The structure of the magnetic field in the massive star-forming region W75N

0.43 MB·English

by  G. Surcis

Astronomy&Astrophysicsmanuscriptno.15825 (cid:13)c ESO2011
January12,2011
The structure of the magnetic field in the massive star-forming
region W75N
G.Surcis⋆,1,W.H.T. Vlemmings1,S.Curiel2,B.Hutawarakorn Kramer3,4,J.M.Torrelles5,andA.P.Sarma6
1 Argelander-Institutfu¨rAstronomiederUniversita¨tBonn,AufdemHu¨gel71,D-53121Bonn,Germany
e-mail:[email protected]
2 InstitutodeAstronom´ıa(UNA),ApdoPostal70-264,Cd.Universitaria,04510-MexicoDF,Mexico
3 Max-PlanckInstitutfu¨rRadioastronomie,AufdemHu¨gel69,D-53121Bonn,Germany
4 NationalAstronomicalResearchInstituteofThailand,MinistryofScienceandTechnology,RamaVIRd.,Bangkok10400,Thailand
1 5 InstitutodeCienciasdelEspacio(CSIC)-UB/IEEC,UniversitatdeBarcelona,Mart´ıiFranque`s1,E-08028Barcelona,Spain
1 6 PhysicsDepartment,DePaulUniversity,2219N.KenmoreAve.,ByrneHall211,Chicago,IL60614,USA
0
2 Received;accepted

n
ABSTRACT
a
J
Context.Adebatedtopicinstarformationtheoryistheroleofmagneticfieldsduringtheprotostellarphaseofhigh-massstars.It
0
isstillunclearhowmagneticfieldsinfluencetheformationanddynamicsofmassivedisksandoutflows.Mostcurrentinformation
1
onmagneticfieldsclosetohigh-massprotostarscomesfrompolarizedmaseremissions,whichallowsustoinvestigatethemagnetic

fieldonsmallscalesbyusingverylong-baselineinterferometry.
]
R Aims.Themassivestar-formingregionW75Ncontainsthreeradiocontinuumsources(VLA1,VLA2,andVLA3),atthreedifferent
evolutionarystages,andassociatedmasers,whilealarge-scalemolecularbipolaroutflowisalsopresent.Veryrecently,polarization
S
observations ofthe6.7GHzmethanol masersatmilliarsecondresolutionhavebeenabletoprobethestrengthandstructureofthe
.
h magnetic fieldover more than2000 AUaround VLA1. Themagnetic fieldisparallel to theoutflow, suggesting that VLA1isits
p poweringsource.TheobservationsofH2O masersat22GHzcangivemoreinformationaboutthegasdynamicsandthemagnetic
- fieldsaroundVLA1andVLA2.
o Methods.TheNRAOVeryLongBaselineArraywasusedtomeasurethelinearpolarizationandtheZeeman-splittingofthe22GHz
r watermasersinthestar-formingregionW75N.
t
s Results.Wedetected 124water masers, 36around VLA1 and88 around VLA2 of W75N,which indicatetwo different physical
a environments around thetwosources, whereVLA1isinamoreevolvedstate.Thelinearpolarizationof themasersconfirmsthe
[ tightlyordered magneticfieldaroundVLA1, whichisaligned withthelarge-scalemolecular outflow,andalsorevealsanordered

magnetic field around VLA2, which is not parallel to the outflow. The Zeeman-splitting measured on 20 of the masers indicates
1
strongmagneticfieldsaroundbothsources(theaveragedvaluesare|B | ∼ 700mGand|B | ∼ 1700mG).Thehighvaluesof
v VLA1 VLA2
themagneticfieldstrengths,whichcomefromtheshockcompressionofthegas,areconsistentwiththemethanolandOHmagnetic
6
field strengths. Moreover, by studying the maser properties we were also able to determine that the water masers are pumped in
5
C-shocksinbothsources.
9
1 Keywords.Stars:formation-masers:water-polarization-magneticfields-ISM:individual:W75N
.
1
0
1 1. Introduction Theyobtainedtheirresultsbyusingmethanolmasersasprobes
1 of magnetic fields. In fact, the bright and narrow spectral line
v: During the formation of low-mass stars, the magnetic field is emission of masers is ideal for detecting, with polarimetric in-
i thoughttoslowthecollapse,totransfertheangularmomentum, terferometry, both Zeeman-splitting and the orientation of the
X and to power the outflow, but its role during the formation of magneticfieldonscalesfromarcsectomilliarcsec(mas).Sofar,
r high-mass stars is still under debate (e.g., McKee & Ostriker themainmaserspeciesusedforthispurposeareH O,OH,and
a 2007,Girartetal.2009).Severalquestionsremainunanswered CH OH masers.Sincetheirmasingconditionsared2ifferentthey
3
becausemassivestar-formingregionsarerareanddistantandof- are located in distinct zones of massive star-forming regions.
tenconsistofalargenumberofprotostars,consequentlyitisdif- The H O masers are detected in the denser zones with hydro-
2
ficulttoobservemagneticfieldsduringtheirprotostellarphase. gennumberdensitiesn betweenapproximately108 and1010
Recentlysomeprogresshasbeenmade.Vlemmingsetal.(2010) cm−3 (Elitzur et al. 198H92), showing a velocity width of about
andSurcisetal.(2009,hereafterS09)haveshownthatthemag- 1 kms−1 and a high brightness temperature T > 109K (e.g.,
b
netic field is orthogonal to large rotating disks and parallel to Reid & Moran1981). Thefirst verylongbaselineinterferome-
molecularbipolaroutflowsevenduringthehigh-massprotostel- try(VLBI)observationsofthelinearandcircularpolarizationof
lar phase and not only during the formation of low-mass stars watermaseremissionsweremadebyLeppa¨nenetal.(1998)and
(e.g., Matsumoto & Tomisaka 2004, McKee & Ostriker 2007). Sarma et al. (2001), respectively.AfterwardsmoreH O maser
2
polarization observations were carried out by various authors,
⋆ MemberoftheInternationalMaxPlanckResearchSchool(IMPRS) which confirmed the importance of this kind of observation in
for Astronomy and Astrophysics at the Universities of Bonn and studyingtheroleofthemagneticfieldinmassivestarformation
Cologne.
2 Surcisetal.:W75N:watermaserspolarization.
(e.g.,Imaietal.2003,Vlemmingsetal.2006).Herewepresent outflowisrobust.
VeryLongBaselineArray(VLBA)linearandcircularpolariza- Sincethewatermasers,unliketheCH OH andOHmasers,
3
tionobservationsofwatermasersinthehigh-massstar-forming arefoundindenserzonesofthestar-formingregions,itiscrucial
regionW75N. tostudytheirpolarizationemissiononasmallscale.Infact,be-
W75N is an active high-mass star-forming region in the causetheyariseclosetobothradiosourcestheycangiveusnew
molecularcomplexDR21–W75(Dickeletal.1978,Persietal. informationontheroleofthemagneticfieldduringthemassive
2006) at a distance of 2kpc. At the resolution of ∼ 1′.′5 three starformation,inparticularattwodifferentevolutionarystages.
Hii regionswere identified (Haschick et al. 1981): W75N(A),
W75N(B), and W75N(C). At 0′.′5 resolution, Hunter et al.
(1994)detectedthreesubregionsin W75N(B):Ba, Bb andBc. 2. Observationsanddatareduction
Athigherangularresolution(∼ 0′.′1),Torrellesetal.(1997)re-
We observed the massive star-forming region W75N(B) in the
solvedthesubregionsBaandBbfurther(renamingthemVLA1
6 -5 transitionofH O (restfrequency:22.23508GHz)with
andVLA3),andimagedathirdweakerandmorecompactHii 16 23 2
the NRAO1 VLBA on November 21st 2005. The observations
region between them, which they named VLA2. A large-scale
weremadeinfullpolarizationspectralmodeusing4overlapped
high-velocity outflow, with an extension greater than 3 pc and
basebandfiltersof1MHzinordertocoveratotalvelocityrange
a total molecular mass greater than 255 M , was also detected
⊙ of ≈ 50 kms−1. Two correlations were performed. One with
from W75N(B) (e.g., Shepherd et al. 2003). Shepherd et al.
128 channels in order to generate all 4 polarization combina-
(2003) proposed a multi-outflow scenario where VLA2 may
tions (RR, LL, RL, LR) with a spectral resolution of 7.8 kHz
drive the large-scale outflow, and VLA1 and VLA3 are the
(0.1 kms−1). The other one with high spectral resolution (512
centers of two other small flows. So far, it has been impossi-
channels;1.96kHz=0.027kms−1),whichonlycontainsthecir-
bletodeterminethemainpoweringsourceofthe3pcoutflow.
cular polarization combinations (LL, RR), to be able to detect
Several authors have suggested VLA1 as the powering source
Zeeman splitting of the H O maser across the entire velocity
(e.g.,Torrellesetal.1997,S09andreferencestherein). 2
range. Including the overheads, the total observation time was
Several maser species (H O, OH, and CH OH) have been
2 3 8h.
detected in W75N (e.g., Torrelleset al. 1997, Baart et al 1986,
The data were reduced using the Astronomical Image
S09),inparticulararoundthetwoHIIregionsVLA1andVLA2. Processing Software package (AIPS) following the method of
The H O maser emissions associated with VLA1 are located
2 Kemball et al. (1995). The bandpass, the delay, the phase
along a linear structure parallel to its radio jet with a proper
and the polarization calibration were performed on the cal-
motion of about 2 mas/yr, while those associated with VLA2
ibrator J2202+4216, which has been used successfully by
show a circular distribution, which is expanding with a veloc-
Vlemmings et al. (2006) for H O maser polarization observa-
ity of about 5 mas/yr (Torrelles et al., 2003, hereafter T03). 2
tions of CepheusA. The fringe-fitting and the self-calibration
OnlyoneH O maserisassociatedwithVLA3(T03).Methanol
2 wereperformedononeofthebrightestmaserfeaturesVLA1.14
masersweredetectedintwogroups(AandB/C)locatednorth-
(Table 1). All calibration steps were initially performed on the
west and southeastof VLA1, respectively.GroupA is alonga
dataset with modest spectral resolution after which the solu-
linearstructureandgroupB/Cshowanarc-likestructure(S09).
tions, with the exception of the bandpass solutions that were
NoCH OHmasersareassociatedwithVLA2(e.g.,Minieretal.
3 obtained separately, were copied and applied to the high spec-
2000,S09).However,VLA2istheplacewherethemostinten-
tral resolution dataset. Stokes I (rms = 7.3 mJybeam−1), Q
siveOHflaretookplace(Alakovetal.2005,Slyshetal.2010).
(rms = 6.5 mJybeam−1) and U (rms = 6.5 mJybeam−1) data
Other OH maser emission sites are situated on a ring structure
cubes were created using the AIPS task IMAGR (beam-size
aroundallthreeradiosources(Hutawarakornetal.2002).Based
2.0mas×0.7mas)fromthemodestspectralresolutiondataset,
onthedifferentactivityinH O andOHmasers,Torrellesetal.
2 while the I and V (rms = 10 mJybeam−1) cubes were imaged
(1997) suggest that these three sources are at different evolu-
fromthehighspectralresolutiondatasetandforthesamefields.
tionarystages,inparticularVLA1isthe“oldest”sourceofthis
TheQandUcubeswerecombinedtoproducecubesofpolarized
region and VLA3 the “youngest”one (drivinga radio Herbig-
intensity and polarizationangle. Since these observationswere
Haroobject,Carrasco-Gonza´lezetal.2010).Inthisevolutionary
obtained between two VLA polarization calibration observa-
scenario,VLA2isatanintermediatestage.
tions2,duringwhichthelinearpolarizationangleofJ2202+4216
In order to determine and investigate the magnetic field of
wasconstantat13◦.5,we wereable toestimate the polarization
W75N(B),polarizationobservationsofOHandCH OH maser
3 angles with a systemic error of no more than ∼2◦. The I and
emissions have been made (e.g., Hutawarakorn et al. 2002;
V cubes at high spectral resolution were used to determine the
Fish & Reid 2007; S09). The magnetic field strength obtained
Zeemansplitting.
fromtheOHmaserpolarizationobservationswas∼7mG(e.g.,
Hutawarakorn et al. 2002; Slysh et al. 2002). During the OH
maser flare near VLA2, Slysh & Migenes (2006) detected a
3. Analysis
strongmagneticfieldofmorethan40mGinseveralmaserspots,
whichincreasedinthenextobservationsupto70mG(Slyshet 3.1.Maseridentification
al. 2010). S09 measured a magnetic field strength of 50 mG
To identify the water maser features in W75N(B) region, we
around VLA1 by studying the circular polarized emissions of
used an identification process divided into three steps: 1) we
methanol masers. Hutawarakorn & Cohen (1996) and S09 in-
used a program called “maser finder” (Curiel, personal com-
vestigatedthelinearpolarizationofOHandCH OH masers,re-
3 munication), which is able to search for maser spots, velocity
spectively,andbothsuggestedthatthemagneticfieldisoriented
alongtheoutflow.AlthoughtheOH masersat1.6and1.7GHz
1 TheNationalRadioAstronomyObservatory(NRAO)isafacilityof
are verysusceptible to both internaland externalFaradayrota-
theNationalScienceFoundationoperatedundercooperativeagreement
tion,at6.7GHztheFaradayrotationisonlyabout5◦indicating byAssociatedUniversities,Inc.
thatthefindingofthelarge-scalemagneticfieldalignedwiththe 2 http://www.aoc.nrao.edu/astro/calib/polar
Surcisetal.:W75N:watermaserspolarization. 3
Fig.1.Resultsofthefullradiativetransferχ2-modelfits,forthemaserthatshowsthehighestlinearpolarizationfraction(VLA1.19)
andforthe maserthatshowsthehighest P aroundVLA2 (VLA2.22).Thefits yieldthe emergingmaserbrightnesstemperature
V
T ∆Ω and the intrinsicmaser linewidth∆V. Contoursindicatethe significanceintervals∆χ2=0.25,0.5,1, 2, 3,7, with the thick
b i
solidcontoursindicating1σand3σareas.
Fig.2.TotalintensityandcircularpolarizationspectrumforthethreemaserfeaturesVLA1.07,VLA1.16,andVLA2.16.Thethick
redlinesarethebest-fitmodelsofIandVemissionobtainedusingtheradiativetransfermethod(seeSect.3.2).Allmaserfeatures
werecenteredtozerovelocity.
channelbyvelocitychannel,withasignal-to-noiseratiogreater alltheidentifiedmaserspots.AfterrunningalltheAIPSscripts,
than a given value (in our case, an 8-sigmalimit was used);2) atablewithalltheimportantparameters(e.g.,spatialresolution,
wefittedtheidentifiedmaserspotsusingtheAIPStaskIMFIT; velocity,intensity)ofeachmaserspotisproduced.Wethenuse
and3)weidentifiedamaserfeature,whenthreeormoremaser theoutputtabletosearchformaserfeaturesthatfulfillthecrite-
spotscoincidedspatially(withinaboxof2by2pixels)andeach riadescribedabove.
ofthemappearedinconsecutivevelocitychannels.Table1con-
tains the positions, velocities and peak flux densities of all the
3.2.Polarizationfitting
maserfeaturesfoundintheregion.
Inthefirstpartoftheprocessandafterevaluatingtheglobal From the maser theory we know that the fractional linear po-
noiselevelofthechannelmap,thecodesearchesformaserspots larizationP oftheH O maseremissiondependsonthedegree
l 2
in the velocity map, excluding the edges of the channel map. of its saturation and the angle (θ) between the maser propaga-
When strong maser spots are found (for instance, with a peak tiondirectionandthemagneticfield(e.g.,Goldreichetal.1973,
fluxdensityhigherthan2Jybeam−1),thelocalrmsaroundthese Nedoluha&Watson1992).Eventherelationbetweenthemea-
strong maser spots (within a predefined box, for instance, 400 sured polarization angle χ and the magnetic field angle on the
by400pixels)isestimated,aswellasthermsoutsidetheboxes. sky (φ ) depends on θ, with the linear polarization vector per-
B
The code identifies a maser spot when the SNR of the candi- pendicular to the field for θ > θ ∼ 55◦ (Goldreich et al.
crit
date (usingthe localrms) isgreaterthana predefinedvalue.In 1973), where θ is the so-called Van Vleck angle. It is there-
crit
the case of W75N, we have adopted a lower limit of 8-sigma. fore important to evaluate θ for every H O maser that shows
2
Thecodeproducesatablewiththespatiallocationandintensity linear polarization emission. In order to obtain the best values
ofall the identifiedmaserspots. Thisprocedureis repeatedfor forθ,weneededtodeterminetheemergingbrightnesstempera-
eachvelocitychannelincludedintheinputfilethatisusedbythe ture(T ∆Ω) andtheintrinsicthermallinewidth(∆V),whichis
b i
code.Thecodealsoproducesascriptforeachvelocitychannel thefullwidthhalf-maximum(FWHM)oftheMaxwelliandistri-
thatcanberuninsideAIPSinordertoproduceGaussianfitsof bution of particle velocities, of the masers with high precision.
4 Surcisetal.:W75N:watermaserspolarization.
Table1.All22GHzwatermaserfeaturesinW75Nforwhichlinearand/orcircularpolarizationwasdetected.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)
Maser groupa RA Dec Peakflux VLSR ∆vL Pl χ ∆Vib Tb∆Ωb PV B|| θc
offset offset Density(I)
(mas) (mas) (Jybeam−1) (kms−1) (kms−1) (%) (◦) (kms−1) (logKsr) (x10−3) (mG) (◦)
VLA1.01 C -95.3612 -303.062 2.76±0.13 10.7 0.38 2.3±0.4 +43.8±8.3 1.1+0.3 9.5+0.7 − − 79+10
−0.3 −0.8 −14
VLA1.02 C -94.2666 -300.461 3.39±0.09 10.5 0.40 3.2±0.5 +26.4±9.4 1.1+0.3 9.6+0.6 − − 86+2
−0.2 −1.1 −13
VLA1.03 C -91.6984 -289.684 30.78±0.19 11.3 0.39 5.1±0.8 +29±1.6 0.8+0.4 10.3+0.4 7.8 106±18 74+10
−0.3 −0.3 −6
VLA1.04 C -72.0367 -217.556 9.77±0.13 10.7 0.46 3.9±0.4 −43.0±1.4 1.5+0.3 9.7+0.3 11.3 −212±51 83+5
−0.4 −0.3 −8
VLA1.05 C -71.5315 -214.531 115.41±0.12 10.6 0.44 1.7±0.2 −56.1±6.5 1.7+0.2 9.3+0.5 2.9 −54±9 74+16
−0.3 −0.1 −14
VLA1.06 C -71.3210 -213.371 76.79±0.12 10.8 0.46 1.5±0.2 −59.3±6.0 1.8+0.2 9.4+0.6 5.1 −115±18 66+11
−0.5 −0.2 −39
VLA1.07 C -70.4789 -210.289 10.16±0.12 11.1 0.45 1.4±0.6 −53.9±11.3 1.9+0.2 9.2+0.6 28.4 599±100 72+8
−0.3 −2.0 −47
VLA1.08 C -69.5948 -206.944 158.28±0.20 11.5 0.51 2.6±0.1 −73.5±5.0 1.8+0.2 9.7+0.5 2.8 65±10 69+20
−0.6 −0.1 −5
VLA1.09 C -11.7886 -121.338 6.14±0.06 10.4 0.33 1.1±0.5 −87.2±3.2 1.0+0.2 9.0+0.2 − − 86+3
−0.1 −2.0 −52
VLA1.10 C -11.4097 -237.705 2.07±0.20 11.4 0.51 2.5±0.4 −77.9±7.2 2.0+0.2 9.5+0.5 − − 86+4
−0.3 −1.5 −4
VLA1.11 C -9.7679 -119.442 0.39±0.03 10.3 0.51 2.7±0.3 −56.1±14.2 1.0+0.2 9.5+0.6 − − 83+7
−0.2 −0.5 −9
VLA1.12d B -2.7366 -15.713 1.13±0.03 12.8 0.38 19.4±0.7 −86.2±0.7 0.7+0.3 10.5+0.1 − − 90+2
−0.1 −0.5 −2
VLA1.13d B -2.6945 -14.427 0.97±0.03 12.8 0.40 12.4±2.5 −50.0±1.4 0.8+0.2 10.1+0.3 − − 90+10
−0.5 −2.5 −10
VLA1.14d B -2.6945 1.568 80.02±0.07 12.6 0.44 13.5±0.7 −16.9±0.6 0.6+0.2 10.6+0.1 − − 90+5
−0.2 −0.4 −5
VLA1.15d B -0.5894 -36.270 63.25±0.17 11.6 0.49 13.1±1.9 −90.0±3.6 0.7+0.8 10.7+0.1 − − 90+7
−0.2 −0.6 −7
VLA1.16d B -0.4210 -35.545 111.47±0.20 11.4 0.55 12.9±1.2 −66.9±0.6 0.7+0.7 10.7+0.4 11.3 −177±27 90+6
−0.2 −0.3 −6
VLA1.17d B 0 0 2.96±0.06 12.7 0.54 22.9±5.1 −73.7±3.3 1.1e 10.4+0.6 42.2 809±182 90+11
−2.3 −11
VLA1.18d B 8.2520 31.166 6.17±0.06 12.4 0.52 22.0±5.0 −89.7±1.2 1.0e 10.3+0.5 − − 90+12
−2.3 −12
VLA1.19d B 8.2941 33.016 11.39±0.07 12.6 0.53 25.7±3.5 +58.2±7.7 0.9e 10.5+0.4 5.9 −92±30 90+6
−1.1 −6
VLA1.20 C 36.2920 -230.717 1.86±0.02 8.1 0.33 2.7±0.3 +47.1±2.9 0.9+0.2 9.5+0.3 − − 84+3
−0.1 −0.3 −11
VLA1.21d B 57.6378 17.914 3.56±0.15 11.2 0.44 − − − − 25.8f −398±102e −
VLA1.22d B 84.3726 42.854 6.71±0.05 9.8 0.43 0.9±0.2 −20.6±6.2 1.6+0.1 8.9+0.6 − − 78+12
−0.3 −1.3 −33
VLA1.23 A 295.3883 281.223 39.41±0.05 21.1 0.83 0.6±0.1 −30.6±8.5 − − − − −
VLA1.24 A 332.4381 316.208 2.00±0.04 21.7 0.68 1.9±0.1 −31.8±8.3 − − − − −
VLA1.25 A 507.5828 121.956 1.33±0.03 23.8 0.51 3.6±0.3 −9.9±7.6 1.5+0.3 9.7+0.3 − − 90+11
−0.4 −1.0 −11
VLA2.01 - 571.4096 -488.743 1.69±0.01 5.2 0.57 − − − − 30.2f 567±243e −
VLA2.02 - 571.9570 -487.320 2.82±0.01 5.0 0.58 − − − − 21.5f 413±146e −
VLA2.03 - 587.4926 -621.986 12.0±0.05 6.8 0.68 0.5±0.1 −89.5±52.7 3.0+0.2 8.7+0.5 − − 80+10
−0.2 −0.1 −36
VLA2.04 - 602.6914 -611.675 6.86±0.06 12.7 0.61 0.8±0.3 +82.8±3.8 − − − − −
VLA2.05 - 604.3755 -620.571 4.62±0.12 10.9 0.50 1.5±0.4 −80.6±32.0 1.9+0.2 9.2+0.1 − − 90+22
−0.3 −2.3 −22
VLA1.26 A 605.6807 212.257 20.73±0.04 18.7 0.53 2.3±0.1 −17.9±2.0 2.0+0.2 9.4+0.3 − − 85+3
−0.4 −0.1 −12
VLA2.06 - 625.2582 -454.392 25.44±0.16 11.2 0.69 0.2±0.1 −47.0±16.6 − − − − −
VLA2.07 - 676.4543 -432.327 1.95±0.08 6.0 0.60 2.0±0.3 +47.7±8.7 2.3+0.2 9.4+0.2 − − 90+15
−0.2 −1.9 −15
VLA2.08 - 677.0438 -432.545 4.65±0.01 6.2 0.59 1.1±0.1 +34.5±3.1 2.5+0.2 9.0+0.3 − − 83+8
−0.2 −0.1 −16
VLA2.09 - 690.0533 -439.411 30.29±0.06 9.6 0.75 0.4±0.1 −58.7±9.1 3.3+0.1 8.6+0.7 3.9 −158±36 76+11
−0.2 −0.1 −40
VLA2.10 - 690.7269 -440.163 48.76±0.06 9.7 0.58 0.4±0.1 −69.3±9.0 2.6+0.1 8.6+0.8 6.2 −186±31 72+8
−0.2 −0.1 −44
VLA2.11 - 697.1685 -459.564 13.99±0.04 -1.1 0.86 1.0±0.3 −39.2±19.2 − − − − −
VLA2.12 - 697.4212 -460.125 3.25±0.01 -2.9 2.34 1.5±0.6 −34.2±12.1 − − − − −
VLA2.13 - 697.8843 -458.710 7.16±0.01 0.5 0.69 1.0±0.2 −38.3±7.5 3.0+0.1 9.0+0.6 − − 82+9
−0.3 −1.1 −13
VLA2.14 - 698.6000 -445.961 62.39±0.07 7.2 0.63 1.7±0.2 −84.8±1.2 2.7+0.3 9.3+0.7 1.1 −38±13 74+15
−0.5 −0.2 −35
VLA2.15 - 698.9368 -461.704 49.20±0.05 -0.7 0.87 2.0±0.2 +78.2±36.9 3.4e 9.6+0.4 − − 68+3
−0.3 −41
VLA2.16 - 699.3157 -446.842 30.98±0.05 6.8 0.58 1.3±0.2 −75.4±1.6 2.5+0.2 9.1+0.7 15.6 470±73 77+13
−0.3 −0.3 −33
VLA2.17 - 701.5051 -448.017 12.73±0.05 8.3 0.53 2.5±0.5 −71.7±1.7 2.1+0.3 9.5+0.8 16.1 425±76 77+13
−0.3 −0.9 −34
VLA2.18 - 702.3892 -449.426 4.32±0.03 9.3 0.82 1.2±0.1 −76.9±4.4 − − − − −
VLA2.19 - 703.4839 -451.126 39.57±0.17 11.6 0.78 2.6±0.3 −72.6±3.7 − − − − −
VLA2.20 - 703.8628 -451.496 72.56±0.16 11.2 1.09 2.0±0.3 −65.1±3.9 3.1+0.1 10.2+0.3 − − 61+42
−0.9 −0.2 −42
VLA2.21 - 705.5048 -458.549 2.48±0.01 0.2 0.53 1.4±0.3 −8.7±5.2 2.1+0.1 9.2+0.2 − − 90+16
−0.3 −2.0 −16
VLA2.22 - 709.9675 -458.458 4.02±0.01 5.0 0.56 0.8±0.1 −8.0±4.4 2.5+0.1 8.9+0.5 20.8 957±239 84+6
−0.2 −0.4 −11
VLA2.23 - 710.3886 -459.141 1.34±0.01 4.9 0.80 2.1±0.1 −53.5±5.5 3.0+0.2 9.4+0.1 − − 87+1
−0.1 −0.1 −9
VLA2.24 - 711.0201 -457.901 198.45±0.20 8.7 0.51 0.7±0.1 −72.0±6.2 2.3+0.2 9.0+0.5 6.1 160±24 66+11
−0.3 −0.3 −42
VLA2.25 - 713.0410 -466.595 16.74±0.02 13.7 0.60 0.8±0.1 −44.5±2.1 2.6+0.2 8.9+0.5 − − 80+10
−0.3 −0.1 −38
VLA2.26 - 716.0303 -475.323 53.72±0.06 -7.7 0.68 1.2±0.5 −89.6±61.8 3.0+0.2 9.1+0.5 2.1 79±18 77+13
−0.3 −1.8 −38
VLA2.27 - 777.7941 -447.278 0.67±0.06 -7.7 0.66 6.1±0.1 −87.7±6.5 0.7+1.8 10.7+0.1 − − 71+13
−0.1 −1.7 −3
aThewatermasersassociatedwithVLA1aredividedinthreegroupsaccordingtotheirpositions(seeSect.4.1).
bThebest-fittingresultsobtainedbyusingamodelbasedontheradiativetransfertheoryofwatermasers(Vlemmingsetal.2006,2010)forΓ+Γν=1.Theerrorsweredeterminedby
analysingthefullprobabilitydistributionfunction.ForT ∼400K(Γν =2)andT ∼2500K(Γν =13)Tb∆Ωhastobeadjustedbyadding+0.48and+1.15,respectively(Nedoluha&
Watson1992,Anderson&Watson1993).
cTheanglebetweenthemagneticfieldandthemaserpropagationdirectionisdeterminedbyusingtheobservedPlandthefittedemergingbrightnesstemperature.Theerrorsweredetermined
byanalysingthefullprobabilitydistributionfunction.
dBecauseofthedegreeofthesaturationofthesewatermasersTb∆Ωisunderestimated,∆Viandθareoverestimated.
eTheconstraintfitdidnotallowustoevaluatetheerrorbarsproperly.
fInthefittingmodelweincludethevaluesforTb∆Ωand∆Vioftheclosestfeatures.
This could be done by fitting the water maser emissions with earlywith(Γ+Γ ),whicharethemaserdecayrateΓandcross-
ν
thefullradiativetransfermethoddescribedbyVlemmingsetal. relaxationrateΓ .Sinceforthe22GHz H O masersΓ . 1s−1
ν 2
(2006, 2010), who successfully used it for water and methanol and Γ dependson the gas temperature,we include in our fit a
ν
masersinCepheusA.Theseradiativetransfermethodsarebased value (Γ+Γ ) = 1s−1, so that we could adjust the fitted T ∆Ω
ν b
onthemodelsforwatermasersofNedoluha&Watson(1992), valuesbysimplyscalingaccordingto(Γ+Γ ).Notethat∆V and
ν i
for which the shapes of the total intensity, linear polarization, θdonotneedtobeadjusted.
andcircularpolarizationspectradependonT ∆Ωand∆V.They We model the observed linear polarized and total intensity
b i
alsofoundthattheemergingbrightnesstemperaturescaledlin- maser spectra by gridding the intrinsic thermal linewidth ∆V
i
Surcisetal.:W75N:watermaserspolarization. 5
between 0.4 and 3.5 kms−1, in steps of 0.025 kms−1, using a aredistributedellipticallyaroundits1.3cmcontinuumpeak.In
least square fitting routine (χ2-model) with an upper limit of Fig. 4 we report two ellipses obtained by fitting the positions
the brightness temperature T ∆Ω = 1011 K sr. In Fig. 1 two ofthewatermasersdetectedbyT03(calledhereellipse 1)and
b
examples are shown of the results obtained by fitting the wa- inthepresentpaper(ellipse2),theirparametersarereportedin
termaseremissionswiththefullradiativetransfermethodcode. Table 2. The major axis of ellipse 2 is 64 mas larger than that
Fromthesefirstfits,wecanthenobtaintheanglesθbyconsid- of ellipse 1, the minoraxes are also in the same ratio. This in-
eringtherelationbetween P andθ(seeVlemmingsetal.2010 dicates an expansion velocity of ∼4.8 mas/yr in all directions,
l
formoredetails).ThebestvaluesforT ∆Ωand∆V arethenin- corresponding to ∼ 46 kms−1 (at 2 kpc). This value is con-
b i
cludedinthefullradiativetransfercodetoproducetheI andV sistent with the proper motion of the water masers as reported
modelsthatwereusedforfittingthetotalintensityandcircular byT03.Sincethe expansionvelocityissymmetricin alldirec-
polarized spectra of the H O masers extracted from the high- tions, we aligned the centers of the two ellipses with the posi-
2
resolutionI andV cubes.Themagneticfieldstrengthalongthe tion of the 1.3 cm continuum peak obtained from a Gaussian
lineofsightisevaluatedbyusingtheequation fit. Consequently we were able to overlay all water masers of
theregiononthecontinuumVLAmapwithanaccuracyofap-
P ∆v
B = Bcosθ= V L , (1) proximately10mas.Inthiswaywecouldalsocomparethepo-
|| 2·AF−F′ sitionsofthemasersaroundVLA1atthetwodifferentepochs.
IfwelookatthenorthernmasersaroundVLA2,weseethatthe
where ∆v is the FWHM of the total intensity spectrum, the
L masersdetectedbyT03arebetteralignedwiththeellipse1than
A coefficient, which depends on T ∆Ω. The circular polar-
F−F′ b themasersdetectedbyuswithellipse2(Fig.4,rightpanel),in-
izationfraction(P )areobtainedfromtheIandV models.
V dicatingthatthosemasersarealso movingnortheastward.This
Another important aspect that must be considered in the
movementmightbeduetotheformationofajet(seeFig.3).
analysis of the water masers is the degree of their saturation,
ThewatermasersassociatedwithVLA1canbedividedinto
which influences P. We are able to estimate it by considering
l three groups,one composedof the northernmasers(VLA1A),
theratiobetweenmaserrateofstimulatedemission(R)andthe
oneof the masersclosest to the centralprotostar(VLA1B, see
sumofmaserdecayrateandthecross-relaxationrate(Γ+Γ ).
ν
Thestimulatedemissionrateisgivenby
Table2.Number,epoch,semi-majoraxis,semi-minoraxis,and
Ak T ∆Ω
R≃ B b , (2) inclination from the elliptical fit of the water maser associated
4πhν withVLA2inthetwodifferentepochs.
whereA = 2×10−9s−1 isthe22-GHzH O maserspontaneous
2
rate(Goldreich&Keeley1972),k andharetheBoltzmannand Ellipse epoch a b θ ref.
B (mas) (mas) (◦)
Planckconstants,respectively,νthemaserfrequencyandT ∆Ω
b
theemergingbrightnesstemperatureadjustedattherightΓ+Γ 1 2-Apr.-1999 71.0 61.1 123.5 T03
ν
value.ThemasersareunsaturatedwhenR/(Γ+Γ ) < 1,inthe 2 21-Nov.-2005 103.1 92.9 120.5 thiswork
ν
onsetofsaturationwhenR/(Γ+Γ )≥1,andfullysaturatedwhen
ν
R/(Γ+Γ )≈100(Vlemmingsetal.2006).Moreover,whenthe
ν
saturationsetsin,themaserlinesstarttobroadenagain,andthis
meansthattheobservedlinewidth(∆v )becomesaslargeas∆V Fig. 5 right panel) and the third one of the southern masers
L i
orlargerdependingonthedegreeofthesaturation. (VLA1C). The east-northernmasers of the VLA1A groupdid
Although the full radiative transfer method is based on a not show emission during the previous observation of epoch
model for general water masers, it should be noted that, when 1999 (T03). All VLA1A masers (VLSR ≃ 19 − 25 kms−1)
this model is applied to saturated masers, it is impossible to are well-aligned with the direction of the thermal radio jet de-
properly disentangle the values of T ∆Ω and ∆V. In this case tected by Torrelles et al. (1997) and the methanol masers de-
b i
the method only provides a lower limit for T ∆Ω and an up- tected by S09. The VLA1B and VLA1C groups show a dis-
b
per limit for ∆V. For high maser brightness temperatures, i.e. tribution quite similar to that shown on April 2nd 1999 (T03).
i
greaterthan109cm−3Ksr(Nedoluha&Watson1992),thecosθ For these two groups T03 reported a proper motion of about
dependence of Eq. 1 breaks down thereby introducing a more ∼2 mas/yr (∼ 19 kms−1). Although we have detected several
complexdependenceonθ.Asaresult,theangleθobtainedfrom water masersclose to thosedetectedby T03,itis verydifficult
thefitisoverestimatedand,inparticular,itcouldalsoreachval- to estimate the propermotion between the two epochs. In fact,
uesof90◦orgreater.Consequently,thevaluesofT ∆Ω,∆V and as is shown on the right panel of Fig. 5, it is difficult to iden-
b i
θobtainedforsaturatedmaserscouldnotbetakenintoaccount tify a maser that arises in both epochs (1999 and 2005). The
intheanalysisoftheregion. LSR radial velocities of VLA1 are between 9 and 25 kms−1,
whichareabout3-4timeshigherthanthoseofmethanolmasers
(S09)andgreaterthanthesystemicvelocityofthewholeregion
4. Results (10kms−1,Shepherdetal.2003).
4.1.Maserdistribution
4.2.Linearandcircularpolarization
In Fig. 3 we show the distribution of the 124 22-GHz water
maserfeaturesdetectedwiththe VLBA. Nowater maseremis- Linearpolarizationwasdetectedin50masers(Table1).Inpar-
sion with a peak flux density less than 80 mJybeam−1 is de- ticular, about 70% of the H O masers associated with VLA1
2
tected. All H O masers are detected around the radio sources show linear polarization emission. The corresponding linear
2
VLA1(29%,36/124)andVLA2(71%,88/124). polarization vectors are shown in Figs. 4 and 5 (left panel).
The water masersassociated with VLA2, whichhavelocal The highest fractional linear polarization was detected in wa-
standardofrestvelocities(V )between-7.7and13.7kms−1, ter masers related to VLA1 (P ∼ 26%). Eight masers associ-
LSR l
6 Surcisetal.:W75N:watermaserspolarization.
42 37 35.4
VLA 1
35.2
)
0
0 35.0
0
2
J
(
N
O
TI
A 34.8
N
LI
C
E
D
34.6
34.4
VLA 2
34.2
20 38 36.46 36.44 36.42 36.40 36.38 36.36
RIGHT ASCENSION (J2000)
Fig.3.Positionsofwaterandmethanolmaserssuperimposedonthe1.3cmcontinuumcontourmapoftheVLA1thermaljetand
VLA2 (Torrellesetal. 1997).Contoursare−3,−2,2,3, 4,5,6,7,8× 0.16mJybeam−1. Blue trianglesindicatethe positionsof
the water maser features detected with the VLBA (this paper). Red circles indicate the position of the methanolmasers features
detectedbyS09.Thetwoarrowsindicatethedirectionofthelarge-scalemolecularbipolaroutflow(PA=66◦).
atedwithVLA1havehighlinearpolarizations,P between10% lyzing the full probability distribution functions. Eight of the
l
and26%,sothatthemedianvalueoflinearpolarizationsforthe H O masersaroundtheradiosourceVLA1(fromVLA1.12to
2
masers in VLA1 is ∼7%. In contrast, the maximum P for the VLA1.19,which are the closestfeaturesto the protostar)show
l
masers associated with VLA2 is 6%, and the median value is valuesforθequalto90◦.Thissuggeststhat,fromthetheoryof
much lower at 1.5%. Circular polarization was detected in 20 the Zeemaneffect, we are observingperpendicularto the mag-
masers,equallydistributedbetweenVLA1andVLA2. neticfield.Consequently,weshouldnotbeabletodetectcircu-
The rms weighted linear polarization angles of the water larpolarizationfromthesemasers.However,since wedetected
masers around VLA1 and VLA2 are hχ i ≈ −67◦±40◦ and circularpolarizationinthreeofthem(VLA1.16,VLA1.17,and
1
hχ i ≈ −72◦±32◦, respectively. Since the water masers of VLA1.19)themeasurementsofθmustbeaffectedbythedegree
2
group VLA1A show ordered linear polarization vectors, it is of their saturation, as discussed in § 3.2. By excluding these 8
worthwhile evaluating its weighted polarization angles, which masers,weevaluatedhθi = 83◦±7◦ andhθi = 85◦±6◦,
VLA1 VLA2
ishχ i=−20◦±9◦. whichimplythatthemagneticfieldsareclosetotheplaneofthe
1A
We were able to fit 42 masers with the method described sky.
in § 3.2, and the results are given in columns 10 and 11 of Incolumn13ofTable1wereportthemagneticfieldstrength
Table 1, and two examples (i.e., VLA1.19, VLA2.22) are re- along the line of sight obtained by fitting the high-resolution
portedinFig.1.Althoughtheemergingbrightnesstemperatures spectraofI andV (see§3.2).ThemasersVLA1.21,VLA2.01,
forVLA1andVLA2arebothintherange108Ksr <T ∆Ω < andVLA2.02donotshowanylinearpolarizationemission,but
b
1010Ksr,theirweightedintrinsicmaserlinewidthsareverydif- they do show circular polarization emission. Consequently we
ferent. They are h∆Vi = 1.0+0.8 kms−1 and h∆Vi = could not get T ∆Ω and ∆V as done for the other masers that
i VLA1 −0.2 i VLA2 b i
2.5+0.6kms−1,respectively. showed linear polarization emission, and to measure the mag-
−F0i.4nally, considering the emerging brightness temperature netic field strength, we used the Tb∆Ω and ∆Vi values of the
and the observed P we determined θ for those masers which closest masers. In Table 1 the circular polarization fraction is
l
we were able to fit for ∆V and T ∆Ω. The results are given alsoreported(column12).
i b
in column 14 of Table 1 with the errors determined by ana-
Surcisetal.:W75N:watermaserspolarization. 7
Fig.4. Left panel: a close-up view of the H O maser features around the radio source VLA2. Right panel: a zoom-in view of
2
theboxedregionoftheleftpanel.Theoctagonalsymbolsaretheidentifiedmaserfeaturesinpresentworkscaledlogarithmically
according to their peak flux density. The maser LSR radial velocity is indicated by color. A 5 Jybeam−1 symbol is plotted for
illustrationinbothpanel.Thelinearpolarizationvectors,scaledlogarithmicallyaccordingtopolarizationfractionP (inTable1),
1
areoverplotted.Twoellipsesarealsodrawninbothpanels.Theyaretheresultsofthebestfitofthewatermasers(crosses)detected
by T03 (dotted ellipse, ellipse 1; epoch 1999) and of those detected in present work (solid ellipse, ellipse 2; epoch 2005). Their
parametersarelistedinTable4.Thesynthesizedbeamis2.0mas×0.7mas.
Fig.5.Leftpanel:aclose-upviewoftheH O maserfeatures(octagon)aroundtheradiosourceVLA1.Rightpanel:azoom-inview
2
ofthewatermasersgroupVLA1B,herearealsoreportedthewatermasers(square)detectedbyT03.Theoctagonalandthesquare
symbolsarescaled logarithmicallyaccordingtotheir peakflux density.ThemaserLSRradialvelocityisindicatedbycolor.The
systemicLSRradialvelocityofW75Nmassivestar-formingregionis10kms−1(Shepherdetal.2003).Symbolsfor3Jybeam−1
andfor1Jybeam−1 areshowninthelowerleftcornerofthepanels.Thelinearpolarizationvectors,scaledlogarithmicallyaccording
topolarizationfractionP (inTable1),areoverplottedontheleftpanel.Thesynthesizedbeamis2.0mas×0.7mas.
1
5. Discussion pendentgroups,theirpolarizationandfinallythetypeofshocks
thatgiverisetothem.
The water masers detected using the VLBA are related to two
different radio continuum sources of W75N(B), VLA1 and
VLA2, which are thought to be in two different evolutionary
stages.Inthenextsections,wediscusstheirnatureastwoinde-
8 Surcisetal.:W75N:watermaserspolarization.
5.1.HOmaserproperties and ∆Ω ≈ 10−2 sr (Γ+Γ = 14s−1). In a tubular geom-
2 VLA2 ν
etry ∆Ω ≈ (d/l)2, where d and l are the transverse size and
As suggested by the difference between their intrinsic thermal
length of the tube respectively. Assuming d approximately the
linewidths, which is 1.5 kms−1, the masers around VLA1 and
sizeofthemaserfeatures,themaserlengthsareallintherange
VLA2 appear to arise under two different physical conditions. 1012cm.l.1014cm.
Using
∆Vi ≈0.5×(T/100)21, (3) 5.2.MagneticfieldinW75N(B)
we estimate the gas temperatures in the H O masing regions 5.2.1. Magneticfieldstrength
2
around VLA1 and VLA2, and find them to be T ∼
VLA1
400 K and T ∼ 2500 K, respectively. The intrinsic ther- The magnetic field strength along the line of sight was deter-
VLA2
mal linewidths of the VLA2 masers are quite similar to each mined from circular polarization measurements for 20 water
other, while the three groups of masers around VLA1 show masers. We detected significant (≥ 3σ) magnetic fields toward
a difference in their ∆V. In particular, we find that the maser 10masersinVLA1and7masersinVLA2.Thedetectedfields
i
groupsVLA1AandChavesimilarintrinsicthermallinewidths alongthelineofsight(B||)rangefrom−400mGto+600mGfor
(h∆Vii = 1.0kms−1)whilethegroupVLA1Bhaslowervalues theH2O masersaroundVLA1andfrom−200mGto1000mG
(h∆Vi = 0.7 kms−1). This can come from the degree of their forthose associated with VLA2 (see Table 1). The wide range
i
saturationthatwecanevaluatebyconsideringEq.2. obtained for BV||LA1 and BV||LA2 is not anomalous, and a similar
Before using Eq. 2 we have to adjust the emerging bright- rangewasreportedforCepheusA(Vlemmingsetal.2006).The
nesstemperaturevaluesasexplainedin§3.2.Atatemperature changinginthesignindicatesthereversalofthemagneticfield
of∼400K,forwhichΓ+Γ =3s−1(Nedoluha&Watson1992), (negativetowardstheobserver,positiveawayfromtheobserver)
ν
the weighted emerging brightness temperature of VLA1B is as already observed in other sources (e.g., Vlemmings et al.
hT ∆Ωi > 1011Ksr, which is close to the limit of full satura- 2006).Sincetheθvaluesarecloseto90◦,i.e.closetotheplane
b
tion,andforVLA1AandCishT ∆Ωi = 1010Ksr.FromEq.2 ofthesky,aslightdifferenceintheseanglescanproduceanin-
b
themasersofVLA1AandVLA1Carethuscompletelyunsat- versionofthesignofthemagneticfields.InthecaseofVLA1,
urated(R/(Γ+Γ ) = 0.5), while VLA1B masersare saturated thepositiveandnegativemagneticfield strengthsmeasuredfor
ν
(R/(Γ+Γ )>5),eveniftherealdegreeoftheirsaturationisun- VLA1B andVLA1Cgroupsindicatea twisted magneticfield,
ν
knownbecauseofunderestimatingTb∆Ω.Theaverageobserved which is not the case for VLA2. Apart from the H2O masers
linewidth (∆v ) of VLA1B group, which is about 0.5 kms−1, VLA2.01andVLA2.02,whicharelocatedontherightedgeof
L
gives us another indication of the saturation state of their H O theellipse2andshowpositivevalues,theothermasersintheup-
2
masers.Asreportedin§3.2,theintrinsicthermallinewidthsfor perleftpartoftheellipse2(fromVLA2.09toVLA2.26)present
thesaturatedmasersareoverestimatedso∆v mightbelargeror a clearseparationbetweenpositiveandnegativemagneticfield
L
atleast equalto ∆V implyingthatthe maserlinesare rebroad- strengths(seeFig.4andTable1).
i
ened as expected in the saturated maser lines. As the VLA1B Based on these detections the absolute weighted magnetic
masers are saturated, they have not been taken into account in field strengths along the line of sight, where the weights are
ourconclusions.FortheothermasersassociatedwithVLA1,we wi = 1/e2i and ei is the error of the i-th measurement, are
observedlinewidthswellbelowtheirintrinsicthermallinewidth, h|BVLA1|i = (81±62)mGandh|BVLA2|i = (145±110)mG.In
|| ||
andthisconfirmsthatthesemasersareunsaturated.Themasers thecaseofVLA1,itismorecorrecttodetermineh|B |iforonly
||
associatedwithVLA2areunsaturated;infact,forT ∼ 2500K theunsaturatedmasers;i.e.,h|BVLA1|i=(74±50)mG.Thelarge
(Γ+Γ = 14s−1, Anderson & Watson 1993), we get ratios of ||
ν errorsareduetothelargescatter ofthe magneticfieldstrength
R/(Γ+Γν)<0.5. values,whichdependontheprojectioneffects,onthedensityof
From the measurements of the maser flux densities (S(ν))
each water maser region,and on the local velocitiesof shocks.
andfeatureangularsizes(Σ)wecanestimatethebrightnesstem- Becauseofthisscatteritisdifficulttoevaluatethetruestrength
peraturebytheequationreportedinS09,
of the full magneticfields of the entire region.To comparethe
results obtained using different maser species we can estimate
Tb = S(ν)· Σ2 −1·ξ , (4) themroughly.Since hθiVLA1 = 83◦+−715◦◦ and hθiVLA2 = 85◦+−636◦◦
[K] [Jy] [mas2]! H2O the absolute weighted magnetic field strengths are |BVLA1| =
h|BVLA1|i/coshθi ∼ 700mG and |B | ∼ 1700mG, re-
|| VLA1 VLA2
whereξ = 1.24×109mas2Jy−1Kisaconstantfactorwhich spectively.SuchhighvaluesarealsofoundbyVlemmingsetal.
includesH2aOll constant values, such as the Boltzmann constant, (2006)inCepheusA.Astheirshockvelocityis10kms−1,mea-
the wavelength, and the proportionality factor obtained for a suring a magnetic field of 600 mG, these authors explain this
Gaussian shape by Burnset al. (1979). ComparingT with the highvaluewiththepresenceofanearbymagneticdynamo.
b
emergingbrightnesstemperaturesobtainedfromthemodel,we If the magnetic field is important throughout the collapse
can easily estimate ∆Ω. The Gaussian fit of the masers give a of a spherical cloud, the cloud forms primarily by flows along
size of 0.4mas for the water masers related to VLA1, which field lines, and the conservation of magnetic flux and of the
indicates that they are unresolved, while a size of 0.5 for the massimply B ∝ n0.47 (Crutcher1999).Althoughthescalingof
masers associated with VLA2, which indicates that they are themagneticfieldin watermaser isduetoshocks,Vlemmings
marginallyresolved.Thebrightnesstemperaturesofthebright- (2006, 2008) empirically showed that even water masers with
est masers in both groups are T > 1.23× 1012 K and numberdensitiesupto1011cm−3 followthisrelation,evenifin
b,VLA1.08
T ≈ 1012 K, for which we find ∆Ω < 10−2 sr non-masinggasofsimilardensitiesthemagneticfieldstrengths
b,VLA2.24 VLA1.08
and ∆Ω ≈ 10−2sr for T = 400 K and 2500 K, respec- are likely lower due to ambipolar diffusion. Hence, from the
VLA2.24
tively.Consideringallthe featuresinTable 1, themaser beam- magnetic field strengths, we can determine the number den-
ing of the two groupsare ∆Ω . 10−2 sr (Γ+Γ = 2s−1) sity of the cloud where the H O masers arise by the relation
VLA1 ν 2
Surcisetal.:W75N:watermaserspolarization. 9
B ∝ n0.47. Considering the pre-shock magnetic field (B = jumpshock(J-shock)oranondissociativecontinuousshock(C-
||
16 mG) and density (n = 109cm−3) obtained by S09 from shock).InthecaseofaC-shock,theintrinsicFaradayrotationat
methanolmaseremissioHn2s,wehavenVLA1 = 3×1010cm−3 and 22GHzcanbeconsiderednegligiblebecausetheionizationstate
nVLA2 = 1×1011cm−3.SimilarcalculHa2tioncanbedoneconsid- of the gasis controlledby cosmic-rayionization,which gener-
H2 ateselectron-molecularionpairsatarateof10−17s−1 (Kaufman
ering the magnetic field strength obtained during the OH flare
&Neufeld1996),whilewehavetoroughlyestimateitscontribu-
nearVLA2(Slyshetal.2010).Theymeasuredamagneticfield
tionforaJ-shock.Inthiscase,theelectrondensityisn =χ n ,
strengthofabout70mG,whichiscomparabletothetotalmag- whereχ istheionizationfraction(χ =10−5−10−4;eKylaefisH&2
netic field strength inferred by S09 (B = 50 mG). Since from e e
Norman1987)andn thenumberdensityoftheH O masers.
the OH maser observations is generally possible to get only In the most conservaHt2ive situation, for which n ∼2 104cm−3,
B and not B and since the flare occurred near VLA2, in this e
||
we determine an internal Faraday rotation of tens of degrees,
case we have to use the value of |B | instead of the value
VLA2
of h|BVLA2|i. For an OH number density of 108cm−3, we find enough to destroy the linear polarization in some cases. This
|| means that either the electron density must be much lower or,
nVLA2 ≈9×1010cm−3forthewatermasersaroundVLA2,close
H2 morelikely,theinternalFaradayrotationmustbenegligible,and
towhatisobtainedconsideringthemethanolmasers.Thesere- consequentlytheshockmustbeaC-shock.
sults, which are close to the extreme upper limit of 1011 cm−3
of thewater maser thermalisation(Elitzuretal. 1989), confirm
the assumptionson the CH OH and OH numberdensities and 5.2.3. Magneticfieldorientation
3
reinforcethemeasurementsofthemagneticfieldstrengthmade
byS09,Slyshetal.(2010)andinthispaper.Thethreeindepen- SupposethatthewatermasersinbothsourcesarepumpedbyC-
dentmagneticfieldstrengthvaluesobtainedfromthreedifferent shocks(see§5.3),thismeansthattheFaradayrotationis<26◦.
maser species located at differentpositionsshow that the mag- SincetheH2O masersassociatedwith VLA1andVLA2 indi-
netic field in massive star-forming regions can reach strength cate thatthereare two differentmagneticfieldsaroundthe two
larger than expected. This large unexpected values of B might sources, it is worthwhile discussing them separately. All water
be due either to the presence of a magnetic dynamo or to the masersshowθ > θcrit ∼ 55◦;i.e.,themagneticfielddirectionis
presenceofveryhigh-velocityshocksthatstronglycompressthe perpendiculartothelinearpolarizationvectorsandclosetothe
gas,andconsequentlyincreasethemagneticfieldstrengthfrom planeofthesky.
50-70mG (pre-shockmethanol-OHregion)to > 1 Gauss (wa- The group VLA1A shows an orientation of the magnetic
termaserregion).Forthelasthypothesiswereferthereaderto field ϕVLA1A = 70◦±9◦, whichisin goodagreementwith what
B
§5.3. isreportedbyS09(ϕB = 73◦±10◦).Thisalsoconfirmsthatthe
With regard to the orientation of the magnetic field, it is magnetic field orientation in this part of the source is close to
worthwhile to discuss the influence of the Faraday rotation in the direction of the large-scalemolecularbipolaroutflow (66◦)
ourobservations. (S09).TheothertwogroupsVLA1BandCshowdisorderedlin-
ear polarization vectors, which cannot give information on the
orientationofthemagneticfield.Thedifferentorientationofthe
5.2.2. Faradayrotation linearpolarizationvectorsforthesetwogroupsmightcomefrom
thedifferentdepthsofthemasers,i.e.highvaluesoftheambient
Themeasurementsofthelinearpolarizationangle(χ)mightbe
Faraday rotation, and therefore we can assume that the masers
disturbedbytheforeground,ambient,andinternalFaradayrota-
withanglesfarfrom−20◦arelocatedathigherdepthsthanthose
tion,whicharegivenby
withanglescloseto−20◦(e.g.VLA1.14).
Although around VLA2 some H O masers show the 90◦-
Φ[◦]=1.35×10−15D[cm]n [cm−3]B [mG]ν−2[GHz], (5) 2
e || flipofthelinearpolarizationangle(e.g.,VLA2.15),whichwas
alsoobservedinsomewatermasersinCepheusA(Vlemmings
where D is the length of the path over which the Faraday ro-
etal.2006),themagneticfieldappearstoberadial(seeFig.4).
tation occurs, ne and B|| are respectively the average electron Consideringhχ i≈−72◦,themagneticfieldorientationisabout
density and the magnetic field along this path and ν the fre- 18◦ thatindica2tesa misalignmentwith the large-scalemolecu-
quency.FortheforegroundFaradayrotation(i.e.therotationdue
larbipolaroutflowandthemagneticfieldofVLA1,eventhough
tothemediumbetweenthesourceandtheobserver)φf is0◦.4at theinclinationofbothellipses(∼60◦)isclosetothem.
22GHz,assumingtheinterstellarelectrondensityandthemag-
The different orientation of the magnetic fields of the two
neticfieldaren ≈0.012cm−3andB ≈2µG,respectively.
e || sources again implies differentlocal physicalenvironmentsfor
All water masers are not in the same plane, but they arise
VLA1 and VLA2, and suggests that VLA1 is the powering
in cloudsat differentdepths.This depthis impossible to deter-
source of the large-scale molecular bipolar outflow, as already
minefromourobservations,soinordertoestimatetheambient
indicatedbyS09.
Faraday rotation (i.e. the rotation due to the ambient medium
wheremasersarise),weassumethemaximumprojectedsepara-
tionasupperlimitof Damongmasersclosetoeachotherwith 5.3.J-orC-shocks?
the highest velocity difference, which is D ≈ 2 × 1015cm.
max
For the electron density,we assume the value reportedby Fish Asreportedinpreviouspapers,theH O masersarethoughtto
2
& Reid (2006), n ≈ 300cm−3, which can produce a rotation bepumpedbythetransitionofshocksgeneratedinthetwopro-
e
of90◦alongthepathamplificationofOHmaseraroundHIIre- tostars (e.g., Torrelles et al. 1997). These shocks can be either
gions.ConsideringB = 16mG(S09),theambientFaradayro- C-shocksorJ-shocks.Inawatermasingregion,C-shocksshow
||
tationisφ <26◦. v . 50 kms−1 and a post-shock T < 4000 K (Kaufman &
a
The internal Faraday rotation, which can destroy the linear Neufeld 1996), while J-shocks have velocities v & 45 kms−1
polarization(Fish & Reid 2006), dependson the type of shock and a post-shock plateau temperature that depends on the pre-
that pumpsthe water masers. They can either be a dissociative shockdensity,theshockvelocity,andtheobservedlinewidthof
10 Surcisetal.:W75N:watermaserspolarization.
thewatermasers(Elitzuretal.1989).Kylafis&Norman(1991) 8isvalidforbothtypesofshocks(Kaufman&Neufeld1996).
investigatedtheeffectsofthetemperatureontheirJ-shockmodel Consideringn =109cm−3asbefore,weget|B |≈500mG
H2 VLA1
up to a temperature of about 1000 K, and they found that the and|B |≈1200mG,whichareconsistentwiththeroughval-
VLA2
brightness temperature starts to increase more slowly when a uesreportedin§5.2,makingthepresenceofamagneticdynamo
temperature∼500Kisreached. unnecessary.Finally,fromourresultswecanstronglydepictthe
WefirstconsiderVLA1.Assumingthepropermotionofthe followingscenario for VLA1: the hotgas in a C-shock pumps
watermasersreportedbyT03,i.e.19kms−1,asthevelocityof the water masers and at the same time the warm dust associ-
theshockandaspre-shockdensity,thedensityofthemethanol ated with this shock emits infrared photons, which pumps the
masersn =109cm−3(S09),theequationreportedinElitzuret methanolmasers in the pre-shockregion, as already suggested
H2
al.(1989)foraJ-shock inS09.
n v ∆v−1 2
T ≈400 0 s 9 K, (6)
plateau (cid:18)107cm−3 100kms−11kms−1(cid:19) 6. Conclusions
where here ∆v = 0.5 kms−1, gives us a most likely post- Weobservedthe22GHzwatermasersinfullpolarizationmode
VLA1
shock temperature TVLA1 ≈ 900 K, which is higher than the in the massive star-forming region W75N with the VLBA. We
plateau
temperaturemeasuredconsideringtheweightedintrinsicveloc- detected124watermasersaroundthetworadiosourcesVLA1
ity obtainedfromourfit(T ≈ 400K).Moreover,fromthe and VLA2, which appear to be in two different evolutionary
VLA1
model of a C-shock described by Kaufman & Neufeld (1996), stages; i.e., VLA1 is more evolved and a radio jet has not
weareabletoestimatethevelocityoftheshockbyconsidering formed yet in VLA2. From our observations, we have shown
the temperature T . The model for a temperature of 400 K that these two sources separated in the sky by ∼ 1′′ (2000
VLA1
givesvVLA1 ∼ 15kms−1,whichisclosetothepropermotionof AU) have different local physical environments. In particular,
s
theH O masers.Therefore,followingourarguments,theshock the linear polarization indicates that the magnetic field around
2
inVLA1ismostlikelyaC-shock. VLA1 is aligned well with the large-scale molecular bipolar
In the case of VLA2, the plateau temperature due to a J- outflow, while the magnetic field is well-ordered around the
shock is TVLA2 ≈ 1000K (where ∆v = 0.7 kms−1, and shell-like expandinggas associated with VLA2. The detection
plateau VLA2 of Zeeman-splitting indicates an absolute weighted magnetic
v = 46kms−1),whichisless thanthetemperaturethatwe
VLA2 field strengths |B | ∼ 700 mG and |B | ∼ 1700 mG.
measured(T ≈ 2500K).Thedifferenceofthetwotemper- VLA1 VLA2
VLA2 We were also able to determine the type of shocks that pump
atures may be due to the turbulence velocity of the gas, which
the water masers associated with both sources. They both are
canbedeterminedbyconsidering∆V.Infact,∆V isrelatedto
i i C-shocks,eventhoughaJ-shockcannotcompletelyberuledout
thethermalandturbulencevelocitiesbytheequation
in the case of VLA2. We suggestVLA1 as the drivingsource
ofthelarge-scalemolecularbipolaroutflow.
∆V = ∆V2 +∆V2 , (7)
i th turb
q
Acknowledgments. We wish to thank an anonymous referee and
andconsideringtheEq.3weareabletoestimatetheturbulence
the editor for making useful suggestions that have highly improved
velocityofthegasifaJ-shockispresent.Thethermalvelocity,
the paper. GS and WHTV acknowledge support by the Deutsche
whichstemsfromthewarmingupofthegasbythecompression
Forschungsgemeinschaft (DFG) through the Emmy Noether Reseach
due to the passage of a shock, for a temperature of 1000 K is grant VL 61/3-1. SC acknowledges support from CONACYT grant
∆VtVhLA2 ≈ 1.6 kms−1, so we have ∆Vturb ≈ 1.9 kms−1. If in- 60581. JMT acknowledges support fromMICINN(Spain) AYA2008-
steadwesupposethattheshockisaC-shockfromthemodelof 06189-C03 (co-funded with FEDER funds) and from Junta de
Kaufman&Neufeld(1996)forT ∼2500K,wegetvVLA2 ∼ Andaluc´ıa(Spain).
VLA2 s
43kms−1, whichiscloseto theexpansionvelocitythatwede-
terminedinthepresentpaper(v = 46kms−1).Bothveloc-
ities are also in goodagreementVwLAit2h the velocityof 43 kms−1 References
measuredbySlyshetal.(2010)fortheOHmasersnearVLA2 AlakovA.V.,SlyshV.I.,PopovM.V.etal.2005,Astron.Lett.,31,375
obtained by multiepoch observations. However, v is close Anderson,N.,&Watson,W.D.,1993,ApJ,407,620
VLA2
tothelimitof45kms−1,whichisconsideredthelimitbetween Baart,E.E.,Cohen,R.J.,Davies,R.D.etal.1986,MNRAS,219,145
Burns,J.O.,Owen,F.N.,&Rudinick,L.1979,AJ,84,1683
the two typesof shocks, so we cannotrule one of them outby
Carrasco-Gonza´lez, C., Rodr´ıguez, L.F.,Torrelles, J.M. et al. 2010, AJ, 139,
onlyconsideringthevelocity.Sinceweobtainwell-orderedhigh 2433
linear polarization fraction and linear polarization vectors, we Crutcher,R.M.1999,ApJ,520,706
have to supposethat the Faradayrotationis atmost ∼20◦; oth- Dickel,J.R.,DickelH.R.&Wilson,W.J.1978,ApJ,223,840
Elitzur,M.,Hollenbach,D.J.,&McKee,C.F.1989,ApJ,346,983
erwise,thevectorsnolongerappeartobealignedwiththelarge
Fish,V.L.&Reid,M.J.2006,ApJS,164,99
structure (i.e., ellipse 2). This condition is only met if the in- Fish,V.L.&Reid,M.J.2007,ApJ,656,952
ternal Faradayrotation is small; in other words, if the shock is Girart,J.M.,Beltra´n,M.T.,Zhang,Q.,etal.2009,Science,324,1408
a C-shock (Kaufman& Neufeld1996). Therefore,we strongly Goldreich.P.,&Keeley,D.A.,1972,ApJ,174,517
Goldreich.P.,Keeley,D.A.,&Kwan,J.Y.,1973,ApJ,179,111
suggestthatthereis aC-shockalso inVLA2evenifa J-shock
Haschick,A.D.,Reid,M.J.,Burke,B.F.,etal.1981,ApJ,244,76
cannotbecompletelyruledout.
Hunter,T.R.,Taylor,G.B.,Felli.M.etal.1994,A&A,284,215
Fromtheequation Hutawarakorn,B.&Cohen,R.J.1996,ASPC,97,532
Hutawarakorn,B.,Cohen,R.J.&Brebner,G.C.,2002,MNRAS,330,349
Bmaser ≈80 nH2 · vs , (8) KImaauif,mHa.n,,HMor.Jiu.c&hiN,Se.u,fDeledg,uDc.hAi,.,S1.,9&96K,AampJe,y4a5,6O,.2250003,ApJ,595,285
[mG] r[108cm−3] [10kms−1]
Kemball,A.J.,Diamond,P.J.andCotton,W.D.1995,A&AS,110,383
Kylafis,N.D.&Norman,C.A.,1987,ApJ,323,346
where n is the pre-shock density and v the velocity of the
H2 s Kylafis,N.D.&Norman,C.A.,1991,ApJ,373,525
shock, we can verify our assumption about the pre-shock den-
Leppa¨nen,K.,Liljestro¨m,T.&Diamond,P.1998,ApJ,507,909
sity and the rough estimation of |BVLA1| and |BVLA2|. Equation Matsumoto,T.&Tomisaka,K.2004,ApJ,616,266