当前位置: X-MOL 学术Environ. Sci. Technol. Lett. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Thermal Decomposition of PFAS: Response to Comment on “Thermal Stability and Decomposition of Perfluoroalkyl Substances on Spent Granular Activated Carbon”
Environmental Science & Technology Letters ( IF 10.9 ) Pub Date : 2021-02-08 , DOI: 10.1021/acs.estlett.1c00061
Feng Xiao 1 , Pavankumar Challa Sasi 1 , Bin Yao 2 , Alena Kubátová 2 , Svetlana A. Golovko 3 , Mikhail Y. Golovko 3 , Dana Soli 1
Affiliation  

Per- and perfluoroalkyl substances (PFAS) used as surface-active agents in nonstick cookware and firefighting foams have generally been assumed to be thermally stable. However, our results published in Environmental Science& Technology Letters showed that thermal decomposition of certain legacy PFAS compounds such as perfluorooctanoic acid (PFOA) can occur at temperatures as low as 150 °C.(1) In addition to the experimental observations presented in our Letter, the measured yield of fluorine (F) during PFAS thermal treatment led us to propose that a free radical pathway might be operative.(1) Kopinke and Frenzel’s comments(2) on our Letter(1) focus on three items: (1) thermal stability of perfluoroalkyl acids (PFAAs) and their salts, (2) what they perceive as missing discussion regarding the over-recoveries observed in one of the four tested extraction methods, and (3) our interpretation of the thermal desorption–pyrolysis–gas chromatography–mass spectrometry (TD–Pyr–GC–MS). They believe our study “provides valuable insights into the behavior of perfluorinated carboxylic acids (PFCAs) and sulfonic acids (PFSAs)”. We appreciate their comments and the opportunity to further discuss certain aspects of thermal decomposition of PFAS. Kopinke and Frenzel first comment that “the authors specify their substrates as “acids” throughout the entire article.” This is not true. As mentioned in the Materials and Methods (page 344) as well as in Table S1,(1) we used PFCAs and the potassium salts of PFSAs in this study. PFCA salts are rarely available on the market. We are well aware that deprotonated PFCAs and PFSAs predominate at intermediate pH. In aqueous solution, a PFCA/PFSA anion is always associated with a counterion (e.g., H+, NH4+, or Na+) to maintain electroneutrality. Future studies are recommended to study the effect of the counterion on the thermal stability of PFCAs and PFSAs. One main conclusion of our study(1) is that thermal stability of PFAS decreases in the following order: PFSAs ≥ PFUnDA > PFDA > PFNA > PFOA > PFBA > HFPO-DA. We stand by this conclusion. Kopinke and Frenzel’s second comment focuses on the over-recoveries of one of our four investigated extraction methods, i.e., extraction with alkaline methanol.(1) They incorrectly state that we used this method for “all the other PFCA data” in our study.(1) The unexpected high or low recovery percentages caused by matrix enhancement and/or suppression is not uncommon when a mass spectrometer serves as the detector. Thus, in our study,(1) we compared four extraction methods before performing large-scale thermal decomposition experiments. Due to unexpectedly high recovery (Figure 1b(1)), this alkaline methanol approach was not selected for the extraction of PFAS from granular activated carbon in our study.(1) Similarly, the extraction approach using methanol only (Figure 1c,d(1)) was also not adopted because of its low recovery. Detailed study regarding the mechanism of the matrix effect would be outside the scope of the original Letter.(1) The method we selected (methanol amended with 100 mmol/L ammonium acetate(1)) achieved recoveries ranging from 79.2% to 111.8% for most of the PFAS (Figure 1a(1)). We provided this information in the Materials and Methods of the original Letter:(1) “One portion of GAC particles was freeze-dried, weighed, and extracted using methanol (Vextr, mL) with 100 mmol/L of ammonium acetate (NH4Ac) to determine the PFAS mass on GAC before thermal treatment” and “... screening tests of different methanol extraction conditions are described in the SI.” Kopinke and Frenzel also comment that “the thermal treatments of PFAS samples are, in most cases, a superposition of physical volatilization and chemical decomposition”. They questioned our choice of the word “destabilization”. We believe that thermal decomposition of PFAS may involve a series of processes such as melting, boiling, sublimation, radical chain-transfer reactions, oxidation (for polyfluoroalkyl substances), and pyrolysis, which cannot be treated simply as physical volatilization and chemical decomposition. The word “destabilization” in our Letter(1) refers to the overall change in stable solid-state PFAS compounds via these processes. Kopinke and Frenzel also question the rationale behind application of the Eyring equation. However, we did not provide fitting results using the Arrhenius or Eyring equation in our Letter,(1) nor did we report a single apparent activation energy; this would have been misleading for such complicated processes. The Eyring-like and Arrhenius-like curves in panels e and f of Figure 2 of the Letter(1) are a guide for the eye and are also meant to inspire our readers’ further reflection on the nature of PFAS thermal decomposition. With regard to Kopinke and Frenzel’s comment on the ordinates in panels e and f of Figure 2, we stand by the units in these two figures(1) as these are commonly used for Eyring(-like) and Arrhenius(-like) curves, including in Dr. Kopinke’s own study.(3) Kopinke and Frenzel comment that “perfluoroheptene (C7F14) is a plausible primary ... product of PFOA (C8F15HO2)” and “the only clearly detected organic product”. While we admit that we lack standards to confirm volatile decomposition products of PFOA corresponding to the broad peak at a Pyr–GC–MS retention time of 2.142–2.270 min (Figure S3a–c of our original Letter),(1) it is unlikely that perfluoroheptene (C7F14) is the only decomposition product generated from PFOA (see Scheme 1 in Figure 1 of this Response), because it is not consistent with the measured yield of F from PFOA in both open (under N2 or CO2 flow) and closed thermal treatment systems (Figure 4 of the original Letter(1)). We believe that thermal degradation of PFOA results from free radical reactions initiated by bond breaking, which generates a range of products. As illustrated in Scheme 2 of Figure 1 of this Response or Figure 4 of our original Letter,(1) thermal decomposition of PFOA may begin with homolytic cleavage of the relatively weak C–C bond located next to the carboxyl group of PFOA. The C–C bond may split, forming a perfluoroalkyl radical (:C7F14; m/z 350.0). The perfluoroalkyl radical may transform to perfluoroheptene at low temperatures (Scheme 1 in Figure 1 of this Response) or further undergo a series of defluorination, or “unzip”, reactions, generating shorter-chain perfluoroalkyl radicals (Scheme 2 in Figure 1). Eventually, these “unzip” reactions may be terminated by production of “dead”, or very short, degradation products from PFOA (Scheme 2 in Figure 1). Kopinke and Frenzel state that “... it is common knowledge that GC analysis does not detect radicals”, a statement with which we agree. However, MS can detect radicals such as those generated in thermal and chemical processes as reviewed by Sablier and Fujii in 2002.(4) Figure 1. Possible schemes of PFOA thermal decomposition. In summary, my co-authors and I are grateful to Drs. Kopinke and Frenzel for their interest in our article. It is important to gain an improved understanding of the stability and decomposition of PFAS in natural and engineered thermal treatments. There is still much to learn about PFAS thermal decomposition at the molecular level, and we appreciate the opportunity to expand the discussion of this complicated process presented in our Letter(1) and Response through this scholarly channel. The authors declare the following competing financial interest(s): U.S. Patent (Application 16/923,637: Degradation of Recalcitrant Organic Pollutants). Patent applicant: University of North Dakota. Name of Inventor: Feng Xiao. Status of application: filed July 8, 2020. The authors declare the following competing financial interest(s): U.S. Patent (Application 16/923,637: Degradation of Recalcitrant Organic Pollutants). Patent applicant: University of North Dakota. Name of Inventor: Feng Xiao. Status of application: filed July 8, 2020.
This article references 4 other publications.


中文翻译:

PFAS的热分解:回应对“废颗粒活性炭上全氟烷基物质的热稳定性和分解”的评论

通常认为在不粘炊具和灭火泡沫中用作表面活性剂的全氟烷基物质和全氟烷基物质(PFAS)具有热稳定性。但是,我们的结果发表在《环境科学与技术快报》上他们认为我们的研究“为全氟化羧酸(PFCA)和磺酸(PFSA)的行为提供了有价值的见解”。我们感谢他们的评论以及有机会进一步讨论PFAS热分解的某些方面。Kopinke和Frenzel首先评论说:“作者在整篇文章中都将其底物指定为“酸”。这不是真的。如材料和方法(第344页)以及表S1,(1)所述,在本研究中我们使用了PFCA和PFSA的钾盐。PFCA盐在市场上很少有。我们很清楚,在中等pH值下,去质子化的PFCA和PFSA占主导地位。在水溶液中,PFCA / PFSA阴离子总是与抗衡离子缔合(例如,H 我们感谢他们的评论以及有机会进一步讨论PFAS热分解的某些方面。Kopinke和Frenzel首先评论说:“作者在整篇文章中都将其底物指定为“酸”。这不是真的。如材料和方法(第344页)以及表S1,(1)所述,在本研究中我们使用了PFCA和PFSA的钾盐。PFCA盐在市场上很少有。我们很清楚,在中等pH值下,去质子化的PFCA和PFSA占主导地位。在水溶液中,PFCA / PFSA阴离子总是与抗衡离子缔合(例如,H 我们感谢他们的评论以及有机会进一步讨论PFAS热分解的某些方面。Kopinke和Frenzel首先评论说:“作者在整篇文章中都将其底物指定为“酸”。这不是真的。如材料和方法(第344页)以及表S1,(1)所述,在本研究中我们使用了PFCA和PFSA的钾盐。PFCA盐在市场上很少有。我们很清楚,在中等pH值下,去质子化的PFCA和PFSA占主导地位。在水溶液中,PFCA / PFSA阴离子总是与抗衡离子缔合(例如,H 如材料和方法(第344页)以及表S1,(1)所述,在本研究中我们使用了PFCA和PFSA的钾盐。PFCA盐在市场上很少有。我们很清楚,在中等pH值下,去质子化的PFCA和PFSA占主导地位。在水溶液中,PFCA / PFSA阴离子总是与抗衡离子缔合(例如,H 如材料和方法(第344页)以及表S1,(1)所述,在本研究中我们使用了PFCA和PFSA的钾盐。PFCA盐在市场上很少有。我们很清楚,在中等pH值下,去质子化的PFCA和PFSA占主导地位。在水溶液中,PFCA / PFSA阴离子总是与抗衡离子缔合(例如,H+,NH 4 +或Na +)以保持电子中性。建议将来进行研究以研究抗衡离子对PFCA和PFSA的热稳定性的影响。我们研究的主要结论(1)是PFAS的热稳定性按以下顺序降低:PFSA≥PFUnDA> PFDA> PFNA> PFOA> PFBA> HFPO-DA。我们支持这一结论。Kopinke和Frenzel的第二点评论集中在我们研究的四种提取方法之一的过度回收率上,即用碱性甲醇提取。(1)他们错误地指出,我们在研究中将这种方法用于“所有其他PFCA数据”。 (1)当质谱仪用作检测器时,由基质增强和/或抑制引起的意外的高或低回收率并不罕见。因此,在我们的研究中 (1)在进行大规模热分解实验之前,我们比较了四种提取方法。由于出乎意料的高回收率(图1b(1)),在我们的研究中未选择这种碱性甲醇方法从粒状活性炭中萃取PFAS。(1)同样,仅使用甲醇的萃取方法(图1c,d( 1))也没有被采用,因为它的回收率低。关于基质效应机理的详细研究将不在原始信的范围之内。(1)我们选择的方法(甲醇用100 mmol / L乙酸铵修正(1))可实现79.2%至111.8%的回收率。大部分PFAS(图1a(1))。我们在原始信函的材料和方法中提供了以下信息:(1)“将一部分GAC颗粒冷冻干燥,称重,然后使用甲醇(由于出乎意料的高回收率(图1b(1)),在我们的研究中未选择这种碱性甲醇方法从粒状活性炭中萃取PFAS。(1)同样,仅使用甲醇的萃取方法(图1c,d( 1))也没有被采用,因为它的回收率低。关于基质效应机理的详细研究将不在原始信的范围之内。(1)我们选择的方法(甲醇用100 mmol / L乙酸铵修正(1))可实现79.2%至111.8%的回收率。大部分PFAS(图1a(1))。我们在原始信函的材料和方法中提供了以下信息:(1)“将一部分GAC颗粒冷冻干燥,称重,然后使用甲醇(由于出乎意料的高回收率(图1b(1)),在我们的研究中未选择这种碱性甲醇方法从粒状活性炭中萃取PFAS。(1)同样,仅使用甲醇的萃取方法(图1c,d( 1))也没有被采用,因为它的回收率低。关于基质效应机理的详细研究将不在原始信的范围之内。(1)我们选择的方法(甲醇用100 mmol / L乙酸铵修正(1))可实现79.2%至111.8%的回收率。大部分PFAS(图1a(1))。我们在原始信函的材料和方法中提供了以下信息:(1)“将一部分GAC颗粒冷冻干燥,称重,然后使用甲醇(在我们的研究中,没有选择从碱性活性炭中提取PFAS的碱性甲醇方法。(1)同样,由于回收率低,也没有采用仅使用甲醇的提取方法(图1c,d(1))。关于基质效应机理的详细研究将不在原始信的范围之内。(1)我们选择的方法(甲醇用100 mmol / L乙酸铵修正(1))可实现79.2%至111.8%的回收率。大部分PFAS(图1a(1))。我们在原始信函的材料和方法中提供了以下信息:(1)“将一部分GAC颗粒冷冻干燥,称重,然后使用甲醇(在我们的研究中,没有选择从碱性活性炭中提取PFAS的碱性甲醇方法。(1)同样,由于回收率低,也没有采用仅使用甲醇的提取方法(图1c,d(1))。关于基质效应机理的详细研究将不在原始信的范围之内。(1)我们选择的方法(甲醇用100 mmol / L乙酸铵修正(1))可实现79.2%至111.8%的回收率。大部分PFAS(图1a(1))。我们在原始信函的材料和方法中提供了以下信息:(1)“将一部分GAC颗粒冷冻干燥,称重,然后使用甲醇(关于基质效应机理的详细研究将不在原始信的范围之内。(1)我们选择的方法(甲醇用100 mmol / L乙酸铵修正(1))可实现79.2%至111.8%的回收率。大部分PFAS(图1a(1))。我们在原始信函的材料和方法中提供了以下信息:(1)“将一部分GAC颗粒冷冻干燥,称重,然后使用甲醇(关于基质效应机理的详细研究将不在原始信的范围之内。(1)我们选择的方法(甲醇用100 mmol / L乙酸铵修正(1))可实现79.2%至111.8%的回收率。大部分PFAS(图1a(1))。我们在原始信函的材料和方法中提供了以下信息:(1)“将一部分GAC颗粒冷冻干燥,称重,然后使用甲醇(V extr,mL)与100 mmol / L醋酸铵(NH 4Ac)来确定热处理前GAC上的PFAS质量”和“ ...在SI中描述了不同甲醇提取条件的筛选测试。” Kopinke和Frenzel还评论说:“在大多数情况下,PFAS样品的热处理是物理挥发和化学分解的叠加”。他们质疑我们对“稳定化”一词的选择。我们认为,PFAS的热分解可能涉及一系列过程,例如熔融,沸腾,升华,自由基链转移反应,氧化(对于多氟烷基物质)和热解,这些过程不能简单地视为物理挥发和化学分解。我们的信函(1)中的“稳定化”一词是指通过这些过程稳定的固态PFAS化合物的整体变化。Kopinke和Frenzel也质疑应用Eyring方程背后的原理。但是,我们没有使用Letter中的Arrhenius或Eyring方程来提供拟合结果,(1)也没有报告单一的表观活化能。对于如此复杂的过程,这可能会产生误导。Letter(1)图2中的面板e和f中的类似Eyring和Arrhenius的曲线是眼睛的指南,也旨在激发读者对PFAS热分解性质的进一步思考。关于Kopinke和Frenzel对图2的面板e和f中的纵坐标的评论,我们支持这两幅图(1)中的单位,因为它们通常用于Eyring(-like)和Arrhenius(-like)曲线,包括在Kopinke博士自己的研究中。(3)Kopinke和Frenzel评论说,“全氟庚烯(C 我们没有使用Letter中的Arrhenius或Eyring方程来提供拟合结果,(1)也没有报告单个表观活化能;对于如此复杂的过程,这可能会产生误导。Letter(1)图2中的面板e和f中的类似Eyring和Arrhenius的曲线是眼睛的指南,也旨在激发读者对PFAS热分解性质的进一步思考。关于Kopinke和Frenzel在图2的面板e和f中对纵坐标的评论,我们支持这两幅图(1)中的单位,因为它们通常用于Eyring(-like)和Arrhenius(-like)曲线,包括在Kopinke博士自己的研究中。(3)Kopinke和Frenzel评论说,“全氟庚烯(C 我们没有使用Letter中的Arrhenius或Eyring方程来提供拟合结果,(1)也没有报告单个表观活化能;对于如此复杂的过程,这可能会产生误导。Letter(1)图2中的面板e和f中的类似Eyring和Arrhenius的曲线是眼睛的指南,也旨在激发读者对PFAS热分解性质的进一步思考。关于Kopinke和Frenzel对图2的面板e和f中的纵坐标的评论,我们支持这两幅图(1)中的单位,因为它们通常用于Eyring(-like)和Arrhenius(-like)曲线,包括在Kopinke博士自己的研究中。(3)Kopinke和Frenzel评论说,“全氟庚烯(C (1)我们也没有报告单一的表观活化能;对于如此复杂的过程,这可能会产生误导。Letter(1)图2中的面板e和f中的类似Eyring和Arrhenius的曲线是眼睛的指南,也旨在激发读者对PFAS热分解性质的进一步思考。关于Kopinke和Frenzel对图2的面板e和f中的纵坐标的评论,我们支持这两幅图(1)中的单位,因为它们通常用于Eyring(-like)和Arrhenius(-like)曲线,包括在Kopinke博士自己的研究中。(3)Kopinke和Frenzel评论说,“全氟庚烯(C (1)我们也没有报告单一的表观活化能;对于如此复杂的过程,这可能会产生误导。Letter(1)图2中的面板e和f中的类似Eyring和Arrhenius的曲线是眼睛的指南,也旨在激发读者对PFAS热分解性质的进一步思考。关于Kopinke和Frenzel对图2的面板e和f中的纵坐标的评论,我们支持这两幅图(1)中的单位,因为它们通常用于Eyring(-like)和Arrhenius(-like)曲线,包括在Kopinke博士自己的研究中。(3)Kopinke和Frenzel评论说,“全氟庚烯(C Letter(1)图2中的面板e和f中的类似Eyring和Arrhenius的曲线是眼睛的指南,也旨在激发读者对PFAS热分解性质的进一步思考。关于Kopinke和Frenzel对图2的面板e和f中的纵坐标的评论,我们支持这两幅图(1)中的单位,因为它们通常用于Eyring(-like)和Arrhenius(-like)曲线,包括在Kopinke博士自己的研究中。(3)Kopinke和Frenzel评论说,“全氟庚烯(C Letter(1)图2中的面板e和f中的类似Eyring和Arrhenius的曲线是眼睛的指南,也旨在激发读者对PFAS热分解性质的进一步思考。关于Kopinke和Frenzel对图2的面板e和f中的纵坐标的评论,我们支持这两幅图(1)中的单位,因为它们通常用于Eyring(-like)和Arrhenius(-like)曲线,包括在Kopinke博士自己的研究中。(3)Kopinke和Frenzel评论说,“全氟庚烯(C7 F 14)是PFOA(C 8 F 15 HO 2)的合理的主要产物。尽管我们承认我们缺乏标准来确认PFOA的挥发性分解产物对应于Pyr–GC–MS保留时间为2.142–2.270 min时的宽峰(我们原始信的图S3a–c),(1)这不太可能认为全氟庚烯(C 7 F 14)是唯一从PFOA生成的分解产物(参见本响应图1中的方案1),因为它与在开放(在N 2或CO下)下从PFOA测得的F收率均不一致。2个流量)和封闭式热处理系统(原始Letter(1)的图4)。我们认为,PFOA的热降解是由键断裂引发的自由基反应引起的,自由基产生了一系列产物。如该响应图1的方案2或我们原始信函的图4所示,(1)PFOA的热分解可能始于均一裂解裂解PFOA羧基旁的相对较弱的C–C键。C–C键可能分裂,形成全氟烷基(:C 7 F 14m / z350.0)。全氟烷基可以在低温下转变为全氟庚烯(此响应的图1中的方案1),或进一步进行一系列脱氟或“解压缩”反应,生成较短链的全氟烷基(图1中的方案2)。最终,这些“解压缩”反应可能会因PFOA产生“死”或非常短的降解产物而终止(图1中的方案2)。Kopinke和Frenzel说:“……众所周知,气相色谱分析无法检测到自由基”,我们同意这一说法。但是,MS可以检测到自由基,例如Sablier和Fujii在2002年综述的在热和化学过程中生成的自由基。(4)图1. PFOA热分解的可能方案。总而言之,我和我的合著者对Drs表示感谢。Kopinke和Frenzel对我们的文章感兴趣。重要的是要更好地了解天然和工程热处理中PFAS的稳定性和分解。关于PFAS热分解在分子水平上仍有很多知识,我们很高兴有机会通过这种学术渠道扩大对我们的Letter(1)和Response中提出的这一复杂过程的讨论。作者宣称以下竞争的经济利益:美国专利(申请16 / 923,637:难降解有机污染物的降解)。专利申请人:北达科他大学。发明人名称:冯晓。申请状态:2020年7月8日提交。作者声明以下竞争的经济利益:美国专利(申请16 / 923,637:难降解有机污染物的降解)。专利申请人:北达科他大学。发明人名称:冯晓。申请状态:2020年7月8日提交。
本文引用了其他4个出版物。
更新日期:2021-04-13
down
wechat
bug