Snow lines around embedded protostars

A solution to the missing carbon problem in disks?

Posted by Sébastien Maret on April 27, 2016

Embedded protostars are very young stars that are still accreting gas and dust from their parental molecular cloud. This is the earliest phase of the formation of sun-like star. Because embedded protostars are not burning deuterium in their interior yet, they are extremely cold, from a few tens to a few hundreds Kelvins, depending on the distance to the center of the protostar. As a result, volatile elements that compose the protostellar envelope – such as carbon monoxide, water, or complex organic molecules – condense onto dust particles. The specific distance from the protostar is called the snow line. Because not all chemical species condense at the same temperature, the snowline also varies from one species to the other. For example, the carbon monoxide snowline is located much farther away that the water snowline, because carbon monoxide is more volatile than water.

Snow lines are important for the formation of planets, because they determine their chemical composition. Indeed, in the protosolar nebulae, the water snowline was located somewhere between Mars and Jupiter. This is the reason why inner solar planets are composed mostly of refractory elements, while outer solar system planets are gaseous giants. Snowlines can also influence the composition of extra-solar planet atmospheres. In addition, dust particles can grow and form planetesimals more efficiently close to snowline.

In a recent study, Sibylle Anderl, myself are our colleagues from the CALYPSO team have used the NOEMA interferometer to measure the carbon monoxide snowline around several embedded protostars. For this we have used C18O (a rare carbon monoxide isotopologue) and N2H+ (the diazenylium ion) line observations. In four protostars, we have observed an anti-correlation between these two species, with the C18O emission centered on the protostar, and surrounded by the N2H+ emission (see the figure below for an example). This anti-correlation was indeed expected, as the N2H+ ion is destroyed by reactions with carbon monoxide: within the snowline, CO reacts quickly with N2H+, and the latter is depleted from the gas. The anti-correlation has been observed previously in prestellar cores such as Barnard 68 and the HD 163296 protoplanetary disk. However, it is first time that it is searched for in a large sample of embedded protostars.

CO snowline in NGC 1333 IRAS 4B

Figure above: NOEMA observations of the C18O 1-0 (left panel) and N2H+ 1-0 lines (right panel) in the embedded protostar NGC 1333 IRAS 4B. Note the anti-correlation between these two species. The white circles show the position of the carbon monoxide snowline. It is located at 460 astronomical units from the protostar, i.e. about 15 times the distance between Neptune and the Sun.

In order to constrain the exact position of the CO snowline, we have compared our observations with chemistry model coupled with a radiative transfer model. Although the anti-correlation between N2H+ and C18O was well reproduced by our model, we have found that the predicted position for the CO snowline was slightly closer from the protostar than one would have expected from CO ice evaporation laboratory measurements. This suggests that the CO ices are not pure, but that some of the CO may be trapped into less volatile CO2 or H2O ices. In addition, the CO abundance within the snowline is about an order of magnitude lower than expected. Indeed, within the water snowline, one would expect most of the carbon to be locked into gaseous CO, so the CO abundance should be close to the carbon elemental solar abundance (10-4 with respect to H2). Instead, we have measured CO abundances of 10-5 in all the protostars of our sample.

Interestingly, a comparably low CO abundance has been measured in the TW Hya disk. The authors of this study suggested that it may be the result of active carbon chemistry in the disk, with CO being rapidly converted into carbon chains, or perhaps CO2. Our results suggest an alternative scenario: the low CO abundance was simply inherited from the protostellar phase. Although this scenario is appealing, there is still some uncertainty on the CO abundance measurement: the H2 density – which we use to normalize our abundance – is somewhat uncertain in the innermost region of the protostar. In order to better constrain it, we have just submitted an ALMA cycle 4 proposal to observe high density line tracers at high angular resolution. In addition, we will observe the C18O at higher angular resolution in order to detect a second CO snowline closer to the protostar, as expected if some of the CO is indeed trapped into more refractory ices.

The paper “Probing the CO and methanol snow lines in young protostars – Results from the CALYPSO IRAM-PdBI survey” by Anderl et al. is available on arXiv.