BBCEAS反演算法是利用绝对光强的变化反演HONO和NO2的浓度, 气溶胶消光会影响光谱拟合结果, 导致光谱拟合结果较差且拟合残差较大, 如图4(b)所示. 相较于无气溶胶消光影响时的光谱拟合(拟合残差的标准偏差为6.30 × 10–10 cm–1), 存在气溶胶消光影响时光谱拟合残差的标准偏差为7.10 × 10–9 cm–1, 增加了约1个数量级. 相反, 基于迭代反演的OP-BBCEAS算法, 其利用了DOAS反演算法的优势消除了气溶胶颗粒Mie散射消光和光源波动宽带变化的影响, 在有和无气溶胶消光影响时光谱拟合无明显差异, 积分时间为60 s时HONO和NO2的探测灵敏度(1σ)分别为95 × 10–12和270 × 10–12.
由高斯误差传播可以确定系统的测量误差. OP-BBCEAS系统的测量误差主要是由高反镜镜片反射率的标定误差、标准吸收截面的测量误差和光谱拟合反演误差组成. 其中高反镜镜片反射率的标定误差为5%, 文献报道的HONO[25]和NO2[26]标准吸收截面的测量误差分别为5%和4%, 光谱拟合反演的误差为4%. 根据误差传递函数, HONO和NO2的总测量误差分别为8.1%和7.5%.
-
为验证基于迭代算法的OP-BBCEAS系统测量准确性, 开展了OP-BBCEAS系统和常规封闭腔BBCEAS系统HONO和NO2测量对比实验, 常规封闭腔BBCEAS系统的详细描述见文献[11]. 两台系统均放置于合肥市科学岛安徽光学精密机械研究所综合实验楼6楼(31.89°N, 117.17°E), 距地面约20 m. 常规封闭腔BBCEAS系统采样管长度约为2 m, 采样口放置于OP-BBCEAS系统附近. 考虑到实际测量时OP-BBCEAS系统采用过滤空气作为吹扫保护气流, 为验证过滤空气作为吹扫保护气流是否能够保证高反镜镜片反射率的稳定, 标定了实际大气测量前后的高反镜镜片反射率曲线, 如图5所示. 环境空气测量前后的高反镜镜片反射率变化差异< 0.1%, 表明采用过滤空气作为吹扫保护气流能够保证实际测量时高反镜镜片反射率的稳定.
图 5 原始镜面反射率曲线(黑色)和实际大气测量后镜面反射率曲线(红色)
Figure 5. Initial mirror reflectivity curve (black line) and mirror reflectivity curve after the atmospheric measurements (red line).
基于迭代算法的OP-BBCEAS系统和常规封闭腔BBCEAS系统测量不同PM2.5浓度下HONO和NO2浓度的时间序列, 如图6所示. 测量期间, PM2.5浓度变化范围为2—149 μg/m3, HONO浓度变化范围为0.77 × 10–9—2.80 × 10–9, 平均浓度为(1.75±0.49) × 10–9; NO2浓度变化范围为5.20 × 10–9—25.24 × 10–9, 平均浓度为(13.76 ± 5.12)×10–9. 图7(a)和图7(b)分别为轻度(PM2.5<75 μg/m3)和中度(PM2.5>75 μg/m3)不同气溶胶污染状况下两台BBCEAS系统测量HONO和NO2浓度的相关性, 不同PM2.5污染程度下两台BBCEAS系统测量的HONO和NO2浓度均显著性相关. PM2.5浓度小于75 μg/m3时两台BBCEAS系统测量的HONO和NO2浓度的相关性系数R2分别为0.998和0.999, HONO和NO2浓度的测量差异分别为2.4%和6.3%. PM2.5浓度大于75 μg/m3时两台BBCEAS系统测量的HONO和NO2浓度的R2分别为0.999和0.999, HONO和NO2浓度的测量差异均为4.0%. 两台BBCEAS系统测量的HONO和NO2浓度差异均在系统的测量误差(HONO: 8.1%. NO2: 7.5%)范围内, 其可能是由于光学腔/采样损耗或测量空气团的差异造成的.
图 6 基于迭代算法的 OP-BBCEAS系统和常规封闭腔BBCEAS系统测量不同PM2.5浓度(小时均值)下HONO和NO2浓度时间序列, 红色点线为OP-BBCEAS系统测量结果, 黑色点线为封闭腔BBCEAS系统测量结果
Figure 6. Time series of HONO and NO2 concentrations measured by OP-BBCEAS system based on iterative algorithm and conventional close-path BBCEAS system at different hourly average PM2.5 concentrations. The red dotted line is the measurements of open-path BBCEAS system, and the black dotted line is the measurements of close-path BBCEAS system.
图 7 基于迭代算法的OP-BBCEAS系统和常规封闭腔BBCEAS系统在不同PM2.5浓度下测量HONO和NO2浓度的相关性 (a)轻度(PM2.5<75 μg/m3)和中度(PM2.5>75 μg/m3)气溶胶污染状况下两台BBCEAS系统测量HONO浓度的相关性; (b)轻度(PM2.5<75 μg/m3)和中度(PM2.5>75 μg/m3)气溶胶污染状况下两台BBCEAS系统测量NO2浓度的相关性
Figure 7. Correlation of HONO and NO2 concentrations measured by OP-BBCEAS system based on iterative algorithm and conventional close-path BBCEAS system at different PM2.5 concentrations: (a) The correlation between HONO concentration measured by two BBCEAS instruments in light (PM2.5<75 μg/m3) and moderate (PM2.5>75 μg/m3) aerosol loading; (b) the correlation between NO2 measured by two BBCEAS instruments in light (PM2.5<75 μg/m3) and moderate (PM2.5>75 μg/m3) aerosol loading.
-
本文介绍了基于迭代算法的开放光路宽带腔增强吸收光谱技术应用于大气HONO和NO2的定量测量. 开放光路的测量模式避免了因光学腔和采样造成的吸附损耗和二次生成的影响, 采用过滤空气作为吹扫保护气流, 能够保证开放腔测量状态下高反镜镜片反射率的稳定. 开放光路测量时, 气溶胶颗粒Mie散射消光显著减少了测量时的有效光吸收光程, 相较于空腔时368 nm处的最大有效吸收光程1.94 km, 最大有效吸收光程减少为1.33 km. 常规BBCEAS反演算法受气溶胶消光影响, 存在气溶胶消光时的光谱拟合较差, 拟合残差的标准偏差较无气溶胶消光影响时增大约1个数量级. 采用迭代反演算法反演HONO和NO2浓度, 通过多次迭代计算有效吸收光程, 利用DOAS反演HONO和NO2浓度, 消除了气溶胶颗粒Mie散射消光和光源波动宽带变化的影响, 在有和无气溶胶消光影响时其光谱拟合无明显差异, 在积分时间为60 s时HONO和NO2的探测灵敏度(1σ)分别为95 × 10–12和270 × 10–12. 开展轻度(PM2.5< 75 μg/m3)和中度(PM2.5>75 μg/m3)不同气溶胶污染状况下基于迭代算法的OP-BBCEAS系统大气HONO和NO2的测量, 并且与常规封闭腔BBCEAS系统测量结果进行了对比. 对比结果显示, 不同PM2.5污染程度下两台BBCEAS系统测量的HONO和NO2浓度均显著性相关(R2 > 0.99), HONO和NO2浓度的测量差异(HONO ≤ 4.0%, NO2 ≤ 6.5%)均小于系统测量误差, 验证了迭代算法应用于OP-BBCEAS系统实际大气HONO和NO2测量的可行性. 未来将利用基于迭代算法的OP-BBCEAS系统开展大气自由基及其活性前体物的测量, 探究自由基及其活性前体物对大气氧化性的影响.
-
气态亚硝酸(HONO)作为羟基(OH)自由基的重要前体物, 在大气中浓度低、寿命短、易损耗且活性强, 针对大气HONO的高灵敏度测量具有一定的挑战. 本文介绍了基于迭代算法的开放光路宽带腔增强吸收光谱(OP-BBCEAS)技术应用于大气HONO和NO2的测量. 常规BBCEAS技术通过将经滤膜过滤后的环境空气由泵压入/抽入光学腔内进行测量, 尽管可以减小气溶胶消光对测量的影响, 但针对一些活性组分的测量则需要考虑光学腔和采样造成的吸附损耗和二次生成等壁效应. 本文采用OP-BBCEAS技术, 开放光路的测量模式避免了上述壁效应的影响, 基于迭代反演算法通过多次迭代确定有效吸收光程, 然后采用差分光学吸收光谱的光谱拟合方法对光谱中HONO和NO2的吸收进行定量, 克服了气溶胶颗粒Mie散射消光和光源波动的宽带变化影响. 在轻度(PM2.5 < 75 μg/m3)和中度(PM2.5 > 75 μg/m3)不同气溶胶污染状况下测量了实际大气HONO和NO2浓度, 并同时与常规封闭腔BBCEAS系统开展了测量对比. 不同PM2.5污染程度下两台BBCEAS系统测量的HONO和NO2浓度均显著性相关(R2>0.99), HONO和NO2浓度的测量差异(HONO ≤ 4.0%, NO2 ≤ 6.5%)均小于系统测量误差(HONO: 8.1%, NO2: 7.5%), 验证了迭代反演算法应用于OP-BBCEAS系统实际大气测量的可行性.Nitrous acid (HONO), as an important precursor of hydroxyl (OH) radical, has a low concentration, short lifetime, easy loss and high reactivity in the atmosphere. Thus, the high sensitivity detection of atmospheric HONO is a challenge. In this paper, we report an open-path broadband cavity enhanced absorption spectroscopy (OP-BBCEAS) system based on the iterative algorithm for simultaneous measurement of atmospheric HONO and NO2. In the conventional BBCEAS system, a pump is used to drive the ambient air into the optical cavity through the filter membrane for measurement, which can reduce the influence of aerosol particle extinction. However, the influence of wall loss and secondary formation caused by the optical cavity and sampling should be considered for reactive component measurements. The OP-BBCEAS with open-path configuration is adopted in this paper to avoid being influenced by wall effect. The effective absorption optical path is calculated by the iterative retrieval algorithm through multiple iterations, and the absorption of HONO and NO2 are then quantified by the spectral fitting method of differential optical absorption spectroscopy, which removes the broadband change influence of the Mie scattering extinction by aerosol particles and the light intensity fluctuation. The atmospheric HONO and NO2 with light (PM2.5 < 75 μg/m3) and moderate (PM2.5 > 75 μg/m3) aerosol loading are measured by the OP-BBCEAS system based on iterative algorithm, and compared with the counterparts by the conventional close-path BBCEAS system. The concentrations of HONO and NO2 measured by the two BBCEAS systems are in good agreement (R2 > 0.99) for different PM2.5 concentration levels, and the measurement differences of HONO and NO2 concentrations (HONO ≤ 4.0%, NO2 ≤ 6.5%) are less than the systematic measurement errors (HONO: 8.1%, NO2: 7.5%), which verifies the feasibility of iterative algorithm applied to OP-BBCEAS system for atmospheric measurement.
-
Keywords:
- broadband cavity enhanced absorption spectroscopy /
- open-path /
- iterative algorithm /
- differential optical absorption spectroscopy
[1] Elshorbany Y F, Kurtenbach R, Wiesen P, Lissi E, Rubio M, Villena G, Gramsch E, Rickard A R, Pilling M J, Kleffmann J 2009 Atmos. Chem. Phys. 9 2257
Google Scholar
[2] Kleffmann J 2007 Chemphyschem 8 1137
Google Scholar
[3] Perner D, Platt U 1979 Geophys. Res. Lett. 6 917
Google Scholar
[4] Legrand M, Preunkert S, Frey M, Bartels-Rausch T, Jourdain B 2014 Atmos. Chem. Phys. 14 9963
Google Scholar
[5] Cui X, Yu R, Chen W, Zhang Z, Pang T, Sun P, Xia H, Wu B, Dong F 2019 J. Lightwave Technol. 37 2784
Google Scholar
[6] Wang L, Zhang J 2000 Environ. Sci. Technol. 34 4221
Google Scholar
[7] Fiedler S E, Hese A, Ruth A A 2003 Chem. Phys. Lett. 371 284
Google Scholar
[8] Gherman T, Venables D S, Vaughan S, Orphal J, Ruth A A 2008 Environ. Sci. Technol. 42 890
Google Scholar
[9] Wu T, Zha Q, Chen W, Xu Z, Wang T, He X 2014 Atmos. Environ. 95 544
Google Scholar
[10] Min K E, Washenfelder R A, Dubé W P, Langford A O, Edwards P M, Zarzana K J, Stutz J, Lu K, Rohrer F, Zhang Y, Brown S S 2016 Atmos. Meas. Tech. 9 423
Google Scholar
[11] Duan J, Qin M, Ouyang B, et al. 2018 Atmos. Meas. Tech. 11 4531
Google Scholar
[12] Hoch D J, Buxmann J, Sihler H, Pöhler D, Zetzsch C, Platt U 2014 Atmos. Meas. Tech. 7 199
Google Scholar
[13] Johansson O, Mutelle H, Alexander E P, et al. 2014 Appl. Phys. B-Lasers O. 114 421
Google Scholar
[14] Liang S, Qin M, Xie P, et al. 2019 Atmos. Meas. Tech. 12 2499
Google Scholar
[15] Wang H, Chen J, Lu K 2017 Atmos. Meas. Tech. 10 1465
Google Scholar
[16] Grilli R, Mejean G, Kassi S, Ventrillard I, Abd-Alrahman C, Romanini D 2012 Environ. Sci. Technol. 46 10704
Google Scholar
[17] Dorn H P, Apodaca R L, Ball S M, et al. 2013 Atmos. Meas. Tech. 6 1111
Google Scholar
[18] Venables D S, Gherman T, Orphal J, Wenger J C, Ruth A A 2006 Environ. Sci. Technol. 40 6758
Google Scholar
[19] Chen J, Wenger J C, Venables D S 2011 J. Phys. Chem. A 115 12235
Google Scholar
[20] Varma R M, Venables D S, Ruth A A, Heitmann U, Schlosser E, Dixneuf S 2009 Appl. Opt. 48 159
Google Scholar
[21] Wu T, Chen W, Fertein E, Cazier F, Dewaele D, Gao X 2012 Appl. Phys. B-Lasers O. 106 501
Google Scholar
[22] Suhail K, George M, Chandran S, Varma R, Venables D S, Wang M, Chen J 2019 Spectrochim. Acta A 208 24
Google Scholar
[23] Horbanski M, Pöhler D, Lampel J, Platt U 2019 Atmos. Meas. Tech. 12 3365
Google Scholar
[24] Tang K, Qin M, Fang W, et al. 2020 Atmos. Meas. Tech. 13 6487
Google Scholar
[25] Stutz J, Kim E S, Platt U, Bruno P, Perrino C, Febo A 2000 J. Geophys. Res. Atmos. 105 14585
Google Scholar
[26] Voigt S, Orphal J, Burrows J P 2002 J. Photoch. Photobio. A-Chem. 149 1
Google Scholar
[27] Greenblatt G D, Orlando J J, Burkholder J B, Ravishankara A R 1990 J. Geophys. Res. 95 18577
Google Scholar
[28] Moosmüller H, Varma R, Arnott W P 2005 Aerosol Sci. Tech. 39 30
Google Scholar
[29] Platt U, Meinen J, Pöhler D, Leisner T 2009 Atmos. Meas. Tech. 2 713
Google Scholar
-
图 2 氮气(N2)谱(黑色)、氦气(He)谱(红色)和镜面反射率曲线(蓝色)
Fig. 2. Nitrogen (N2) spectrum (black line), helium (He) spectrum (red line) and the derived curve of mirror reflectivity (blue line).
图 3 环境空气中气溶胶颗粒Mie散射对透射光谱强度和有效吸收光程的影响 (a)气溶胶颗粒Mie散射对透射光谱强度的影响; (b)气溶胶颗粒Mie散射对有效吸收光程的影响
Fig. 3. Influence of Mie scattering of aerosol particles in ambient air on transmission spectral intensity and effective absorption optical path: (a) Influence of Mie scattering of aerosol particles on transmission spectral intensity; (b) influence of Mie scattering of aerosol particles on effective absorption optical path.
图 4 基于迭代反演的OP-BBCEAS算法和常规BBCEAS反演算法存在气溶胶消光影响时的大气HONO和NO2反演实例 (a)基于迭代反演的OP-BBCEAS算法存在气溶胶消光影响时的光谱拟合结果, 拟合残差的标准偏差为1.75 × 10–4; (b)常规BBCEAS反演算法存在气溶胶消光影响时的光谱拟合结果, 拟合残差的标准偏差为7.10 × 10–9 cm–1
Fig. 4. Examples of HONO and NO2 retrieval of iterative retrieval algorithm and conventional retrieval algorithm with the influence of aerosol extinction: (a) Spectral fitting results of iterative retrieval algorithm with the influence aerosol extinction, the standard deviation of fit residual is 1.75 ×10–4; (b) spectral fitting results of conventional retrieval algorithm with the influence aerosol extinction, the standard deviation of fit residual is 7.10 × 10–9 cm–1.
图 5 原始镜面反射率曲线(黑色)和实际大气测量后镜面反射率曲线(红色)
Fig. 5. Initial mirror reflectivity curve (black line) and mirror reflectivity curve after the atmospheric measurements (red line).
图 6 基于迭代算法的 OP-BBCEAS系统和常规封闭腔BBCEAS系统测量不同PM2.5浓度(小时均值)下HONO和NO2浓度时间序列, 红色点线为OP-BBCEAS系统测量结果, 黑色点线为封闭腔BBCEAS系统测量结果
Fig. 6. Time series of HONO and NO2 concentrations measured by OP-BBCEAS system based on iterative algorithm and conventional close-path BBCEAS system at different hourly average PM2.5 concentrations. The red dotted line is the measurements of open-path BBCEAS system, and the black dotted line is the measurements of close-path BBCEAS system.
图 7 基于迭代算法的OP-BBCEAS系统和常规封闭腔BBCEAS系统在不同PM2.5浓度下测量HONO和NO2浓度的相关性 (a)轻度(PM2.5<75 μg/m3)和中度(PM2.5>75 μg/m3)气溶胶污染状况下两台BBCEAS系统测量HONO浓度的相关性; (b)轻度(PM2.5<75 μg/m3)和中度(PM2.5>75 μg/m3)气溶胶污染状况下两台BBCEAS系统测量NO2浓度的相关性
Fig. 7. Correlation of HONO and NO2 concentrations measured by OP-BBCEAS system based on iterative algorithm and conventional close-path BBCEAS system at different PM2.5 concentrations: (a) The correlation between HONO concentration measured by two BBCEAS instruments in light (PM2.5<75 μg/m3) and moderate (PM2.5>75 μg/m3) aerosol loading; (b) the correlation between NO2 measured by two BBCEAS instruments in light (PM2.5<75 μg/m3) and moderate (PM2.5>75 μg/m3) aerosol loading.
-
[1] Elshorbany Y F, Kurtenbach R, Wiesen P, Lissi E, Rubio M, Villena G, Gramsch E, Rickard A R, Pilling M J, Kleffmann J 2009 Atmos. Chem. Phys. 9 2257
Google Scholar
[2] Kleffmann J 2007 Chemphyschem 8 1137
Google Scholar
[3] Perner D, Platt U 1979 Geophys. Res. Lett. 6 917
Google Scholar
[4] Legrand M, Preunkert S, Frey M, Bartels-Rausch T, Jourdain B 2014 Atmos. Chem. Phys. 14 9963
Google Scholar
[5] Cui X, Yu R, Chen W, Zhang Z, Pang T, Sun P, Xia H, Wu B, Dong F 2019 J. Lightwave Technol. 37 2784
Google Scholar
[6] Wang L, Zhang J 2000 Environ. Sci. Technol. 34 4221
Google Scholar
[7] Fiedler S E, Hese A, Ruth A A 2003 Chem. Phys. Lett. 371 284
Google Scholar
[8] Gherman T, Venables D S, Vaughan S, Orphal J, Ruth A A 2008 Environ. Sci. Technol. 42 890
Google Scholar
[9] Wu T, Zha Q, Chen W, Xu Z, Wang T, He X 2014 Atmos. Environ. 95 544
Google Scholar
[10] Min K E, Washenfelder R A, Dubé W P, Langford A O, Edwards P M, Zarzana K J, Stutz J, Lu K, Rohrer F, Zhang Y, Brown S S 2016 Atmos. Meas. Tech. 9 423
Google Scholar
[11] Duan J, Qin M, Ouyang B, et al. 2018 Atmos. Meas. Tech. 11 4531
Google Scholar
[12] Hoch D J, Buxmann J, Sihler H, Pöhler D, Zetzsch C, Platt U 2014 Atmos. Meas. Tech. 7 199
Google Scholar
[13] Johansson O, Mutelle H, Alexander E P, et al. 2014 Appl. Phys. B-Lasers O. 114 421
Google Scholar
[14] Liang S, Qin M, Xie P, et al. 2019 Atmos. Meas. Tech. 12 2499
Google Scholar
[15] Wang H, Chen J, Lu K 2017 Atmos. Meas. Tech. 10 1465
Google Scholar
[16] Grilli R, Mejean G, Kassi S, Ventrillard I, Abd-Alrahman C, Romanini D 2012 Environ. Sci. Technol. 46 10704
Google Scholar
[17] Dorn H P, Apodaca R L, Ball S M, et al. 2013 Atmos. Meas. Tech. 6 1111
Google Scholar
[18] Venables D S, Gherman T, Orphal J, Wenger J C, Ruth A A 2006 Environ. Sci. Technol. 40 6758
Google Scholar
[19] Chen J, Wenger J C, Venables D S 2011 J. Phys. Chem. A 115 12235
Google Scholar
[20] Varma R M, Venables D S, Ruth A A, Heitmann U, Schlosser E, Dixneuf S 2009 Appl. Opt. 48 159
Google Scholar
[21] Wu T, Chen W, Fertein E, Cazier F, Dewaele D, Gao X 2012 Appl. Phys. B-Lasers O. 106 501
Google Scholar
[22] Suhail K, George M, Chandran S, Varma R, Venables D S, Wang M, Chen J 2019 Spectrochim. Acta A 208 24
Google Scholar
[23] Horbanski M, Pöhler D, Lampel J, Platt U 2019 Atmos. Meas. Tech. 12 3365
Google Scholar
[24] Tang K, Qin M, Fang W, et al. 2020 Atmos. Meas. Tech. 13 6487
Google Scholar
[25] Stutz J, Kim E S, Platt U, Bruno P, Perrino C, Febo A 2000 J. Geophys. Res. Atmos. 105 14585
Google Scholar
[26] Voigt S, Orphal J, Burrows J P 2002 J. Photoch. Photobio. A-Chem. 149 1
Google Scholar
[27] Greenblatt G D, Orlando J J, Burkholder J B, Ravishankara A R 1990 J. Geophys. Res. 95 18577
Google Scholar
[28] Moosmüller H, Varma R, Arnott W P 2005 Aerosol Sci. Tech. 39 30
Google Scholar
[29] Platt U, Meinen J, Pöhler D, Leisner T 2009 Atmos. Meas. Tech. 2 713
Google Scholar
计量
- 文章访问数: 1243
- PDF下载量: 46
- 被引次数: 0
