E. Arijs, D. Nevejans, J. Ingels and P. Frederick

GEOPHYSICAL RESEARCH LETTERS, VOL. 10, NO. 4, PAGES 329-332, APRIL 1983

SULFURIC ACID VAPOUR DERIVATIONS FROM NEGATIVE ION COMPOSITION DATA BETWEEN 25 AND 34 KM

E. Arijs, D. Nevejans, J. Ingels and P. Frederick

Belgian Institute for Space Aeronomy, Ringlaan 3, B-1180 Brussels, Belgium.

Abstract : Negative ion composition measure- Results and discussion ments obtained during a valve controlled balloon

flight were used to derive (H2SO 4 + HSO ) number Two typical spectra obtained in the moderate densities between 25 and 34 km altitude.YThe data resolution mode are shown in figure 1. As

are compared to similar results obtained for observed, the amplitudes of the mass peaks due to

other stratospheric temperatures. The implica- HSOT(H•O_) (H. SO,) _{• 3 m z w} ions (mass numbers 160 195 _'

tions of the results on our present ideas about and 293) decrease •ith decreasing altitude, which

aerosol formation are briefly discussed. illustrates the variation of sulfuric acid number density with altitude.

Introduction As pointed out by Arnold and Fabian (1980) the

sulfuric acid number density [H2SO4] can be Although H2SO 4 plays a key role in the forma- deduced through the formula :

tion of aerosols (Turco et al., 1982), measure-

_

merits of its concentration in the stratosphere k[•] [H2SO4] = a[n +] [ns] (1)

have not been possible until very recently.

Stratospheric negative ion composition measure- where [n +] is the total positive ion number

ments (Arnold and Henschen, 1978; Arijs et al., density, a the ion-ion recombination coefficient, 1981) have shown the existence of HSO. cluster k the reaction rate coefficient for switching

ions formed through switching reaction• between reactions between NO• cluster ions and sulfuric

H SO_ and NO -cluster ions. Subsequent laboratory acid, [n•] the sum o•the_number densities of all

s• 4 udles (Viggiano 3 et al., 1980) revealed the rate HSO A cluster ions •nd [nN] the total density of constants of these reactions and made it possible all-ions with a NO 3 corel' The values of k and

to deduce sulfuric acid concentrations in the are assumed to be the same for all ion-molecule stratosphere (Arnold and Fabian, 1980). The and ion-ion reactions.

technique explained briefly hereafter, has been Recently it was discovered (Arnold et al., exploited extensively by Arnold and coworkers 1982) that other sulfur bearing g•ses may contri- (Arnold et al., 1981; Viggiano and Arnold, 1981; bute to the formation of HSO. cluster ions.

Viggiano and Arnold, 1983). Therefore formula (1) should b• replaced by :

Most data published so far were obtained

during autumn. In the present paper we report k[nN] ([H2SO4] + [HSOy]) = a[n +] [ns] (2)

some recent measurements performed during summer

conditions and briefly discuss the implication on where HSO includes HSO 3 and HSO_ and any other the present ideas about aerosol formation. sulfur co•pound that can react with NO cluster

ions to form HSO• cluster ions.

Experimental and measurements The calculation of the ion abundances [•] and [n•] to be used in formula (2) is straight-

The data shown and discussed hereafter were forward. From spectra as those shown in figure 1, obtained during a flight performed with a valve the ion abundances are derived assuming that ion controlled balloon on 16 June 1982 from the CNES count rates reflect ion number densities. This launching base at Gap-Tallard (S. France). The sounds reasonable in view of the moderate balloon borne ion mass spectrometer has been resolution used, which yields small mass discri- described in detail before (Arijs et al., 1980; mination. The results are not affected by cluster

Ingels et al., 1978; Nevejans et al., 1983). break-up _n since [n•] and [n•] are the number _{• - -}

The mass range of the spectrometer is limited de sities of all xons with HSO 4 cores and NO_

to 330 amu in the high resolution mode, but use cores respectively and collisional induced dis •

of low resolutions allows detection of ions with sociation does not modify the core of the cluster

masses larger than 330 amu. ions. Due to the limited mass range of the mass

For the negative ion measurements two reso- filter (330 amu) it may be expected that some lution modes have been used. At float altitude ions will not be detected. This is especially spectra were obtained in a moderate resolution true at altitudes above 30 km where mass 391,

mode (m/•m • 17), adequate to resolve the major i.e. HSO•(H2SO4)3, represents a considerable

ions. During descent, series of multiple scans fraction of the fotal ion signal (Arijs et al., with moderate resolution were recorded alternated 1981). In order to calculate the signal due to with single scans in the total ion mode (no DC on missing ions, the spectra obtained in the total the quadrupole rods). In this way spectra were ion mode were used, assuming that all ions with

obtained from 34 km down to 25 km altitude, masses larger than 330 amu are due to HSO 4

either at float altitude or during the descending cluster ions, as confirmed by previous measure- phase of the balloon flight. ments (Arijs et al., 1981; Arnold et al., 1982).

For the ion-ion recombination coefficient Copyright 1983 by the American Geophysical Union. recent data based on in-situ measurements (Rosen

and Hofmann, 1981), laboratory work (Smith and

Paper number 3L0363. Adams, 1982) and theoretical calculations (Bates,

0094-8276/83/003L-036353.00 1982) are available. In order to compromize

329

330 Arijs et al.' Stratospheric Sulfuric Acid

500

400 aoo

100 0

.... I .... I ... I ... I ' ' '

50 100 150 200 250 300

MASS NUMBER (AMU)

Fig. la. Typical negative ion spectrum in the moderate resolution mode obtained at 32 km after summation of 6 scans of 16Os each.

from the negative ion spectra at different alti-

• tudes are shown in figure 2, together with a

300• compilation of data previously obtained by our

• group and by the group of the Max Planck Institut z of Heidelberg (MPIH hereafter) (Viggiano and O Arnold, 1983) In view of the uncertainties on k

O '

200• and •, the errors on the [H SO + HSO ] values

• are estimated to be a factor %f43. TheYdata are

• also compared with different model calculations.

z The numbered curves are H2SO 4 vapour pressure

100D

O calculations performed by us assuming that the water vapour mixing ratio in the stratosphere is

z o constant and that the aerosols consisted of 0 -- liquid droplets of a homogeneous water/sulfuric

acid mixture. Following Hamill et al. (1977) we accepted that the water vapour is in equilibrium

with the H2SO4-H20 droplets. Using the vapour

pressure dala of Gmitro and Vermeulen (1964), which seem to be only reliable for water vapour pressure calculations (Verboff and Banchero,

between these different data we have used a 1972) the weight percentage of H_SO 4 in the

parametrization of the form : aerosols at different altitudes (•r different

temperatures) was calculated. From this and the

• = 6xlO -8 (300/T)0'5+ 1.25x10 -25 [M](300/T) 4 (3) temperature, the partial vapour pressure of

sulfuric acid was derived with the formula re-

where T is temperatur•3in Kelvin, [M] total •s ported by Ayers et al. (1980). Curves 1 and 2

number density in cm and • is in cm s . represent the results of such a computation for a This parametrization gives • values, repre- 3 ppm water mixing ratio and a U.S Standard senting an average of the data in the literature Atmosphere temperature profile, for spring-fall referred to above, between 30 and 20 km altitude. and summer conditions respectively.

Above 30 km the value given by formula (3) is in As was noted by Viggiano and Arnold (1983) better agreement with the recent data of Bates there is good agreement between curve 1 and the (1982). Using the extreme values of • as reported MPIH data in the altitude region 35 to 28 km. If in the cited references (Smith and Adams, 1982; a constant water vapour mixing ratio of 1.5-ppm Bates, 1982) results in a difference of a factor is accepted, this agreement is even excellent

of 1.5 in the derived [H2SO 4 + HSO ] values. (curve 1A). This however is probably fortuituous, The total positive ion density is d•p•ced _A/• from regarding the low H20 mixing ratio (1.5 ppm),the

the continuity equation [n +] = (Q/a) , where large uncertainties on the thermochemical data the ion pair production rate Q is calculated with used to calculate the H_SO 4 vapour pressure

the parametrization of Heaps (1978). (Verhoff and Banchero, 197•; Ayers et al., 1980)

For the reaction rate coefficient k, the and the errors on the deduced H2SO 4 number densi-

ties. As can be seen the agreement between our

recently corrected values, published by Viggian8

et al. 1(1982) were used. A value of 1.1 x 10-- data and curve 2 is rather poor. In order to cm-m-m-m• s - was reported for the reaction of H2SO 4 investigate this, we have plotted the temperature wi•h NO;(HNOq)• and since the latter was the most versus time as measured by a tiny bead thermistor

abundant io• •ith NO• core in our spectra, only ₃ during the ascent period of the flight. This this value will be applied in formula (2). measurement took place at night so that errors

The (H2SO 4 + HSO .) number densities obtained introduced by solar radiation are insignificant.

z 300 o

z

o 200

:D 100

50 100 150 200 250 300

300•

o

z o 0

35 .

5 25•-- ///•.-•'• ,'," .... 3UN81 -

< P '//•"•'"- "• 'T4'- ' ß ., 5EP81

Z?o ,o ,o ,o ,o ,o

H2• • ß H• NUMBER DENSITY (cm -3)

MASS NUMBER(AMU) Fig. 2. [H2SO 4 + HSO_] obtained in-situ by

Fig. lb. Moderate resolution negative ion different experiment• compared to vapour pressure

spectrum obtained at 28 km after summation of 5 calculation and recent models. Curve T refers to

scans. the work of Turco et al. (1981), H are model

Both spectra were smoothed as described before calculations of Hamill et al. (1982), R.A. repre-

(Arijs eta!., 1982). senting the radical agglomeration case.

Arijs et al.: Stratospheric Sulfuric Acid 331

The result is represented by curve 1 in figure 3.

For comparison U.S. Standard Atmosphere tempera-

tures at the same altitudes are represented by 240 curve 2 and 3 for summer and spring-fall condi-

tions respectivelyß Strong fluctuations probably • related to balloon motions, occur on the observed w temperature, especially at ceiling altitudeß •

Nevertheless it can be concluded that the • 230

measured temperature is closer to the spring-fall • profile after 0 30 UT, which corresponds to about ß w •

30 km. For the derivation of the H2SO 4 vapour • pressure we have therefore adopted two tentative • temperature profiles, which fit better to the 220 measured one. Using a simple linear relation

T = 195.2 + 1.15 z, where z is the altitude in km, results in curve 3A of figure 2. Taking the same relationship for z • 28 and T = 185 + 1.5 z for higher altitudes, results in curve 3B.

The agreement between experimental data and

, I , J , I , I

24.00 00.30 01.00 01.30

UNIVERSAL TIME (hours.minutes)

Fig. 3. Temperature versus time during ascent in the balloon flight of 16 June 1982.

the H2SO 4 vapour pressures now obtained is within Curve 1 : temperature as measured; 2: U.S.

the experimental errors down to about 28 km. Standard Atmosphere temperature for summer con- According to model calculations (Turco et al., ditions, derived from the conversion of pressure 1979) the concentration of H_SO. is much larger measurement to altitude; 3: same for spring-fall than that of HSO 3 above 28 Mm. It is therefore

reasonable to conclude that above this altitude

the major compound detected by the present

technique is H2SO 4. The agreement of our data 1981 data are somewhat larger, due to the lower

with our calculated values of sulfuric acid resolution used (Arijs et al., 1982).

vapour therefore suggests that the H SO_ number It is clear that more measurements at dif-

density above 28 km is mainly con{rolled by ferent stratospheric temperatures are required to

evaporation from aerosols. This confirms the elucidate the problem.

present ideas about aerosol formation and the

measurement of Viggiano and Arnold (1983) who Conclusive remarks

drew the same conclusion.

Below 28 km the data values are much higher A comparison of sulfuric acid vapour pressure

than the vapour pressure values, indicating calculations with the data set on (H. SO 4 + HSO ) supersaturation leading to aerosol growth. concentrations between 28 and 35 •m altitude

It is tempting to compare the data to the seems to support the present ideas about the

models of Turco and Hamill shown in figure 2. interaction of H2SO. with aerosols.

This however is poemature since the measured It should be no%ed however that this inter- concentrations are referring to H2S74 plus all pretation is relying on information, still suffe-

other

sulfur compounds which can gzve rise to ring from some inaccuracies due to different HSO_ ions _{4 '} These include HSO_, the concentration _• experimental factors which need further in-

of which can exceed the H_SO 4 concentration below vestigations. Among those we cite the errors on 28 km (Turco et al., 198•. Therefore a compari- the sulfuric acid vapour pressure calculations

son should be made with the total concentration due to the uncertain values of the thermodynamic

of all sulfur compounds resulting in H•O 4 quantities involved (Verhoff and Banchero, 1972).

clusters. Unfortunately no recent models gzvzng As noticed from figure 2 small changes in this quantity as a function of altitude and temperature and water mixing ratios have a con- temperature are available yet. siderable effect on the sulfuric acid partial

It is evident however from figure 2 that pressure. A precise measurement of both water (H_SO. + HSO ) concentrations are found which are mixing ratio and temperature is therefore needed.

considerably yhigher than in the experiments of Also the influence of other substances, such as Viggiano and Arnold (1983). At present however it impurities or possibly HNO•dissolved in aerosols

is difficult to conclude whether this is due to (Kiang and Hamill, 1974) s•hould be investigated.

higher stratospheric temperatures, or to recent Below 28 km the interpretation of the data is volcanic eruptions (E% Chich•n, April 1982). hampered by the fact that the present technique The latter might have increased considerably the cannot distinguish between H2SO. and other sulfur

amount of SO_ in the stratosphere and consequent- compounds which may lead to H•O i cluster ions.

ly the resulting formation of H2SO 4 and HSO . It Laboratory measurements of the ion-molecule

should be noted that all our data were obtained reactions of •SO3, HSO 5 and other sulfur com-

during descent, so that contamination effects pounds with NO_ cluster ₃ ions may help to assess cannot be responsible for our higher values. their role. Furthermore model calculations of Furthermore this difference is not due to the use these sulfur compounds would be very helpful.

of a value of the ion-ion recombination coeffi-

cient • and the ion production rate Q, being We would like to express our gratitude to our different from those used by the referred technical staff (particularly E. Van Ransbeeck,

authors. On the contrary even higher(H2SO 4 + HSOy) M. Neven, F. Cooremans, M. Mathijs and G. Van den

concentrations are found with their • and Q Wijngaert) and to the CNES personnel who contri- values and our ion abundances. buted greatly to the realization of the experi-

It should also be noted that the errors on our ments. We also thank Dr. A. Viggiano and Dr. F.

332 Arijs et al.: Stratospheric Sulfuric Acid

Arnold for contributing data, prior to publica- tion. The project was partly financied by the NFWO (Belgian National Science Foundation) under

contract hr. 2.0009.79.

References

ARIJS, E., D. NEVEJANS and J. INGELS, Unambiguous determination of major stratospheric positive ions, Nature, 288, 684-686, 1980.

ARIJS, E., D. NEVEJANS and J. INGELS, Strato-

able opening device for a balloon borne mass spectrometer, Rev. Sci. Instr., 49, 782-784,

1978.

KIANG, C.S. and P. HAMILL, H2SO4-HNO3-H_O ternary system in the stratosphere, Nature,•50, 401-

402, 1974.

NEVEJANS, D., P. FREDERICK and E. ARIJS, Micro- processor based data acquisition and control system for a balloon borne quadrupole mass spectrometer, Bull. Acad. R. Belt, C1. Sci., in press, 1983.

spheric positive ion composition measurements, ROSEN, J.M. and D.J. HOFHANN, Balloon borne mea- ion abundances and related gas detection, J. surements of electrical conductivity, mobility Arm. Terr. Phys., 44, 43-53, 1982. and the recombination coefficient, J. Geophys.

ARIJS, E., D. NEVEJANS, P. FREDERICK and J. Res., 86, 7406-7420, 1981.

INGELS, Negative ion composition measurements SMITH, D. and N.G. ADAMS, Ionic recombination in in the stratosphere, Geophys. Res. Lett., •, the stratosphere, Geophys. Res. Lett., •,

121-124, 1981. 1085-1087, 1982.

ARNOLD, F. and R. FABIAN, First measurements of TURCO, R.P., P. HAHILL, O.B. TOON, R.C. WHITTEN gas phase sulfuric acid in the stratosphere,

Nature, 283, 55-57, 1980.

ARNOLD, F., R. FABIAN and W. JOOS, Measurements of the height variation of sulfuric acid vapor concentration in the stratosphere, Geophys.

Res. Lett., •, 293-296, 1981.

ARNOLD, F. and G. H•NSCH•N, First mass analysis of stratospheric negative ions, Nature, 275, 521-522, 1978.

and C.S. KIANG, A one dimensional model des- cribing aerosol formation and evolution in the stratosphere. I. Physical processes and mathematical analogs, J. Arm. Sci., 3__6, 699- 717, 1979.

TURCO, R.P., O.B. TOON, P. HAMILL and R.C.

WHITTEN, Effects of meteoric debris on strato- spheric aerosols and gases, J. Geophys. Res., 86, 1113-1128, 1981.

ARNOLD, F., A.A. VIGGIANO and H. SCHLAGER, Impli- TURCO, R.P., R.C. WHITTEN and O.B. TOON, Strato- cations for trace gases and aerosols of large spheric Aerosols: Observation and Theory, Rev., negative ion clusters in the stratosphere, Geophys. Space Phys., 20, 233-279, 1982.

Nature, 297, 371-376, 1982. VERHOFF, F.H. and J.T. BANCH•RO, A note on the AYERS, G.P., R.W. GILLET and J.L. GRASS, On the equilibrium partial pressures of vapors above

vapor pressure of sulfuric acid, Geophys. Res. sulfuric acid solutions, A.I.Ch.E Journal, 18,

Lett., •, 433-436, 1980. 1265-1268, 1972.

BATES, D.R., Recombination of small ions in the VIGGIANO, A.A. and F. ARNOLD, Extended sulfuric troposphere and lower stratosphere, Planet. acid vapor concentration measurements in the Space Sci., 3_•0, 1275-1282, 1982. stratosphere, Geophys. Res. Lett., 8, 583-586, GMITRO, J.I. and T. VERMEULEN, Vapor-liquid 1981.

equilibria for aqueous sulfuric acid., VIGGIANO, A.A. and F. ARNOLD, Stratospheric A.I.Ch.E. Journal, 10, 740-746, 1964. sulfuric acid vapor-New and updated measure- HAMILL, P., O.B. TOON and C.S. KIANG, Micro- ments, J. Geophys- Res., submitted, 1983.

physical processes affecting stratospheric VIGGIANO, A.A., R.A. PERRY, D.L. ALBRITTON, E.E.

aerosol _1977. particles, J. Atm. Sci., 34, 1104-1119, FERGUSON _in stratospheric and F.C. FEHSENFELD, negative ion The chemistry, role of H2SO J •

HAMILL, P., R.P. TURCO, O.B. TOON, C.S. KIANG and Geophys. Res., 85, 4551-4555, 1980.

R.C. WHITTEN, On the formation of sulfate VIGGIANO, A.A., R.A. PERRY, D.L. ALBRITTON, E.E.

aerosol particles in the stratosphere, J.

Aerosol. Sci., in press, 1982.

H•APS, M.G., Parametrization of the cosmic ray ion-pair production rate above 18 Pan, Planet.

Space Sci., 26, 513-517, 1978.

INGELS, J., E. ARIJS, D. NEVEJANS, H.J. FORTH and G. SCHAEFER, Liquid helium cryopump and reli-

FERGUSON and F.C. FEHSENFELD, Stratospheric negative ion reaction rates with H_SO. in the

stratosphere, J. Geophys. Res. , 87,2 73•0-7342,

1982.

(Received January 14, 1983;

accepted February 17, 1983.)