Solar System: Self-shielding in the solar nebula

D. Wood*, V. Greener†, D. P. Hampshire†*School of Engineering and †Department of Physics,University of Durham, Durham DH1 3LE, UKe-mail: d.p.hampshire@durham.ac.uk

Solar System

Self-shielding in thesolar nebula

Variations in the abundance of isotopesof elements in primitive meteoritescarry the record of chemical and

nuclear processes that occurred during theformation of the Solar System. Here Iexplore the possibility that photochemicalself-shielding of carbon monoxide, aprocess that is known to occur in molecularclouds, may also have been important in thesolar nebula. In the solar nebula, theprocess is based on far-ultraviolet radiationfrom the growing Sun, which is effectiveover a small distance in the inner part of thenebula. In order to acquire their observedisotope compositions, all of the solid matterin the present inner Solar System must havebeen processed through this region, andsubsequently expelled to greater distancesby an X-wind or similar mechanism1.

Self-shielding in the ultraviolet photo-dissociation of CO is thought to be responsible for large isotopic fractionationeffects that are apparent in carbon and oxygen in molecular clouds2,3. Dissociationof CO, in the wavelength range 91–110 nm,occurs almost entirely by a predissociationmechanism, in which a transition occurs toan excited state with a sufficiently long lifeto exhibit vibrational and rotational struc-ture, and hence to show narrow, isotopicallyseparated absorption bands.

For example, the absorption bandsaround 105.2 nm are separated by 45 cm11

for 12CO and 13CO (ref. 3), which is a largeseparation compared with the Dopplerwidth. Thus, when a cloud is irradiated byan ultraviolet continuum, the wavelengththat corresponds to 12CO is more rapidlyattenuated than that which corresponds tothe less abundant 13CO. As a consequence,the interior of the cloud continues to under-go photodissociation of 13CO, but not of12CO. This results in an increase in the13C atoms/12C atoms ratio of the dissociationproducts in the cloud interior. Millimetre-wavelength observations reveal the comple-mentary enhancement of 12CO/13CO in thecloud interior2. The same effect must occurfor oxygen isotopes, and even more strongly,because of the larger isotope ratios:16O/18O4500; 16O/17O42,500.

The same type of isotopic self-shieldingeffect should occur in the T-Tauri stage ofsolar evolution. The proto-Sun provides astrong source of ultraviolet radiation, and it has a gas/dust disk in which the gas predominantly consists of H2, CO and N2.Irradiation of the disk produces isotopicallyindiscriminatory dissociation of CO at theinner edge, and preferential production of17O and 18O atoms in the interior (inapproximate proportion to their overallabundances). Further chemical processingwithin the disk, to produce solid and/or liquid condensates, should preferentiallyinvolve the heavy-isotope-enriched, reactiveatoms rather than the more inert CO mol-ecules. I have thus elucidated a process forenriching the rocky components and watervapour in 17O and 18O, along a slope-1 trajectory in the three-isotope graph.

During the Sun’s T-Tauri stage, materialthat accretes along magnetic field lines fallsonto the protostar at high latitudes, raisingthe temperature locally and providing asource of intense far-ultraviolet radiation1.This radiation illuminates the outflowingdisk wind, and produces the same kind ofisotopically selective photodissociation ofCO as is seen in molecular clouds3. Thephysical scale of this effect depends on density, temperature, dust/gas ratio, and soon. Observed and computed scales for themolecular-cloud case correspond to visibleoptical depths of the order of 0.5–1.0 (ref. 4),which are primarily due to dust scattering.

The most striking implication of theself-shielding model is that the overall iso-topic composition of oxygen in the primor-dial Solar System must have resembled thatin the meteoritic calcium–aluminium-richinclusions. It has long been argued that calcium–aluminium-rich inclusions areprimary condensates from a hot solar neb-ula5. On the basis of radiometric ages, theyare the oldest known rocks that formed inthe Solar System6. It is therefore unsurpris-ing that their oxygen isotopic abundances

860 NATURE | VOL 415 | 21 FEBRUARY 2002 | www.nature.com

Physics

Oxygen drips upwardsfrom superconductors

In a famous experiment, a magnet is levitated above a superconductor thathas been cooled to 77 K in liquid nitro-

gen. Here we show that in this set-up, liquid oxygen drops form on the super-conductor, drip upwards, hit the magnetand then boil.

In this arrangement, oxygen (whichboils at 92 K) condenses out of the air into the liquid nitrogen. Liquid oxygen issufficiently paramagnetic that it can bepulled into the high-field region betweenthe magnet and the superconductor to forma bridge (Fig. 1).

When a strong neodymium–iron–boronmagnet is levitated above a YBa2Cu3O7

superconductor, the bulk and surface-screening supercurrents provide both buoy-ancy and a restoring force on the magnet.The bulk supercurrents are produced by thepenetration of the magnetic field into thesuperconductor in the form of quantizedflux lines that are pinned by defects andimpurities in the material.

Large drops of liquid oxygen can bridgethe gap between the magnet and the super-conductor. Small drops can instigate a con-tinuous process in which a new drop beginsto form on the superconductor after an oldone has been pulled across the gap onto themagnet and boiled away.

An advantageous feature of this demon-stration is that it does not require the handling of liquid oxygen. For furtherdetails and a demonstration of the super-conductivity, magnetism and thermody-namics in action, see http://www.dur.ac.uk/d.p.hampshire/drop.html.

brief communications

Figure 1 A bridge of liquid oxygen between a levitating neodymium–iron–boron magnet and a YBa2Cu3O7 superconductor. The oxygen

drop develops on the surface of the superconductor, drips upwards and boils on the magnet; as the old drop boils, a new drop forms on

the superconductor. Scale bar, 10 mm.

© 2002 Macmillan Magazines Ltd

should be the same as those in the Sun. However, all other meteoritic and plan-

etary materials analysed have 18O/16O and17O/16O ratios that are several per cent greaterthan the calcium–aluminium-rich inclusionvalues7, implying that this matter has beenenriched in the heavy isotopes (in roughlyconstant 18O/17O ratios) by passing throughthe inner part of the accretion disk. Thismatter could have been returned to the colder, planet-forming regions by an X-windprocess, involving an outflow of matter con-stituting as much as several tenths of a solarmass1. Photochemical self-shielding by O2

has been considered previously in the con-text of earlier models of the solar nebula8,9.

Photochemical self-shielding may alsobe important in the photodissociation of N2,which is isoelectronic with CO. If nitrogenanomalies were caused by isotopically selective photodissociation of N2 molecules,there must have been a chemical trap for the15N-enriched atoms; candidate traps includeH2 to form NH and more complex mol-ecules, and metal grains to form nitrides.

A direct test of the self-shielding modelis being carried out by the Genesis mission,which will return a sample of solar wind toEarth for isotope analysis.Robert N. ClaytonEnrico Fermi Institute and Departments ofChemistry and Geophysical Sciences, University ofChicago, Chicago, Illinois 60637, USAe-mail: r-clayton@uchicago.edu

1. Shu, F. H., Shang, H. & Lee, T. Science 271, 1545–1552 (1996).

2. Bally, J. & Langer, W. D. Astrophys. J. 255, 143–148 (1982).

3. Van Dishoeck, E. F. & Black, J. H. Astrophys. J. 334,

771–802 (1988).

4. Warin, S., Benayoun, J. J. & Viala, Y. P. AIP Conf. 312,

387–392 (1994).

5. Grossman, L. Geochim. Cosmochim. Acta 36, 597–619 (1972).

6. Manhès, G., Göpel, C. & Allègre, C. J. Comptes Rendus de l’ATP

Planétologie 323–327 (1988).

7. Clayton, R. N. Annu. Rev. Earth Planet. Sci. 21, 115–149 (1993).

8. Kitamura, Y. & Shimizu, M. Moon Planets 29, 199–202 (1983).

9. Navon, O. & Wasserburg, G. J. Earth Planet. Sci. Lett. 73, 1–16

(1985).

Competing financial interests: declared none.

Atmosphere science

Clean air slots amidatmospheric pollution

Layering in the Earth’s atmosphere ismost commonly seen where parts ofthe atmosphere resist the incursion of

air parcels from above and below — forexample, when there is an increase in temperature with height over a particularaltitude range. Pollutants tend to accumu-late underneath the resulting stable layers1–5,which is why visibility often increasesmarkedly above certain altitudes. Here wedescribe the occurrence of an oppositeeffect, in which stable layers generate a layer of remarkably clean air (we refer to

these layers as clean-air ‘slots’) sandwichedbetween layers of polluted air.

We have observed clean-air slots in vari-ous locations around the world, but theyare particularly well defined and prevalentin southern Africa during the dry season(August–September). This is because atthis time in this region, stable layers arecommon and pollution from biomassburning is widespread.

We were readily able to identify clean-airslots from our research aircraft by lookingtowards the horizon as our altitudechanged. Visibility was very limited in thepolluted air above and below a slot, butincreased suddenly and markedly uponentering the clean air. The clean-air slotthus appears momentarily as a thin, white,horizontal layer (Fig. 1).

In 26 flights over southern Africa duringa period of 6 weeks, we saw 12 well definedclean-air slots. These were most commonlyencountered in the morning, a fact that isprobably explained by heating of theground during the day, with a resultingincrease in convective activity which causesstable layers to dissipate. During the night,as the land cools, stable layers and clean-airslots can be re-established.

All of the clean-air slots occurred withinnarrow regions a few hundred metres thick,where the atmosphere was stable and the airwas very dry; they were generally locatedabout 2 km above sea level. The particleconcentrations contained in these slots wereonly about one-third of those measured inthe polluted air above and below them.

In southern Africa, pollution beneaththese clean-air slots is produced as a result ofwidespread biomass burning, and is aug-mented by industrial emissions. The pollut-ed air above the slot probably derives frompollutants that have been carried up by con-vective activity in regions that are not domi-nated by stable layers. Once in this higherband, the polluted air can presumably betransported horizontally over large distances.

Stable layers are common in southernAfrica during the dry season, because of theoccurrence of atmospheric subsidence6. Ifthe subsiding air originates high in the troposphere, it will generally be clean anddry. The subsiding air thus has both of the

brief communications

NATURE | VOL 415 | 21 FEBRUARY 2002 | www.nature.com 861

principal attributes of a clean-air slot. Inaddition, the dryness of the air will enhancevisibility, because the particles will bedevoid of condensed water7. The stability ofthe layer of subsiding air to intrusions ofpolluted air from both below and above willcause it to retain its pristine state. The hori-zontal transport of layers of clean marineair may also have been responsible for someof the clean-air slots that we observed.Peter V. HobbsDepartment of Atmospheric Sciences, University ofWashington, Seattle, Washington 98195, USAe-mail: phobbs@atmos.washington.edu

1. Wu, Z. et al. J. Geophys. Res. 102, 28353–28365 (1997).

2. Swap, R. J. & Tyson, P. D. S. Afr. J. Sci. 95, 63–71 (1999).

3. Newell, R. E. Nature 398, 316–319 (1999).

4. Garstang, M. et al. J. Geophys. Res. 101, 23721–23736 (1996).

5. Cho, J. Y. N. et al. Geophys. Res. Lett. 28, 3243–3246 (2001).

6. Tyson, P. D. & Preston-Whyte, R. A. The Weather and Climate of

Southern Africa (Oxford Univ. Press, Cape Town, 2000).

7. Kotchenruther, R. A. & Hobbs, P. V. J. Geophys. Res. 103,

32081–32089 (1998).

Competing financial interests: declared none.

COMMUNICATIONS ARISING

Climate change

Terrestrial export oforganic carbon

Dissolved organic matter in the oceansrepresents one of the biosphere’s prin-cipal stores of organic carbon. A large

proportion of this matter is drained fromthe continents — particularly from north-ern peatlands, which contain 20% of theglobal soil carbon1. Freeman et al.2 havesuggested that rising temperatures mayenhance this transport of dissolved organiccarbon (DOC) from peatlands to theoceans. We argue here that warming canaffect DOC export in different ways,depending on whether it is accompanied by increased or decreased precipitation. Analteration in the rate of relocation of organ-ic carbon from the continents to the oceanscannot therefore be predicted on the basisof temperature change alone.

The increase in DOC transport pro-posed by Freeman et al.2 is based on anobserved 65% increase in DOC concentra-tions in British lakes and streams during the1990s, when the mean air temperature was0.66 7C higher than during the three pre-ceding decades (there was no reportedincrease in temperature during the 1990sthat was concomitant with the increasingDOC), and on a positive relationshipbetween experimentally manipulated temp-eratures and DOC leakage from peat soil.

However, we question whether this issufficient evidence for a simple and directrelationship between increased temperatureand increased export of DOC. Althoughthere are broad patterns in DOC concentra-tion in rivers between climatic zones3, the

Figure 1 Clean-air slot near Maputo, Mozambique, photographed

on 24 August 2000.

© 2002 Macmillan Magazines Ltd