26 Pages
Read an excerpt
Gain access to the library to view online
Learn more


PRESENTED  BY :   MAR ION   FRU ECHTL   SUPERV I SORS :   PROF .   T .   RÖCKMANN     DR .   C .   J ANSSEN   Laboratory Measurements of the Anomalous Isotopic Composition of Ozone

  • iso- topic species proceeds

  • ?? rh

  •  o2  

  • altitude dependence

  • ozone isotope

  • ozone

  • mass-dependent species

  •  h?  à?

  • stratospheric ozone isotopes



Published by
Reads 14
Language English
Document size 5 MB
Report a problem

Laboratory Measurements of the
Anomalous Isotopic Composition
of Ozone

DR.C.JANSSENRole of ozone
-OHprecursor §  UVabsorp/on
3 31 §  O +hνàO( P)+O( P)2§  O +hνàO( D)+O3 2
31 O( P)+O +MàO +MO( D)+H OàOH+OH 2 32
O +hνàO+O3 2
O+O àO +O 3 2 2
Pollutant Greenhousegas
§  absorbsandemitsIRradiaon§  oxidizesbiological/ssue
2O2 §  radiave forcingTP:0.4W/m
§  RH+OHàRO +H O2 2 2§  radiave forcingSP:-0.05W/m
O 2 IsotopeMarker
NO +hνàNO+O2 3
§  Sensi/vemarkerofpar/cular
§  Transferofisotopeanomalyinto
othertracegases“Initial Training Network on Mass-Independent
+ NO + SO + CH2 2 4 2-N O NO SO OH H O 4 2 5 3 2
+ H O 2
+ NO + H O + SO+ NO 2 22 3
+ OH + NO + hv + OH 1HNO NO O( D) CO CO 3 2 O 23
+ N O 2+ O + NH + C H2 5 8
+ hv
CO NO O N O 2aerosol 2
ANRV273-EA34-08 ARI 22March2006 18:1
17 The O anomaly in the atmosphere
O strato.3
17" Ospecies O topo.380
17 17 18Δ O=δ O–0.52*(δ O)
NO3 δ-nota<-n:
CO 182 $ ' 40 R18strato. SA" O = #1 *1000 ‰& ) 18R% ( STDH O2 2 CO tropo.20 2
Air O Mass-dependentSO 24
Bulk Earth’s silicate species
Standard mean ocean water
Atmospheric H O ! 2
-50 0 50 100 150 200
18 0! O ( / )00SMOW
(Thiemens 2006) Figure 7
18andC O.Undernonsaturationconditions,thephotodissociationoftheminoriso-
in a three-isotope oxygen plot defines a slope 1. There are now measurements of
17C O in interstellar molecular clouds, although the errors associated with the ob-
servations are large compared with meteoritic isotope measurements (Bensch et al.
(2004) provide details of astronomical observations of interstellar regions and their
Following the suggestion of the self-shielding process in 1983 as a means to
produce meteoritic oxygen isotopic compositions (Thiemens & Heidenreich 1983),
the self-shielding process occurring in the early Solar System. The analysis demon-
strated that the rapidity of oxygen isotope exchange reactions is a major limit on
!3metalatomsorhydrogenisinefficientatpressureslowerthan10 atmandanebular
232 Thiemens
Annu. Rev. Earth Planet. Sci. 2006.34:217-262. Downloaded from
by Utrecht University on 09/06/11. For personal use only.
! O ( / )
SMOW O anomaly 3
Stratosphere Troposphere
49 50Figure 2. Altitude profiles of enrichment of (left) O and (right) O measured during five balloon3 3
flights at middle latitude (Aire-sur-l’Adour, France). Horizontal error bars are 2s errors while the vertical(Krankowsky et al. 1995) (Krankowsky et al. 2007)
bars indicate the height range over which ozone was collected. The lines represent the enrichment
expected from measured atmospheric temperature profiles and the known temperature dependence of
isotope fractionation in ozone formation (see Figure 1).
of the total values measured. There appears to be a higher stratosphere increases the isotope ratios increase as well, at
contribution of this additional isotope fractionation above middlelatitudesfrom6.5%in20kmtoabout7.5%in28km
49 50the equator compared to middle latitudes. for O and from 7% to 8% for O . In the cold polar3 3
stratosphere at temperatures below 200 K both heavy
3.3. Polar Region, Kiruna, Sweden isotope enrichments are lower and nearly equal around
[15]ThefourflightsfromKirunawereflowninNovember/ 6%. When toward summer the temperatures in the strato-
DecemberandinMayandresultsarealsoshowninFigure3. sphere rise enrichments increase accordingly without a sign
Unfortunately in one May flight a descent system was not of an additional effect even at altitudes of 30 km. Very
part of the payload and all 4 data were obtained around different is the situation at middle and equatorial latitudes:
22 km. For both winter flights the ozone isotope enrich- The precision and accuracy of the measurements and the
ments are substantially lower and almost equal, in very better known temperature dependence permit a quantifica-
good agreement with prediction from the temperature tion of an additional isotope effect which is more pro-
dependence. Atmospheric temperatures were very low, nounced above the equator than at middle latitudes. In past
particularly during the flight on 28 November 2002 when publications the source of this effect has been discussed a
the payload was inside the cold polar vortex for most of the number of times: Miller et al. [2005] identified the Chap-
time. Considerably different is the situation in May. As puis, Huggins, and Hartley band photolysis of ozone, and
expected,withhigheratmospherictemperaturesofabout230 their calculations show that both altitude dependence as
to 240 K the enrichments are much higher. Averaging those well as the magnitude of the effect agree quite well with our
49values shows that atmospheric temperatures still determine middle latitude data at 32 km which are 1.9% ( O)and3
50the isotope ratios with no or little additional contributions. 2.7% ( O ). At the same altitude the calculation by Miller3
et al. [2005] results in 1.5% and 2.9%, respectively. The
equatorial balloon data suggest even higher fractionations4. Conclusions
than Miller et al. [2005] predict. In both cases the fraction-
[16]Alargedatabaseofstratosphericozoneisotopedata ation in the heavier isotopologue is about twice that in the49 50for both O and O is now available, covering the3 3 lighter. Liang et al. [2006] concluded that the ozone
Northern Hemisphere. Very high enrichments, found in formation process is primarily responsible for the enrich-
earlier flights, were not confirmed in the samples analyzed ments but that there is an additional contribution of a few
from the eleven balloon flights presented in this paper, percent at high stratospheric altitudes. Remote sensing
consistent with the reanalysis of older data in the work by results [Johnson et al.,2000;Haverd et al.,2005]can
Mauersberger et al. [2001]. The temperature of the atmo- 16 16 18 16 18 16separate the O O O from the O O Omolecules,but
sphere determines foremost the magnitude of the enrich- the absolute accuracy and precision of the data are not
ments in both heavy isotopomers. As the temperature in the enough to resolve latitudinal variations. Haverd et al.
effect on ozone isotopic composition, since the sum of the
reaction rates of these processes is at least 2 orders of
magnitude less than the photolysis rate. Following Johnson
668 686et al. [2000], we assume that O , O and O are in3 3 3
REVIEW ARTICLE photochemical steady state and obtain the following
expressions for the modeled fractionations,
Dissociation of ozone by reaction with NO was studied general, photo-dissociation in the UV range involves 2
2k j b6 3686d O $ recen%1 tly by Chakraborty "9# and Thiemens (pers. commun.). both reactions (7) and (8), whereas in visible wavelengths 3
k j2 7 The left-over ozone shows a normal mass-dependent only reaction (7) takes place. Reactions (7) and (8) have
!"enrichment, as seen in most of the chemical reactions. been decoupled for UV range photo-dissociation in a
j k "#1%b k4 6668 3 40d O $ ! %1 "10#3 Among the other two dissociation pathways, two ear- recent experiment . Knowing the fact that N quenches 2j k K k2 eq 25 1 –11 3lier studies dealt with photo-dissociation of ozone by UV O( D) quite efficiently (2.6 " 10 cm /mol/s at
37,39 21where k is the rate constant for thand e ivisible th reactlightion, j is, thewhich showed slight deviation from 298 K) , O photo-dissociation was carried out with a i i 3
photolysis rate for the ith reaction, k =k +k , b=k /mass dependence in the case of UV dissociation. In the Hg resonance lamp in the presence of N at different par-6 6a 6b 6a 2
(k +k)andK =k/k .Thepurephotolytic6a 6b eq 8f 8rfollowing section, we will discuss the present under- tial pressures. It is noted that at lower N pressure, the 2
fractionations are obtained by setting k /k =k/(k +2 4 2 6astanding about photo-dissociation and surface-induced slope is 0.63 as observed for UV range photo-dissocia-
Figure 2. Absolute (not fractional) precisions (see text) as k )=1,k /k =1andk /k =2inequations(8)and(9)6b 6a 6b 8f 8r dissociation of ozone. tion, but with increasing N pressure, the slope gradually 686 668 2a function of altitude for d O (left) and d O (right) and rearranging in terms of the total fractionations:3 3 The photo-dissociation of ozone produces isotopically increases and reaches a saturation value of unity. It is
#$ light oxygen and as a consequence the left-over ozone interpreted by the authors that at higher N pressure, j 2686 3d O $ %1 13 phot becomes heavier. However, the fractionation pattern is quenching of O( D) by N is more efficient compared to j 27in fractionation, rather than absolute fractionations. The
668 686 #$ not similar in visible and UV wavelengths. In the visible quenching by O , and, therefore, the follow-up reaction k 32 686generalincreasesinboth d O and d O withaltitudeare3 3 $ d O !1 %1 "11# 17 183 range it is strictly mass-dependent (!! O/!! O = 0.54), (8) cannot take place. Therefore, the change in slope re-2k binterpreted in the following discussion. 6
whereas in the UV region it shows deviation from mass flects that pure UV photo-dissociation (reaction (7)) is a [8]Ozoneformation,
17 18!" %1dependency (!! O/!! O = 0.63, Figure 2, after Chak- mass-independent process, which proceeds with a slope #$j k "#1%b"#k !k4 6a 6b668 3O!O !M! O "2#2 3 40 17 18d O $ %1$ !3 raborty and Bhattacharya ). In a critical analysis, it was (!! O/!! O) value of unity. In contrast, reaction (8) phot j k K k2 eq 25
#$is generally considered the major source of ozone isotopic seen that the product O-atom in reaction eq. (7) is in the has the property of a normal chemical reaction and these 668& d O !1 %1 "12#3 1fractionation in the stratosphere [e.g., Brenninkmeijer et al., excited state (O( D)) for higher photon energy (" < two reactions (7) and (8) together give rise to the devia-
2003]. However laboratory experiments on ozone photo- 310 nm), and for lower photon energies (" > 310 nm), the tion from mass dependency during UV photo-dissocia-
We substitute the observed fractionations into (11) and (12), 3dissociation [e.g., Bhattacharya and Thiemens,1988]and product O-atom is in ground state (O( P)) (ref. 41). The tion. 32/Ttogether with the equilibrium constant K =1.94eeqrecent analysis by Miller et al. [2005] suggest that ozone O3 anoreaction marate lycoefficient in dof eq. es(8) t is r comparatively uctionvery re An acinteresting tion feature s has recently been discovered [Kaye and Strobel, 1983] ad the ozone-formation ratephotolysis should also cause significant isotopic fractiona- –10 3 21 35,42high (1.1 " 10 cm /mol/s at 298 K) when the during ozone dissociation induced by a surface . It is 686 parameters of Janssen et al. [2003]: k /k =0.93+1.03'4 2tion.Hereweinferthephotolyticfractionations,(d O )3 phot 1%3 %5668 participating O-atom is in the O( D) state compared to seen that dissociation yields isotopically light oxygen and 10 (T % 298); k /k =1.27+2.0' 10 (T % 298) andand (d O ) from our observations of total fractiona- 6 23 phot 3
686 668 when the participating O-atom is in the O( P) state the left-over ozone gets enriched in a mass-independent b = 0.427, evaluated at the local temperature. The localtions d O and d O ,usingamodelcomprisingreactions3 3 –15 3 2118 temperature is expected to be th(8.0e te m" p10eratu cmre o/mol/sec f ozone at 298 K) . Therefore, in fashion. (2)–(8) with Q = O.
formation, since the photolytic lifetime of ozone is shortAtmos phere Laboratory
O !hn! O !O "3# 3 2
O !Q!M! OOQ!M "4#2 Left-over Ozone
OOQ!hn! O !Q "5a#2
! O!OQ "5b# Dissociation by UV wavelengths
50 (Slope = 0.63 ± 0.03)
OQ!O!M! OOQ!M "6a#
! OQO!M "6b#
OQO!hn! O!OQ "7#
Dissociation by visible wavelengths
10 (Slope = 0.54 ± 0.02)
Q!O ! O!OQ "8a#2
-30 10 50 90
O!OQ! Q!O "8b#2
18!" O (‰)Initial Ozone
We assume that no other reactions significantly affect the Product Oxygen50isotopic composition of ozone. In particular, we do not Figure 3. Vertical profiles of O.Solidcircles:mas-3 -30
include ozone removal processes other than photolysis; in spectrometric data [Mauersberger et al.,201].Other
the 12–40 km altitude range, these will have a negligible symbols: this work. 17 18Figure 2. Covariation plot between !! O and !! O for photo-dissociation of ozone in UV (Haverd et al. 2005) ( Chakraborty & Bhattacharya 2003c)
(circles) and visible wavelengths (squares). The left-over ozone (filled symbols) is enriched and
3of4 the product oxygen (unfilled symbols) is depleted with respect to the initial ozone. The best-fit
line for the visible light dissociation gives a slope of 0.54 ± 0.02 indicating a mass-dependent
process, while the UV dissociation gives a slope of 0.63 ± 0.03 reflecting slight deviation from
40mass dependency (after Chakraborty and Bhattacharya ).
CURRENT SCIENCE, VOL. 84, NO. 6, 25 MARCH 2003 769
O (‰)
!"Visible light lamp
Experimental Setup
Low Vacuum
High PS PS
sample (HVP)
Ni - foil
UV-C lamp
Dual Inlet
Reaction HVP LN2 chamber
Cold trap Visible light lamp
Experimental Setup
Low Vacuum
High PS PS
sample (HVP)
Ni - foil
UV-C lamp
Dual Inlet
Reaction HVP LN2 chamber
Cold trap PhotolysisVisible light lamp
Experimental Setup
Low Vacuum
High PS PS
sample (HVP)
Ni - foil
UV-C lamp
Dual Inlet
Reaction HVP LN2 chamber
Cold trap PhotolysisVisible light lamp
Experimental Setup
Low Vacuum
High PS PS
sample (HVP)
Ni - foil
UV-C lamp
Dual Inlet
Reaction HVP LN2 chamber Analysis
Cold trap Photolysis