1992 Annual CEDAR Meeting

Metallic Layers in the Mesosphere

John M.C. Plane

School of Environmental Sciences

University of East Anglia

Norwich

United Kingdom

This work has been supported by NSF Atmospheric Sciences (Aeronomy)

and the SERC of the United Kingdom

3 a o P w o 13 P O a O C in n tr

o 0-

»-•

•

1-t

fD to

Per

seid

met

eor

ions

inat

mos

pher

e

t

G

} I f it r

y JSr r

2 p̂f

I' i

XI

o r-^

o'

o -♦<

3 (t

r*+

-

o o 5* o =r

-1 o D Ql

• (6

Rat

ioof

free

met

al/N

oin

nn

eso

sph

ere

f 1 Metol precursor4

in tantalum Doat

Thermocouple

Photomultiplier

1^1 Quortz end—window^—^*~and interference filter

in

Dye loser

Suprosil window

set ot Brewster's ongle

CENTRAL

CHAMBERPUMP

Suprcsil window->>y—^

Excimer loser

Thermocouple

HeaterControl

ThermocoupleRead-out

HV

Power

Supply

Piooam

meter

Excimer

Laser

Li

t-e-

Beam—steerer

Iris

Lt

Filter

/ Beam-7 steerer

Beamsteerer

Beamexpander

Pump

HV

Power

Svpply

Photodiode

..J.

Boxcar

Averaging System

o

#

cr^

rn

Po

tv

Hj

Cl p 1J

u-1

prt

f

^1-

rx

}i-

r T

•s

?s-

s>

I

1 a_

o P

rrr>

Oo

•-

+O

O(>

>g

..o

I p Od 3 <

»

I—1

.0"S

>5r ?

>V

*"

LIF

sign

al(a

rb.

unit

s)

LIF

aign

al(o

rb.

unit

s)

I- t?

S o s: (A 0

t:

^ to r

O r

lo

IOT

CM

IJ)DU

So

ECO

Eo

1.0E-29::

1.0E-30-:

l.OE-31

150

Li + O2 + N2

No + O2 + N2

K + O2 + N2

2000

V)

i?3O

o

Eto

Eo

T/K

400 300 250

1.0E-9"__^ T-t T• I •

i

200

• Kolb et al.

• Ager and Howard

T

Na + O3 (this work)

NaO + O3 (this work)

1.0E-10--

5.0E-112.5E-3

i

Ager and Howard

—I3.0E-3 3.5E-3

44.0E-3

1 /T(K-1)

+

4.5E-3 5.0E-3

Co'6lirtg colls and

insulation \

Waeeif-tftJdled metal

vdpeut injector

Carrier gasinlet

No- + Oj^ ♦ K)j —• •*•

NaO^ ♦ O KkOf O.

Pressure

Photomultipliertube \

VSfltnplingorifice

QutdrupoleiMaat

Photoionizatlonport

Pre filter

jLonsource

hil86 eHUKtlTi^ehflflHeltfWh

Turbo molecular

pump housing

Reactant gas Qinlets

Ion transfer' optics

Mechanicalbooster pump

CU ctwMttj-k,. [n«] |[0»l[W|i"l(r'ja-M)M

=: (i t t 0-4)^10""^300 K.

Neutral and lonii

Rg^ctiQjo

Neutral Chemistry with Oxygen only

* 1. Na+ O3 --> NaO + O2

* 2. NaO + O --> Na(2p,2S) + O^

* 3.

4a.

4: 41b.

* 5.

6.

* 7.

Na +Oj + M --> NaOj +M

NaO + Oj -> Na+202

NaO + O3 ~> NaOj + Oj

NaOj + O --> NaO +Oj

NaO + Oj+ Nj -> NaOj + Nj

Na03 +0 --> NaO +Cy,

Neutral Qtemdstry wiitli H, O and COj

* 8. NaO + Hp --> NaOH + OH

* 9a. NaO + H^ --> NaOH + H

* 9b. NaO + Hj -> Na + H^O

* 10. NaO + H --> Na + OH

ixtfe CO'XSt'ftAt

in the Mesosphere

Rate Coefficient

i.5 X10''exp(-220/tenip)

1.5 X10-'exp(-383/temp)(branchingratio^ = 0.2)

4.7 X10-»(temp/200)-'

3.2 X10-'° exp(-550/temp)

1.5 X10'' exp(-^36/tenip)

1.0 X10-'eKp(-1300/temp)

5.3xl0^(200/temp)

3 XlO"'** sqr(temp/290)

8 X10 '̂° exp(-374/temp)

l.'l X10-'exp(-1100/temp)

1.1 X10-'exp(-1400/temp)

3 X10-'° exp(-668/temp)

o 11. NaOj +H --> NaOH +0 3x10"'" exp(-2000/temp)

12. NaOH + H -> Na + HjO 5 X10"'° exp(-1000/temp)

* 13. NaO + COj +Nj --> NaCOj +Nj 1.3x 10-"(200/temp)

o 14. NaCOj + O -> NaOj +CO, 1 X10-'exp(-1400/teinp)

o 15. NaCOj +H --> NaOH +CO^ 1 X10-9exp(-1400/temp)

*- 16. NaOH +CO2 +Nj -> NaHCOj +N^ 1.9 X10"^* (200/temp)

17. NaHCOj +H --> Na +HjO +CO, 1 X10''exp(-1800/teinp)

Ionic Chemistry

* 18. Na + Oj* -> Na+ +02 1.4 X 10"'

* 19. Na + NO* --> Na* + NO 1 X 10'

c 20. Na^ +Nj +N^ -> NaN,* +N, 2.5 X10" (teinp/200)-' ^

0 21. Na.X+ + e- --> Na + X 1 X10^ (200/tetnp)''̂

nK)tcx:lieiniical reactions

* 22. NaO^ +hC5 —> Na02 + O2 4 X 10"'

23. NaOH + hOJ --> Na + OH 1 X 10-5

c 24. NaOj + htD -> NaO +O2 1 X 10^

* 25. Na + hGJ -> Na+ 2 X 10-5

D

o

Cp.\u*WJ (^HtUcr ei/JUiuX^ 4/c

Jiv^ydf b«A h iJLAJ^Ot^/lt

Meteors

X = N2 Ablation

ondensation

Na.X

NaHCQ

C0..11

COo ,M NaOH

NaCOo

-0.2^

+0.8

l^y

-0.jZ,

Sodium Superoxide

0 Y+0.3

Sodium Ozonide

-0.74-0.79 +0.85

+0.98+0.43 +0.95

-0.32-0.72

(a) NaHC03(^A') (b) NaC03(2A')

-0.77-0.72

+0.37 +0.87

-0.43

-0.78

(c) HCO3" (U') (d) CO3 (202)

Model of Sodium in the Mesosphere

n^teoric input of Na of (1 ±0.2) x10^ atoms cm'̂ -s ' (Hughes. Gadsden)ablation profile from Hunten et al. (1980)

M 11 " ""vertical mixing

be solved as afunction of altitude

panWonlng of sodiun, is g.«r«d b, the o.ly.He seasonal ..d lafi»de dependence of ,.n,>e,a..re, p«ss»e. edd, dlffus.oncoefficient is included

seasonal de^ndenc of nUno, spedes O, Hand 0, is included

variable parameters.meteoric input flux

total Na at 65 km

/c(NaOH + H)

the photolysis cross-section of NaOHjt(NaHCO, + H)

aX

Q)

s

0)•d

105--

100--

00

105"

100--

40ON, Summer, Noon

10 100 1000

Concentration / cm""^

40ON, Summer, Midnight

10 100 1000

Concentration / cm~^

NaHCOs^

1E4

NaHCOsN "

1E4

•d^3

105"

100"

90"

85"

40®N, Summer

I r . I Mitl

Noon

Midnight

100 1000 1E4

Concentration / cm""^

6

•P

Comparison between Model and Tilgner and von Zahn (1988)70ON, Winter. Midnight

110 . . L. ^

100--

Column density = 5.0 (5.5±1.0) x lO^ cm 2Halfwidth (FWHM) = 10.5 (10.810.4) km i

I

Scale height =2.0 (2.5) km |I

Peak Na = 4680 (4940±560) cm~^~

Peak altitude =89 (B9.0±0.9)^^Scale height = -1.9 (—3.0) km

.(Scale height = -0.7 (-0.3) km

1— —I \—

2000 4000 6000

Concentration cm ^

-L

I

I

4-

8000

110-^

106

6looA

95"

-d13. %

90"

< 85-L

AH JiOU -I

75-

winter 35/86

suBMr 87

sunaer 88

OA «. y/cA~K"ury.

Sodium Density Profiles at

4 »-

winter (midnight)

- suunamer (noon)

1000 2000 3000 4000

Concentration / cm*"^

5000

lU,r

c\2

I

6o

m

aV

o

ou

(d

Temperature Dependence of Na Column Density, 40®N

7E9

6E9"

5E9-

4E9"

3E9 -

2E9-0^

1E9--

O BiUs and Gardner (1991)

# Model prediction

o o

o

170 180 190 200 210

Average Temperature (85 - 95 km) / K

220

Characteristics of Sudden Sodium Layers (SSLs)

form explosively at altitudes between 93 and 102 km, i.e. several

kilometres above the permanent global Na layer

have lifetimes of tens of minutes to a few hours, and then disappear

rapidly

the Na atom density in SSLs can be up to 40 times greater than in the

background Na layer: SSLs appear to come from a fresh source of Na

can extend over several hundred kilometres horizontally

appear most commonly between 1500 and 0000 hrs local time

occur quite frequently at low and high latitudes, but very rarely at mid-

latitudes cf. sudden Fe layers

strong correlation between sporadic-^ and SSLs

also appear when there is strong local heating (e.g. from 170 to 220 K in

under an hour) e.g. due to the passage of a tidal wave

o

2100 LST 2314 LST105

^e" 1 i—

!100 "

>— 1

95 -<^ _

^ Na90

851 I

80

1

1

30 March 1989^ : 1 1 I ^

0 3000 10000 20000 30000 40000

Density (cm"^)

K6^€ ^Gwi-dACr^ oaA Tey Lu.^

2 Chemical mechanisms

The striking correlation between SSLs and sporadic-£ suggests thatSSLs could be formed by the action of electrons or ions on a reservoir.

These charged particles could have large kinetic energies if they were causedby auroral precipitation, although this would only explain SSLs at high latitudes. Possible reservoirs include:

• neutral so<Mnin compounds involved in dissociative electron

attachment reactions (von Zahn and Murad, 1990)

e.g. NaHCOj + e" --> Na + HCO3-

Our recent ab initiocalculations (Rajasekhar and Plane, 1992) indicate

that this reaction is about 50 kJmol ' endothermic, so that NaHCOj is

stable with respect to thermal electrons. Reaction with atomic O mayalso be important; we have recently shown that the reactionNaO, +O --> NaO +Oj, k(300 K) = 1.6 x 10"" cm'molecule 's '

is fast (Helmer and Plane, 1992), so that a sudden increase in [O] willproduce Nafrom NaOj.

• Na* ions; these recombine with electrons extremely slowly, so the rate

of formation of molecular Na ions is important

e.g. Na* + X + N2 --> Na.X-^ + N^

Na-' + Oj --> Na0* + 02

The formation of the molecular Na ions would then be followed by

dissociative recombination with electrons;

e.g. Na.X'" + f --> Na + X

The rates for the clustering reactions are not known. However, this

mechanism may work if the rate of re-ionisation of Na by the reactions

e.g. Na + NO^, O2* --> Na-^ + NO.Oj

was slowed down by the removal of and O2* by the high electron

density in a sporadic-^ layer.

Also, if a large concentration of negative ions could be produced

(Swider, 1992):

e.g. e" (fast, 6eV) + O, -> O" + O

then O' + Na'^ -> O + Na

Unfortunately, O" also reacts very rapidly with O;

0 +0 -> 0, + e"

so that the concentration of negative ions will be limited.

Na cluster ions on the surface of dust particles. These are formed

from the recondensation of meteoric silicates, and are thought to be

about 10 nm in diameter:

(Na,Y,r (aos, +charged particle -> (Na^Yy)"^(Na^Y^r (g) +e- --> x.Na +y.Y

Thus, several Na atoms are formed for every electron that is consumed,

in accord with very recent observations (Gardner and coworkers, 1992).

Mesospheric Chemistries of Other Meteoric Metals

Potassium

• seasonal column abundance appears to be constant cf. sodium

• difficult to explain since the alkalis should all have very similarchemistries

• only the reaction K+Oj +Nj has been measured -twice as fast asNa reaction

Calcium

depleted relative to sodium by about 120, although the Ca/Na ratioin meteorites is about unity

the reaction Ca +O, + has apositive temperature dependence atlow temperatures -tnis could explain the summertime enhancementof Ca to Na

the reaction Ca +O, --> CaO +O, is very fast - perhaps furtherreactions such as

CaO + CO2 + N2 --> CaCO, + Nj

CaO + HjO + N2 --> Ca(0H)2 +

remove Ca permanently

Metal + O2 + N2

1.OE-29 V

m 1.0E-3p^|

0/ 1.0E-31

1.0E-32

1 .OE-33 - -

1 .OE-34

200 1000

/K

500.0

o

o 400.0

0)

C 300.0

X)c

^ 200.0o

D

•\ 100,0D

0.0

X 19820 1983• 1984• 1985

X

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

< 80

70-

Ca X 120

1000 2000 3000Number Density (cm-3)

4000

Magnesium

cannot be seen spectroscopically from the ground because of theHartley band of ozone

the reaction Mg + Oj + Nj ~> MgOj + Nj is extremely slow

the reaction Mg + O3 remains to be measured; if MgO formation israpid, then formation of MgCOj and Mg(0H)2 may remove Mg

the ionic chemistry of Ca and Mg is more complex than Na, becausethe Group 2 metals can form stable oxide cations:

Mg+ + 02 + N2 --> MgOj"^

Mg^ + Oj --> Mg0+ + 02

However,Mg02'̂ + 0 -> Mg0"^ + 02

MgO'̂ + O -> Mg+ + 02

compete very effectively with dissociative recombination withelectrons.

Iron

• Fe is depleted relative to Na by factors ranging from 3 to 20; the Felayer is more variableand is a few kilometres lower

• recent lidar observations (Gardner and coworkers) indicates largeenhancements at the equinoxes (comprehensive winter observationsyet to be reported?)

• Fe, like Mg and Ca, has a closed electronic s shell. Thus, it reactsvery slowly with O2. The O3 reaction is unknown at present.