氨和有机胺对羰基硫与α-Fe2O3非均相反应的影响

张拴勤 孔令东 赵 希 罗兰·詹森 陈建民

(复旦大学环境科学与工程系，上海市大气颗粒物污染防治重点实验室，上海200433)

1 Introduction

Carbonyl sulfide(COS)is a trace constituent of the earth's atmosphere，and it is one of the most abundant sulfur-containing species with a rather uniform mixing ratio of about 500×10−6(volume fraction)in the troposphere.1−4Because of its chemical inertness in the troposphere，the lifetime of carbonyl sulfide is longer than that of the other sulfur-containing gases.Once COS is transported into the stratosphere，it can be photodissociated and oxidized to form gaseous sulfur dioxide which can be finally converted into sulfate aerosol.5This sulfate-containing aerosol plays a significant role in atmospheric chemical processes，ozone depletion，and atmospheric radiation budget，6，7which is expected to have an impact on climate.Significant attentions have been paid to its sources and sinks in the atmosphere.It is estimated that the global sources of COS(about(1.31±0.25)Tg·a−1)were less than its sinks(about(1.66±0.79)Tg·a−1)，8indicating that atmospheric chemistry of COS still have uncertainties.

Recently，using infrared spectroscopy，some studies indicated that COS can be catalytically oxidized on the surfaces of single component aerosols as well as multiple component aerosols to produce C，and the components of aerosol can influence the reactivity of the aerosol.9−15Thus heterogeneous oxidation of COS is considered to be a potential sink for COS.However，previous studies mainly focus on the single component aerosol but not much on multiple component aerosols.In fact，real atmospheric aerosols are complex and many factors can influence the heterogeneous reactions of atmospheric trace gases.For example，a mineral dust particle may be covered by various chemical species during the transport process，then the outer layer of particle surface may be different from the mineralogy of the original particle，which can make the reactivity of the particle alter and finally enhance or suppress the reactivity of trace gases with the particle.16In our previous work，we found that the NaCl-containing hematite could decrease the reactivity of COS while the nitrate-containing hematite could enhance the reactivity，14，15which implied that the mixture aerosol had different influence on the heterogeneous oxidation of COS.So the heterogeneous oxidation of COS on multiple component aerosols is as important as that on single component aerosol.

Atmospheric particulate matter comprises a large fraction of organic material.Yang et al.17has found that nitrogen-containing organic salts are widely present in aerosols in the urban atmosphere.Amines are major nitrogen-containing atmospheric organic compound.Low molecular weight alkylamines are emitted by a variety of widespread anthropogenic and biogenic sources including industry emissions，vehicles exhaust，animal husbandry，sewage treatment plants，and incinerators.18−20Field observations have detected the short chain alkylamines in marine，rural and urban atmospheres in the gas phase，particle phase，and aqueous fog and rain drops.21，22Because of their basicity，ammonia，and organic amines are often used as probe molecules to investigate the surface acidity on metal oxides.23−25It has been reported that amines can easily form amine salts with acids or react with O3，OH，or NO3to partition to the particle phase.For example，Qiu and Zhang26found that organic amines can react efficiently with the acidic aerosols(such as sulfuric acid)to form alkylaminium salts，promoting the particle growth.Ge et al.27have investigated the reaction of alkylamine with ozone，and found that the reaction rates of different alkylamines were different，which may explain the intriguing finding in several field studies where organic aerosol were detected in aerosol samples.Therefore，these organic amines may significantly contribute to the formation of atmospheric secondary aerosol.28Besides，ammonia and amines are easy to be absorbed onto atmospheric particles，29which would inevitably alter the surface properties of particles and finally influence the heterogeneous oxidations of COS with these particles.However，the research on how basic substances influence heterogeneous oxidation of COS is limited.In order to obtain information concerning the heterogeneous oxidation of COS on complicated mineral dust particles and to give a further understanding of reaction mechanism，the heterogeneous oxidation of COS on hematite with pre-adsorbed ammonia and amines were studied.

In this paper，the heterogeneous oxidation of COS on hematite were studied using in situ diffuse reflectance infrared Fourier transform spectroscopy(DRIFTS)，ammonia and several amines including methylamine，trimethylamine，triethylamine，phenylamine，pyridine，and pyrrole were selected.The effects of ammonia and amines on the heterogeneous oxidation of COS on hematite were also investigated.We found that ammonia and amines can enhance the reactivity of COS on hematite and the reactivity was different between basic substances.In addition，surface hydroxyl(M―OH)and oxygen ion(M―O−)had a significant influence on the reaction.Surface M―O−was the dominant active site in the conversion of COS in the presence of basic substances.Based on these observations，the reaction mechanism of COS onto hematite was discussed.

2 Experimental

2.1 Materials

Hematite(99.999%purity)was purchased from Sigma-Aldrich.For comparison，hematite was also prepared according to the procedure reported previously.14Powder X-ray diffraction(XRD，Rigaku D/MAX-II X-ray diffractometer with Cu Kαradiation)was used to verify the crystalline structures of the commercial and synthesized samples.Fig.1 shows the X-ray diffraction(XRD patterns of the commercial and synthesized hematite，the characteristic peaks can be indexed according to the hematite crystalline phase(JCPDS，Card No.33-0664).The Brunauer-Emmett-Teller(BET)surface areas of the commercial hematite，determined using an ASAP 2010 automatic equipment at 77 K，was 14.2 m2·g−1.Therefore，the commercial hematite instead of synthesized were used in our experiment.The commercial hematite samples were kept in a desiccator with saturated sodium chloride solution for 48 h before in situ DRIFTS experiments.

Gaseous oxygen(O2，99.999%purity，Shanghai Yunguang Specialty Gases Inc)，argon(Ar，99.999%purity)，carbonyl sulfide[1000×10−6(volume fraction)，COS/N299.999%purity]were introduced through an air dryer before use.

Ammonia(NH3/N2)，methylamine(CH3NH2/N2)，trimethylamine[(CH3)3N/N2]，phenylamine(CH3CH2NH2/N2)，triethylamine[(C2H5)3N/N2]，pyridine(C5H5N/N2)，and pyrrole(C4H5N/N2)，the concentration of these basic substances were 100×10−6and were purchased from Shanghai Qingkuan Trade Co.，Ltd.The coverage of amine absorbing onto the hematite was obtained by titration using hydrochloric acid(HCl，5.15×10−4mol·L−1)as titrating solution.

2.2 In situ DRIFTS experiment

Fig.1 X-ray diffraction patterns of commercial hematite and synthetic hematite

The experiments reported here were performed by using in situ DRIFTS，similar to our previous work.15，16The DRIFTS spectra were recorded on the Nicolet Avatar 360 FTIR spectrometer，which equipped with a high-sensitivity mercury cadmium telluride(MCT)detector cooled by liquid N2and a Spectra-Tech Diffuse Reflectance Accessory.About 30 mg(±0.02 mg)hematite sample was placed into the ceramic crucible in the chamber(the volume of chamber is about 0.025×10−5m3)and the temperature in the chamber was kept at 298 K by using an automatic temperature controller.After being purged with Ar(100 mL·min−1)for 60 min，basic substance(20 mL·min−1)was introduced into the chamber to allow adsorption onto hematite for 15 min，afterwords，Ar(100 mL·min−1)was introduced into the chamber for 60 min again to blow off excess amines，and then a background spectrum was recorded.After collecting the background spectrum，a mixture of gases[COS(50%)，O2(21%)，and Ar(29%)]were introduced into the chamber at a total flow rate of 100 mL·min−1for 3 min，and the inlet and outlet of the chamber was then closed simultaneously，after which the IR spectra in the closed system were recorded automatically every 10 min for 4 h.All the IR spectra were recorded at a resolution of 4 cm−1for 100 scans.

3 Results and discussion

3.1 In situ DRIFTS monitoring of heterogeneous oxidation of COS on hematite with pre-adsorbed methylamine

Fig.2 In situ DRIFT spectra of hematite pre-adsorbed by methylamine(MA)as a function of time after exposure to 5×10−4(volume fraction)COS at 298 K

Fig.2 shows in situ DRIFT spectra of the heterogeneous oxidation of 5×10−4(volume fraction)COS on hematite pre-adsorbed by methylamine(MA)as a function of time at 298 K.As can be seen in Fig.2(A)，a pair of peaks at 2071 and 2051 cm−1，which was the characteristic absorption peaks of gaseous COS，30decreased drastically in intensity as the reaction processed，indicating that COS can react on the surface of hematite in the presence of methylamine.The decrease of gaseous COS was accompanied with the increase of absorption at 2341 and 2360 cm−1by gaseous CO2.31In Fig.2(B)，the bands observed at 1226，1385，and 1652 cm−1were the characteristics of the formation of adsorbed bicarbonate，corresponding to the δ(COH)，ν3(OCO)s，and ν2(OCO)avibrational modes，respectively.32−35The bands at 1436 and 1326 cm−1were attributed to the ν(OCO)aand ν(OCO)svibrational modes of carbonate，and the bands at 1162，1077，and 965 cm−1arose from the surface.These results suggest that COS is converted into gas-phase CO2，surface HCO3−，surface，and surfacespecies on hematite with methylamine pre-adsorbed.

Previous studies10，14had reported obvious negative peak of hydroxyl groups(OH)appeared in the higher wavenumber spectral region due to its consumption during the heterogeneous oxidation of COS.However，in our research，surface M―OH at 3650 cm−1remained very weak during the oxidation process(Fig.2(A))，which inferred that O2−(i.e.，Fe―O−)might be the dominant active site for the conversion of COS in the presence of ammonia and amines.Fig.3 shows DRIFT spectra of hematite with only MA exposure before the introduction of COS.The loss of M―OH band in the region from 3704−3550 cm−1after adsorption of MA on hematitewas clearly observed.The band at 2953 cm−1arose from the asymmetric stretching of CH3in methylammonium salt，the band at 1418 cm−1from the C―H bending mode，and the band from 1586 to 1508 cm−1from N―H bond of primary ammonium salt.36These results indicated that the adsorbed MA could interact with M―OH on hematite to form methylammonium salt.Therefore，if methylamine was introduced in this system，M―OH would be converted to M―O−active site and consequently resulting in the reduced role of M―OH in the conversion of COS.

3.2 Effects of different amines on the heterogeneous oxidations of COS on hematite

Fig.3 In situ DRIFT spectrum of hematite with methylamine exposure at 298 K

To compare the effects of different amines on the heterogeneous oxidation of COS，hematite samples pre-adsorbed by trimethylamine，triethylamine，phenylamine，pyridine，pyrrole，and ammonia were exposed to reactant gases.The pre-adsorption time of ammonia and different amines prior to the exposure of COS remained the same.The pre-adsorption capacities of ammonia and different amines on hematite surface may be different under the same pre-adsorption time.Here we compared their reactivity towards the COS heterogeneous conversion approximately.Spectral results showed oxidation of COS formed the similar products.The concentrations of COS can be obtained from the absorbance peak area of COS between 2092 and 2015 cm−1.The blank experiment showed that wall losses of COS were of no importance in this chamber.Fig.4 shows the concentration of COS against the reaction time with ammonia and amines pre-adsorbed onto hematite.It is evident that the discrepancies between these basic substances are large.The reaction rates influenced by these basic substances are in the order of methylamine＞trimethylamine＞ammonia＞triethylamine＞pyridine＞pyrrole＞phenylamine≈pure hematite.Affected by basic substances，the heterogeneous oxidation of COS occurs rapidly at the initial stage.With the consumption of the adsorbed COS，the surface products gradually occupied the active sites on hematite，resulting in the slower reaction with the increasing of reaction time.

Although the concentration data show that the reactivity of COS on the surface of hematite depends on the properties of basic substances，it is worthwhile to discuss the kinetics of COS conversion.In order to determine the reaction order with respect to COS，three replication experiments were conducted for each sample.ln[COS]and[COS]−1were plotted in two different ways.The one is either ln[COS]or[COS]−1are plotted versus time，37and the other is the bilogarithmic plot.38，39For a first order reaction，a plot of ln[COS]versustime should be linear，while for a second order reaction，a plot of[COS]−1as function of time should be linear.We designated these plots as Type A plots，and Fig.5 shows the result described in this manner.As can be seen from Fig.5，the reaction is obviously second order for COS on pre-adsorbed sample.

Moreover，the oxidation of COS on hematiteis related to the active sites on the surface of hematite，and the kinetics of oxidation of COS with hematite may be described by the general equation:

Fig.4 Concentrations of COS as a function of time with different pre-adsorbed basic substances

Fig.5 First and second order Type Akinetics plots for the loss of COS with trimethylamine pre-adsorption

where k is the rate constant，[hematite]is the concentration of active sites on hematite，[COS]is the concentration of COS at certain reaction time(we regarded the absorbance peak area of COS as its concentration)，[amine]is the concentration of basic substance onto hematite，and m，n，p are the reaction orders for COS，hematite，and amine，respectively.As the concentration of active sites on hematiteisusually considered to be kept constant with respect to COS，[hematite](which is the case at the initial stage)can be assumed to be constant.39

The second approach utilizes the rate law using Eq.(2)obtained from the logarithmic transformation of Eq.(1).In the Eq.(2)，ln[amine]may be not a constant，which make more difficult to analyze the plot.In order to simplify this analysis，we assumed pln[amine]be constant and then analyzed the bilogarithmic plot.We designate this as Type B plot.Fig.6 shows the napierian logarithm of the initial rate of COS loss versus napierian logarithm of the initial COS concentration.As can be seen from Fig.6，the reaction order m for COS on hematite determined by the slope was m=1.89(±0.09)，which confirmed that the reaction order for COS was second.Therefore，the assumption that pln[amine]as a constant is valid.Namely，either p=0 or the concentration of basic substance was constant.Similarly，the reaction order for COS was also verified to be second when pre-loaded by other basic substances.Therefore，basic substances can influence the reaction order of COS，resulting in great enhancement of the conversion reactivity of aerosol.

Fig.6 Bilogarthnic plot of the initial rate of COS consumption onto hematite as a function of[COS]at 298 K with trimethylamine pre-adsorption

As shown in Fig.4，the observed oxidation of COS occurred rapidly in 120 min，so the consumption reactivity of COS in 120 min were calculated and Fig.7 shows the effects of different basic substances on the reaction rates of COS oxidation.As can be seen from Fig.7，hematite pre-adsorbed by basic substances showed different reactivity.Hematitewith methylamine pre-adsorbed has the highest reactivity，about 4.5 times higher than that of pure hematite，while the effect of phenylamine and pyrrole is not obvious.Given the different nature of these basic substances themselves，the coverage，adsorption ability，basicity and other properties on the surface of hematite are not consistent even though being introduced for the same time period，thus the reactivity of COS is different(as shown in Fig.7).The differences in Fig.7 resulted from two possible conditions.If the final coverage of all the basic substances is approximately the same，the basicity of these amines will play the main role in the reactivity of conversion of COS.Higher basicity of the amine can facilitate the conversion of COS and the reactivity has positive correlation with the basicity of pre-adsorbed amines.Pyrrole and phenylamine have the weakest weak basicities and pyrrole is even a weak acid.Therefore，the reactivity affected by these two amines was closed to that of pure hematite.Meanwhile，the basicities of methylamine，trimethylamine and triethylamine are close to one another(pKa=10.66 for methylamine，9.81 for trimethylamine and 10.72 for triethylamine)，but the order of heterogeneous reaction rates is methylamine＞trimethylamine＞triethylamine，for which the steric hindrance of amines may be another factor governing the reactivity.The large methyl groups in trimethylamine and the large ethyl groups in triethylamine may obstruct the interaction between COS and surface active sites distributed on the hematite，resulting in the decrease of the reactivity.If the coverage of these basic substances under the same pre-adsorption time is different the synergistic effect of coverage and basicity of basic substances will result in different reactivity of COS.Both can increase the surface basicity of hematite as well as the amount of M―O−，which enhances the reactivity of COS.

Fig.7 Effects of basic substances on the rates of COS oxidation on the surface of hematite

3.3 Effect of coverage of trimethylamine on heterogeneous oxidation of COS on hematite

To study the coverage of amines，a series of experiments with different amines exposure time were carried.Titration by hydrochloric acid(HCl，5.15×10−4mol·L−1)was used to obtain the coverage of amine absorbing onto the hematite.40The coverages on hematite samples with trimethylamine pre-adsorption for 5，10，15，and 30 min were 1.24×1018，1.44×1018，1.72×1018，and 2.06×1018molecules·g−1，respectively.Obviously，the coverage of trimethylamine increased over time.Correspondingly，the conversion reactivity of COS increased with the coverage of trimethylamine(Fig.8).For different basic substance，this trend keeps the same.Therefore，the coverage of basic substances is also the factor that enhances the heterogeneous reactivity.

3.4 Role of surface adsorbed water on

heterogeneous oxidation of COS on hematite Surface water is an important determinant in atmospheric heterogeneous process and hydrolysis of COS on oxides，an effective way to remove COS in industrial tail gas，is strongly affected by the amount of adsorbed water.41Therefore，the role of surface water cannot be neglected.In order to study the role of surface water，experiments were performed at 298 K over a series of hematite samples.The hematite samples had been pre-evacuated for 60 min at 298，348，398，448，and 498 K for 60 min，and then trimethylamine pre-adsorbed on their surface of hematite sample.Results shown in Fig.9 suggest that the reactivity of hematite decrease with an increase in temperature，e.g.，the reactivity of COS at 498 K is about 2 times lower than that at 298 K.Pre-evacuation of hematite is known to mainly reduce the surface water adsorbed at low temperature，while surface dehydration occurs at higher temperature.The observed decrease in the conversion reactivity of COS was related to the decrease of some physisorbed water and the removal of some M―OH.At low temperature，the physisorbed water，which is not available to the re-generation of M―OH，desorbes as temperature increase.At high temperature，both surface adsorbed water and some M―OH were removed，leading to the decrease in trimethylamine adsorption and subsequently the decrease of surface Fe―O−formed from the interaction between Fe―OH and basic substances.Consequently，the reactivity decreases at higher temperatures.This result clearly indicates that the surface water plays a significant role in the conversion of COS with basic substance pre-adsorption.

Fig.8 Effect of coverage for trimethylamine(20 mL·min−1)on the reactivity of COS

Fig.9 Second-order rate coefficients for COS conversion on hematite as a function of pre-treatment temperature

3.5 Mechanism of heterogeneous oxidation of COS on surface of hematite

The metal oxides with empty or half-empty d atom orbits，such as hematite，is apt to adsorbing O2and H2O(g)and to form the surface hydroxyl.9As proposed by He et al.，10−13the oxidation of COS involves the formation of intermediate product HSCO2−，and M―OH were the main active sites for the heterogeneous reaction of COS.Chen et al.14also verified M―OH and M―O−were main active sites and confirmed that M―O−played a dominant role in the loss of COS on hematite.As shown above，M―OH and M―O−on hematite are crucial to the conversion of COS.It was confirmed by our experiment that reactivity of hematite pre-adsorbed by basic substances was much higher than that of pure hematite.This discrepancy in reactivity can be related to the pre-adsorption of basic substances.Amines and ammonia adsorb onto the oxide surface，resulting in the conversion of active sites due to their interaction with Lewis and Brönsted acid sites distributed on the surface.23Simmons and Beard42proposed that surface M―OH can interact with an organic base as follows:

where X is atomic nitrogen and MO−is surface oxygen ions that can react with COS.When ammonia and amines were introduced into the reaction system，they can interact with M―OH，converting M―OH to M―O−active sites.Chen et al.14found that M―O−played a dominant role in the conversion of COS on hematite.Generally，metal oxide surface has a large number of M―OH and a small amount of M―O−.In this study，the role of each M―OH may be weaker than that of M―O−in the conversion of COS when on ammonia and amines were not introduced into the reaction system，but their overall reactivity was great because of their large number.When a small amount of ammonia or amine was introduced into the reaction system，more M―O−is generated，enhancing the reactivity of COS greatly due to its strong role in the conversion of COS.In addition，for the nucleophilic attack on the positively charged carbonyl carbon of COS，the negatively charged M―O−is easier than that of electrically neutral M―OH.Therefore，M―O−will play a major role in the conversion of COS when more M―O−is available.This may be the reason why the increase of the negative peaks of M―OH during the reaction is not obvious.Meanwhile，basic ammonia and amines on hematite can neutralize the H2CO3and acidic sulfate formed during the heterogeneous oxidation of COS，and make heterogeneous oxidation of COS proceed.For basic substances with strong basicity，its neutralization is easier，resulting in more high reactivity of COS.

On the basis of the discussion above，we proposed the following mechanism of the heterogeneous oxidation of COS on hematite.Ammonia and amines pre-adsorbed onto hematite first reacted with Fe―OH to form more Fe―O−，Fe―O−，and Fe―OH attacked the carbon of COS，forming HSCO2−species on the surface of hematite.Then HSCO2−species can be oxidized by oxygen，followed by the formation of surface HCO3−，surface SO42−species and gaseous CO2.The intermediate product(HSCO2−)can easily interact with O2.Therefore，the intermediate with low concentration were not observed in DRIFTS spectrum.Moreover，basic substances neutralized the acidic products and made heterogeneous oxidation of COS proceed，enhancing the reactivity.In addition，complex formation of COS with amine may also lead to other mechanism of the heterogeneous reaction and further studies are needed to clarify these aspects.

4 Conclusions and atmospheric implications

The heterogeneous oxidation of COS on hematite pre-adsorbed by ammonia and amines(here methylamine，trimethylamine，triethylamine，phenylamine，pyridine，and pyrrole)were investigated by using DRIFTS.Gaseous surfaceandspecies as final products were observed by DRIFTS.The study on kinetics showed that these pre-adsorbed basic substances on hematite can significantly enhance the heterogeneous reactivity and change the reaction order from first to second.The basicity and coverage of basic substances as well as surface water also played important roles in the heterogeneous oxidation of COS.As pointed out in the introduction section of COS，the global sources of COS were less than its sinks，which indicates the atmospheric process of COS still has uncertainties.Therefore，the oxidation of COS on multiple component aerosols is as important as that on signal component aerosol.These findings indicated that the multiple component aerosols may implicate a complicated chemical process and the heterogeneous reaction of trace gases on real atmospheric particles，which provided more data for atmospheric chemistry，should be considered.

(1) Turco，R.P.;Whitten，R.C.;Toon，O.B.;Pollack，J.B.;Hamill，P.Nature 1980，283，283.doi:10.1038/283283a0

(2) Sze，N.D.;Ko，M.K.W.Nature 1979，278，731.doi:10.1038/278731a0

(3) Sze，N.D.;Ko，M.K.W.Nature 1979，280，308.doi:10.1038/280308a0

(4) Sze，N.D.;Ko，M.K.W.Atmos.Environ.1980，14，1223.doi:10.1016/0004-6981(80)90225-5

(5) Svoronos，P.D.N.;Bruno，T.J.Ind.Eng.Chem.Res.2002，41，5321.doi:10.1021/ie020365n

(6) Torres，A.L.;Maroulis，P.J.;Goldberg，A.B.;Bandy，A.R.J.Geophys.Res.1980，85，7357.doi:10.1029/JC085iC12p07357

(7) Rasmussen，R.A.;Khalil，M.A.K.;Hoyt，S.D.Atmos.Environ.1982，16，1591.doi:10.1016/0004-6981(82)90111-1

(8) Watts，S.F.Atmos.Environ.2000，34，761.doi:10.1016/S1352-2310(99)00342-8

(9)Wu，H.B.;Wang，X.;Chen，J.M.;Yu，H.K.;Xue，H.X.;Pan，X.X.;Hou，H.Q.Chin.Sci.Bull.2004，49，739.[吴洪波，王晓，陈建民，俞宏坤，薛华欣，潘循晳，侯惠奇.科学通报，2004，49，739.]doi:10.1360/03wb0132

(10) He，H.;Liu，J.F.;Mu，Y.J.;Yu，Y.B.;Chen，M.X.Environ.Sci.Technol.2005，39，9637.doi:10.1021/es048865q

(11) Liu，J.F.;Yu，Y.B.;Mu，Y.J.;He，H.J.Phys.Chem.B 2006，110，3225.doi:10.1021/jp055646y

(12) Liu，Y.C.;He，H.;Xu，W.Q.;Yu，Y.B.J.Phys.Chem.A 2007，111，4333.doi:10.1021/jp069015v

(13) Liu，Y.C.;He，H.J.Phys.Chem.A 2009，113，3387.doi:10.1021/jp809887c

(14) Chen，H.H.;Kong，L.D.;Chen，J.M.;Zhang，R.Y.;Wang，L.Environ.Sci.Technol.2007，41，6484.doi:10.1021/es070717n

(15) Yu，Y.J.;Zhang，S.Q.;Kong，L.D.;Lin，L.;Cheng，T.T.;Chen，J.M.Acta Phys.-Chim.Sin.2011，27，2275.[俞偐偼，张拴勤，孔令东，林 立，成天涛，陈建民.物理化学学报，2011，27，2275.]doi:10.3866/PKU.WHXB20110912

(16) Usher，C.R.;Michel，A.E.;Grassian，V.H.Chem.Rev.2003，103，4883.doi:10.1021/cr020657y

(17) Wang，X.F.;Gao，S.;Yang，X.;Chen，H.;Chen，J.M.;Zhuang，G.S.;Surratt，J.D.;Chan，M.N.;Seinfeld，J.H.Environ.Sci.Technol.2010，44，4441.doi:10.1021/es1001117

(18) Cadle，S.H.;Mulawa，P.A.Environ.Sci.Technol.1980，14，718.doi:10.1021/es60166a011

(19) Westerholm，R.;Li，H.;Almen，J.Chemosphere 1993，27，1381.doi:10.1016/0045-6535(93)90231-S

(20) Chen，C.P.;Veregin，R.P.;Harbour，J.R.;Hair，M.L.Chin.Sci.Bull.1994，39，744.[陈次平，Veregin，R.P.，Harbour，J.R.，Hair，M.L.科学通报，1994，39，744.]

(21) Vanneste，A.;Duce，R.A.;Lee，C.Geophys.Res.Lett.1987，14，711.doi:10.1029/GL014i007p00711

(22) Zhang，Q.;Anastasio，C.Atmos.Environ.2003，37，2247.doi:10.1016/S1352-2310(03)00127-4

(23) Lavalley，J.C.Catal.Today 1996，27，377.doi:10.1016/0920-5861(95)00161-1

(24) Lercher，J.A.;Grundling，C.;EderMirth，G.Catal.Today 1996，27，353.doi:10.1016/0920-5861(95)00248-0

(25) Sarria，F.R.;Blasin-Aube，V.;Saussey，J.;Marie，O.;Daturi，M.J.Phys.Chem.B 2006，110，13130.doi:10.1021/jp061729i

(26) Qiu，C.;Zhang，R.Y.Environ.Sci.Technol.2012，46，4474.doi:10.1021/es3004377

(27)Gai，Y.B.;Ge，M.F.;Wang，W.G.Acta Phys.-Chim.Sin.2010，26，1768.[盖艳波，葛茂发，王玮罡.物理化学学报，2010，26，1768.]doi:10.3866/PKU.WHXB20100705

(28)Yin，S.;Ge，M.F.;Wang，W.G.;Liu，Z.;Wang，D.X.Chin.Sci.Bull.2011，56，1241.[殷 实，葛茂发，王炜罡，刘 泽，王殿勋.科学通报，2011，56，1241.]

(29) Qiu，C.;Wang，L.;Lal，V.;Khalizov，A.F.;Zhang，R.Y.Environ.Sci.Technol.2011，45，4748.doi:10.1021/es1043112

(30) Dohrmann，J.;Glebov，A.;Toennies，J.P.;Weiss，H.Surf.Sci.1996，368，118.doi:10.1016/S0039-6028(96)01038-2

(31) Amenomiya，Y.;Morikawa，Y.;Pleizier，G.J.Catal.1977，46，431.doi:10.1016/0021-9517(77)90230-5

(32) Turek，A.M.;Wachs，I.E.;Decanio，E.J.Phys.Chem.1992，96，5000.doi:10.1021/j100191a050

(33) Lavalley，J.C.;Travert，J.;Chevreau，T.;Lamotte，J.;Saur，O.J.Chem.Sci.Chem.Commun.1979，146.

(34) Morterra，C.;Zecchina，A.;Coluccia，S.;Chiorino，A.J.Chem.Soc.Faraday Trans I.1977，73，1544.doi:10.1039/f19777301544

(35) Molina，R.;Centeno，M.A.;Poncelet，G.J.Phys.Chem.B 1999，103，6036.

(36) The Sadtler Handbook of Infrared Spectra.Bio-Rad Laboratories，Inc.，Informatics Divison:Htercules，California，USA，1978−2004.

(37) Finlayson-Pitts，B.J.;Wingen，L.M.;Sumber，A.L.;Syomin，D.;Ramazan，K.A.Phys.Chem.Chem.Phys.2003，5，223.doi:10.1039/b208564j

(38) Xu，B.Y.;Zhu，T.;Tang，X.Y.;Ding，J.;Li，H.J.Chem.J.Chin.Univ.2006，27，1912.[徐冰烨，朱 彤，唐孝炎，丁 杰，李宏军.高等学校化学学报，2006，27，1912.]

(39) Borensen，C.;Kirchner，U.;Scheer，V.;Vogt，R.;Zellner，R.J.Phys.Chem.A 2000，104，5036.doi:10.1021/jp994170d

(40)Li，Q.X.;Hou，S.Z.;Xing，Y.Z.;Wei，H.W.;Li，M.China Surfactant Detergent&Consmetics 2000，30，50. [李秋小，侯素珍，邢英站，魏海威，李 明.日用化学工业，2000，30，50.]

(41) Rhodes，C.;Riddel，S.A.;West，J.;Williams，B.P.;Hutchings，G.J.Catal.Today 2000，59，443.doi:10.1016/S0920-5861(00)00309-6

(42) Simmons，G.W.;Beard，B.C.J.Phys.Chem.1987，91，1143.doi:10.1021/j100289a025