Molecules traced in absorption

Molecules traced in absorption

IRAM Summer School Lecture 4Françoise COMBES

2

Role of molecular absorption

Sensitive probe of the ISM or IGM, especially for remoteobjects, when emission is affected by dilutionBenefit of the pencil QSO beam

The densest end of the absorbing systems, power-lawN(NH), optical Ly- forest, then DLA, then molecular systems

Very rare systemsLess than 10 today, but will be more numerouswith ALMA

3

Advantages of the absorption Absorption measures are very useful, in particular in the Galaxy, where both emission and absorption can be detected along the same line of sight

Obtention of the physical conditions, T, NSpatial resolution with absorption (QSO size)

However, there is a bias towards cold gas, for absorption

In the Rayleigh-Jeans domainTA* = (Tex -Tbg) (1 - e-τ)

Emission when Tex > Tbg

4

For the atomic line HI at 21cm for instancelarge influence of stimulated emission ("negative absorption")since the ΔT between the two levels F=1, 0 is only ~0.7 K

ƒτdv ~ N/T

In emission, N(cm-2) ~ƒTexτdv ~ƒTadv independent of temperature

While the optical depth of the absorption signal is in 1/T

Experiences ON-source, and OFF-source Ta(ON), Ta(OFF)gives Tex or Tsp

In the millimeter, CO rotation for instance at 2.6mmthere exists the whole rotational ladder

5

Emission: depends on temperature, since Nu/Ntot = gu/Z e-Eu/kT

Nu(cm-2) ~ƒTexτdv ~ƒTadv if τ << 1, and Ntot ~T Nu eEu/kT

Absorption: ƒτdv ~ N/T (1- e-hν/kT)

strongly weighted by the temperature Tex

Since collisional excitation requires 4 104 cm-3 for CO, and 1.6 107 cm-3 for HCN

In hot (kinetic temperature) and diffuse media, the excitation temperature will be very low, ~ 2.76 K

Absorption is weighted by the diffuse medium

6

Molecular absorption in the Galaxy

More difficult to observe, since continuum sources are weaker(S ~ν-α) and smaller. Requires interferometry to resolve and distinguish from emission

explains the work is recent (the last decade)Marscher et al (1991) in front of BlLac

Small filling factor in surface, even of the diffuse CO medium9 3C sources/100 have CO emission (Liszt & Wilson 93)Among them, 60% show absorption

Extinction of only Av~1 mag, but already very abundant chemistry (Lucas & Liszt, 1994)!

7

Liszt & Lucas 2001

Survey of 30 l.o.s. (Lucas & Liszt 96)HCO+ 30% as often as HI absmore frequent than CO

8

13CO, CN, HCO+, HCN, HNC, C2H, N2H+ (Lucas & Liszt 94-98)

with line ratios quite variable from one l.o.s. to the other

Big surprise, the strength of HCO+ absorption, in these diffuse media

-- higher critical density, so HCO+ is "cold"-- chemistry to be revised in diffuse medium!

Some lines are very optically thick (13CO is detected)others τ << 1 (hyperfine lines of HCN, in the ratio 5:3:1 expected)

ΔV = 0.5 - 1km/s

Abundances of CO versus HCO+ variable by 20!Bistability? Chaos ? (Le Bourlot et al 1993)

9

Absorptions sometimes variable over a year time-scalepresence of clumpy material, of sizes 10-100 AU in front on thecontinuum sourceAlso spatial fluctuations in the chemistry (Liszt & Lucas 2000)CO can form rapidly from HCO+ in diffuse clouds

H2 can form at relatively low densitywhenever H2 is there HCO+/H2 = 2 10-9

and then CO forms by recombination of HCO+ (CO turn on)

HCO+ is linearly correlated with OHX(HCO+) = 0.03-0.05 X(OH) even at low column density

CO forms later (when C+ is recombined)

Diffuse clouds have chemical abundances of dark clouds!

10

Liszt & Lucas, 96, 2000

OH and HCO+ tightly correlatedat low column density,contrary to CO

11

Computed temperatures for gas spheresof N(H) = 5 1020cm-2, according to density

CO and C+ column density for the same models (Liszt & Lucas 2000)

H2 formation can occur at low density,while HCO+ is present, but not COthe C is still largely under C+

12

N(CO) increases abruptlywhen N(HCO+) = 1-2 1012 cm-2

slope of the power-law: 1.5

CO and H2 column density from theUV (Federman et al 95)

Slope of the power-law is 2.02

13

ΔV(HCO+) = 15% higher ΔV(CO)

Surprisingly large 13CO abundanceFractionation, much more efficientthan selective photodissociation

12CO + 13C+ --> 13CO +12C+

14

Extragalactic molecular absorptions When the line of sight of a radio-loud QSO crosses a galaxy, and also a molecular cloud (quite rare) absorption in the mm, cm

Prolongation to the high column density of the Lyα absorbers, in particular DLA N(NH) power law

•Lyα forest N ~ 1013 cm-2 (intergalactic filaments)

•HI-21cm 1020 cm-2 (Damped Lyα systems) Outer parts of galaxies

•CO, HCO+.. 1020-24 cm-2 (the center of galaxies)

The number (N) decreases as a power law

15

Comparison with emission

The absorption technique is much more sensitive than emission

At high redshift for instance, the detection limit is 1010MoWhile the absorption limit does not depend on redshiftAs soon as the QSO source behind is detected, the absorptionlimit is in optical depth τThe source is quasi ponctual at mm, up to 1012K

Galactic versus extragalactic:for MW absorption studies, interferometer is required, since absorption is generally buried among strong emission of local molecular clouds

The nearest absorption is Centaurus A, where both are of the same order

16

Centaurus A

Eckart et al 90

In CO line emission and absorption aredetected

Many other lines are detected in absorptiononly (Wiklind & Combes 1997)

17

No temporal variation (Wiklind & Combes 1997) Constraints can be put on the source size, of > 500 AU

Low density gas, low excitation and low Tkin

optically thin lines

Wide absorption in HCO+, could correspond to a nuclear disk

18

Higher redshift absorptions

First high-z absorption towards the BLLac object PKS1413+135(Wiklind & Combes 1994), after many unfruitful searches towards DLAs

Since then, 4-5 systems are known, but remain rare

Half of them are gravitationally lensed objectsPKS1830-211 and B0218+357

The absorbing molecular clouds are in the lensing galaxy a way to find molecules in normal galaxies at high z

Redshifts range up to z~1 (the QSO at z~2), difficult to find higherredshifts QSO, that are strong enough in the mm (steep spectrum)

19

In absorption, detected masses can be only 1 MoLarge variety of line widths, optical depths, sometimesseveral lines are detected along the same l.o.s.

20

Selection of candidates:

-- Strong mm source (0.15 Jy at 3mm) only 100-200

-- already an absorption detected in HI-21cm, or DLAs, or MgII or CaII

-- absence of previous absorption, but known gravitational lens (VLBI) (Webster et al 95, Stickel & Kuhr 93)

-- same as above, without any known redshift: the case of PKS1830-211The redshift was discovered in the mmsweeping of the band (14 GHz = 14 tuning, and already 2 lines)

--sources where the redshift searched is that of the QSO

Mostly negative results!

21

PKS1413+135 z=0.247

McHardy et al 94

Very narrow absorption < 1km/s (2 comp)BlLac, very variable, also in radio

optically thin, N(H2) > 1022 cm-2, Av > 30 mag

22

Temporal variability, and small-scale structureThe opacity ratio between the two components has varied by 2.3over 2 years

Variations due to thel.o.s. change due to the variability of the continuum source

Superluminic sourceCore unresolved 2.3masor 7pc, might be 10μas = 0.03pc

250km/s = 50AU/yr insufficient (100yrs)> 25 000 km/s required must come from the core

23

Compatible with either a multi-component model with similar filling factors or with dense clumps embedded in a diffuse medium

The diffuse component accounts for most of the absorption, whilethe clumps comprise most of the mass

Because of the very narrow velocity widththe cloud along the l.o.s. must be quite small1pc according to size/line-width relationn(H2) ~104 cm-3

variability seen in the CO, not in HCO+(more optically thick)HCO+ more from the diffuse component

24

B3 1504+377 z=0.6727 different molecular linesLarge separation 330km/snuclear ring + spiral armabsorption hosted by the sourceHNC/HCN Tkin = Tex

HCO+ enhanced by 10-100diffuse + clumps

25

B0218+357 z=0.685Gravitational lens (two images A and B)The largest column density 1024cm-2

Two images separation 335mas (1.8kpc)

All three CO isotopesare optically thick

This was an excellent oppotunity to searchfor O2 without atmospheric absorptionLines at 368 and 424 GHzO2/CO < 2 10-3 (Combes et al 97)most of O in OI??

26

H2O detection at 557 GHz, very large τ =40 000

LiH tentative detection

H2O ubiquitous and coldT=10-15 KH2O/H2=10-5

HD and LiH cooling linesLiH 21K above ground

444GHz line of LiH, optically thinvery narrow, LiH/H2 ~3 10-12

27

Two images A and B, separated by HST(335mas)

VLBA measurements (Patnaik et al 93, 95)

The two images separated in A1, A2,B1, B2 (lens potential non spherical)

28

PKS1830-211 z=0.88582

2 images, + Einstein ring

But 2 absorbing systems,one at z=0.19 seen in HI

Frye et al 97

29

Two components, covering each one image of the sourceas confirmed by PdB (Wiklind & Combes 1998)

Slight temporal variability

30

Monitoring, measure of H0

The single dish (without resolving the 2 images) can followthe intensity of the two, since they are absorbing at two V

Monitoring during 3 years (1h per week)==> delay of 24+5 days, H0 = 69 +12 km/s/Mpc

31

Evolution of chemical conditions?Various line ratios have been obtained in the many absorptions at all z

There does not seem to be variations versus z=0 (open circles)but large scatter, even at z=0 (Lucas & Liszt 94, 06)

32

Measure of Tmb (z)Low excitation (diffuse gas) Tex ~TmbThe case for PKS1830-211

Several transitions give the sameresult (slightly lower, due to a microlens)

From UV H2 linesSrianand et al 2000

33

Variation of Constants

The idea of constants variation dates back to Dirac (1937)Jordan (1937, 39); forces other than gravity geological problems with variation of GSolved if mass p/e varies, or charge e (Gamow 1967) Landau (1955) relation with the re-normalisation in QED

Theories motivating this variationKaluza-Klein (Kaluza 1919, Klein 1926) 5th dimensionextra-dimensions (1980', 90') quantum gravityunification of forces with gravitysuperstrings (10 dimensions)

34

Methods In laboratory, atomic clocks d/dt/ < 3.7 10-14 /yr Natural nuclear reactor OKLO (Gabon)Natural fission, 1.8 Gyr ago, 150Sm (Damour & Dyson 1996) / < 1.2 10-7

Fujii et al (2000) / < 0.04 10-7

But: very low redshift (z ~0.1) No test of spatial variations + Method of radioactive isotopes in meteorites (age of solarsystem, comparable precision, 4Gyr) Olive et al (2002) Absorption lines in front of QuasarsOptical: Alcali Doublet (AD) or Many Multiplet (MM)

CMB: COBE, WMAP, Planck will give 0.1% in

35

MMWeb et al 2002

36

Variation of in radio absorptionIn the case of PKS1413, a resolution of 40m/s is required to well resolve the lines!

Decomposition of spectra in several components, minimising 2

Results:y/y = (-0.16 +0.36) 10-5 pour B0218

y/y = (-0.20 +0.20) 10-5 pour PKS1413

comparable precision to the MM method, but no detection

Different redshifts ?Tests with component changes (not sensitive)

37

kinematical bias?Tests at low redshift:Absorptions of HI and HCO+ in the Galaxy in front of remotequasars (Lucas & Liszt 1998)Dispersion of only 1.2km/sCorrespondant to y/y = 0.4 10-5

This error is to be added in quadrature to the previous limits.

To be found: radio absorptions at larger redshift (between 1 and 2)

+ Problems of temporal variations of the absorption

38

Radio HI-21cm and millimetric (molecules)

39

40

X X

X = radio PKS1413 (z=0.24) and B0218 (z=0.68)

O

O= AD

41

Summary and interpretation

Full line: averageDash line: fit with = 0 fixed at z=0.

42

Local tests: open symbolsQSO absorptions: filled symbolsOlive et al (2002) meteorites in solar system

43

Future of the determination

Method AD ( 21 SiIV systems) / = (-0.5 + 1.3) 10-5

Method MM (many multiplet) / = (-0.72 + 0.18) 10-5

49 systems, towards 28 QSOs, => 128 recent

Method radio y/y = (-0.20 +0.4) 10-5 for PKS1413To find other sources

ALMA (mm interferometre, 64 antennae of 12m)

44

H2 bands in absorption, at high zUV lines at high z Foltz et al (1988) N(H2) = 1018cm-2,

Ge Bechtold (97) z=1.97N(H2) = 7 1019cm-2, T=70K n =300cm-3

total N(H) = 1020cm-2, f(H2) = 0.22 dust and strong CISrianand et al (2000), Petitjean et al (2000)

LMC

45

PKS1232+082z=2.3377

Srianand et al 00C+ and CI linesObserved together withH2

The C+ lines, for a given Tyields an upper limitn(H2) < 20cm-3

For its excitationThen H2 is not dense enoughto excite CI, and CMB isonly responsible

46

Conclusion

Absorption is a precious tool to observe cold gasdiffuse, with low excitation

Small masses are detected

Chemistry can be investigated

Gas in galaxies that are not ultra-luminous

Bias in the optical/UV towards low column density