Corresponding author: J. H. Westlake, Johns Hopkins Applied Physics Laboratory, 11100 Johns Hopkins Rd., Laurel, MD 20723, USA. (firstname.lastname@example.org)
 We report on Cassini Ion and Neutral Mass Spectrometer (INMS) observations above Titan's exobase at altitudes of 2225 km to 3034 km. We observe significant densities of CH5+, HCNH+ and C2H5+that require ion-molecule reactions to be produced in the quantities observed. The measured composition and ion velocity (about 0.8–1.5 km/s) suggest that the observed ions must have been created deep inside Titan's ionosphere (below the exobase) and then transported to the detection altitude. Plasma motion from below Titan's exobase to large distances can be driven by a combination of thermal pressure and magnetic forces. The observed outward flows may link the main ionosphere with the more distant wake and provide a source of hydrocarbon ions in the Saturnian system.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 Titan harbors a chemically complex ionosphere that peaks in density at about 1100 km altitude. This ionosphere is primarily produced by solar photoionization on the dayside [Ågren et al., 2009; cf. Cravens et al., 2009a], and either by transport [Cui et al., 2009] or electron impact ionization [Cravens et al., 2006, 2009b] on the nightside. The ionosphere, unprotected by an intrinsic magnetic field, is immersed in Saturn's magnetospheric plasma, likely composed primarily of O+ and H+, flowing past Titan with a velocity of roughly 100 km/s [Young et al., 2005; Thomsen et al., 2010]. The interaction of Saturn's magnetospheric plasma with Titan's atmosphere and ionosphere has been probed over numerous Cassini flybys. Here we report on a fortuitous observation of Titan's ionospheric ions caught up in this complex moon-magnetosphere interaction. One of the goals of this work is to examine how heavy complex ions that must be manufactured below Titan's exobase escape into the surrounding space.
 Prior to the Cassini mission, Keller and Cravens  predicted, using a numerical ionospheric hydrodynamic outflow model for Titan's wake, that ionospheric plasma (with ionospheric composition including C2H5+ and HCNH+) can be transported at speeds of about 1 km/s deep into the wake. The Cassini Ion and Neutral Mass Spectrometer (INMS) data presented in the current paper provide observational evidence that ionospheric plasma is escaping from Titan that could be carried by the flow deep into the wake.
 In this study we report on a serendipitous observation by the Cassini Ion and Neutral Mass Spectrometer (INMS) during the T40 flyby. This observation shows, for the first time, a mass resolved spectrum of the ions flowing away from Titan. Previous observations of ions in Titan's tail region have utilized Cassini Plasma Spectrometer Ion Mass Spectrometer (CAPS-IMS) data [cf.Sittler et al., 2010; Coates et al., 2012], which as studied has a limited ability to resolve compounds of similar mass such as the hydrocarbons observed from Titan. In general the CAPS observations have reported groups of ions with masses near 16 and 28 amu without distinction between adjacent masses within these groups. Other observations have utilized the Radio and Plasma Wave Science Langmuir Probe (RPWS-LP) instrument that measures the electron density and temperature but not the composition of the plasma [cf.Edberg et al., 2011]. We present here the composition and speed of the ions and determine the likely sources of this plasma.
 The T40 flyby took place on the 5th of January 2008. During this flyby Cassini was inbound on a sunlit portion of Titan's tail and outbound on the anti-Saturn facing flank. The flyby geometry is shown inFigure 1 along with count rates for selected masses. The INMS instrument was pointed in the spacecraft ram direction and able to observe Titan's thermal ions between about 400 seconds before and 600 seconds after the closest approach (1014 km).
 As illustrated in Figure 1, the INMS measured high ion count rates while Cassini is in the main ionosphere below the exobase. The count rates then decrease with altitude overall but remain relatively high out to an altitude of about 3000 km. The gaps in the coverage seen in this figure are due to scheduled pulse height discrimination (PHD) scans and transmission scans. When the instrument switches back into the science mode from the PHD scan significant counts are observed in the mass 2, 15, 16, 17, 28, 29, 30, and 39 mass bins. This is the first flyby in which the INMS has observed significant ion counts above about 1600 km.
 The INMS in its so-called “open source mode” [Waite et al., 2004] detects ions within a narrow (3° FWHM) field of view (FOV). Ions incident on the INMS pass through a set of focusing and beam shaping lenses and then are bent along a 90° angle by a quadrupole switching lens which then focuses the ions on the quadrupole analyzer section. The quadrupole switching lens also acts as an energy filter, only allowing ions with energies that are basically thermal into the analyzer. The energy allowed by the quadrupole switching lens is established prior to the flyby by setting its voltages based on the predicted flyby velocity. When the INMS is not pointed in the spacecraft ram direction or a substantial spacecraft potential is present then it is possible to observe ions with velocities greater than the thermal velocity due to the mismatch in the actual ion velocity parallel to the INMS entrance aperture and the expected ion velocity. Ion velocities are derived by combining the spacecraft velocity with the boresight angle with respect to the direction of the spacecraft motion (also called the spacecraft ram direction) and correcting for the spacecraft potential. Enhanced counts are observed at 2008-005T21:41:30 during the T40 flyby, at high altitudes the peak observed count rate takes place when the INMS boresight angle is at a 9° angle with respect to the spacecraft ram. In order for the observed ions to enter the INMS ion source at this large of an angle, they must have a significant velocity transverse to the direction of motion of the spacecraft. The settings on the quadrupole switching lens were such that ions with energies of 0.6, 3.3, 3.6, 3.8, 6.0, 6.2, and 8.3 eV were transmitted for masses 2, 15, 16, 17, 28, 29, and 39 respectively. The velocity of the ions observed is calculated from the observed energy usingequation (1).
where EOBS is the observed energy, m is the mass of the ion, vSC is the spacecraft velocity retrieved from the Cassini SPICE kernels, vion is the velocity of the ion, and ϕSCis the spacecraft potential which was observed to be −1.2 V by the RPWS-LP. For the observation of enhanced counts at 21:41:30 the ionospheric ions are observed with speeds of 0.8–1.5 km/s.Table 1 gives the mass, species and observed ion speed in the Titan frame for the observation at 21:41:30. H2+ ions are observed at a speed of 7.12 km/s. The magnetic field is nearly perpendicular to the observed ion velocity, indicating that this outflow could be field induced. However, the magnitude of the magnetic field is relatively small (∼3.4 nT) at this time and the forces associated with this field are likely to be comparable to the pressure forces in this region [cf. Cravens et al., 2010; Ulusen et al., 2010]. Since the FOV is relatively narrow in both the energy and angular dimensions the instantaneous measurement of the ion's speed has a small uncertainty (∼5%), however due to the limited coverage of velocity space the uncertainty in the bulk velocity of the ions is significantly larger. The ion density observed by the INMS near 3000 km was about 10 cm−3, which is roughly 30% of the total electron density observed by the RPWS-LP. The T40 flyby was novel in the INMS dataset because it presented the ideal observation scenario of a rolling spacecraft resulting in capture of ions with non-thermal velocities, significant ion densities above the exobase, low spacecraft potential, and relatively low ion bulk velocity that allowed the INMS to observe these ions.
Table 1. The Mass, Species and Observed Speed in the Titan Frame of the Ions Observed at 2863 km Altitude at 2008-005T21:41:30
Observed Speed (km/s)
Figure 2 shows the combined mass spectrum of the ions observed from 2225 km to 3034 km with the densities summed over this altitude range. Peaks are observed at mass 2, 15, 16, 17, 27, 28, and 29 Da, while less distinct peaks are observed at mass 3, 14, 18, 26, 30, 39, and 41 Da. The composition observed in this spectrum is clearly similar to the composition observed in Titan's ionosphere below the exobase. Mass 2 is attributed to H2+, freshly ionized or transported from Titan's ionosphere. Masses 15, 16, and 17 are identified as CH3+, CH4+, and CH5+, which are the ionization and chemical products of CH4. When CH4 is ionized it produces in various amounts: CH4+, CH3+, CH2+, and H+. However, the CH3+ in Titan's ionosphere is primarily produced through the reaction of N+ and N2+ with CH4. CH5+ is produced through the reaction of CH4+ with neutral methane. We identify HCNH+ and C2H5+ as the constituents at masses 28 and 29 respectively. Some amount of mass 28 is likely to be associated with N2+, however it is impossible to discern this signal from that of HCNH+due to the limited mass-resolution of the INMS. We note that the N2+ signal will be closely related to the mass 14 (N+) signal, both of which are produced through the photoionization of neutral N2. Given the low count rate in mass 14 the N2+ contribution to mass 28 will be at least an order of magnitude less than that which is measured, indicating that the majority of the counts in mass 28 are due to HCNH+. HCNH+ is produced in Titan's ionosphere primarily by the reaction of C2H5+ with HCN. In the main ionosphere below the exobase CH3+is rapidly lost through ion-molecule reactions with CH4 producing C2H5+. The ions CH5+, C2H5+, and HCNH+ are all primarily lost in the ionosphere through the relatively slow electron recombination process and therefore have large chemical timescales [Cravens et al., 2010]. We also identify the signals at mass 39 and 41 with the C3H3+ and C3H5+, which are hydrocarbons observed in Titan's ionosphere. The chemical identification in this mass spectrum is consistent with that indicated by exo-ionospheric models [cf.Keller and Cravens, 1994].
 The enhancement in the ion signal at high altitudes on T40 could be related to a change in the INMS FOV direction, or it could be an enhancement in the ion density. The RPWS-LP indicates that the spacecraft potential during this time period is nearly constant at −1.2 V. The altitude profile of this ion observation is shown inFigure 3. The high altitude observations in the INMS instrument show a steadily declining count rate for the Mass 28 and 29 channels, and a nearly flat, reduced count rate for the mass 17 channel and the total counts until roughly 2700 km altitude where the count rate begins to increase. The count rate increases until it peaks at about 400 counts/IP at an altitude of 2863 km. There is also a roughly 20% enhancement in the RPWS-LP electron density coincident with this peak. This enhancement indicates that this parcel of plasma is slightly denser than its surroundings. The INMS-observed count rate enhancement, however, is significantly larger than the RPWS-LP electron density enhancement indicating that the change in boresight angle is contributing to the density enhancement. We note that the INMS signal rapidly declines as the INMS boresight angle further increases causing the INMS FOV to move away from the ion distribution. Had the spacecraft been continuously pointed in the ram direction during this time the INMS would likely have completely missed this parcel of plasma. We, therefore, attribute the signal enhancement to the INMS field of view moving such that more of the observed plasmas' distribution is captured.
3. Discussion and Conclusions
 We hypothesize that the ions observed above 2225 km are produced in Titan's ionosphere below the exobase and then transported to their observed position. It is possible in principle that these ions are created at the observation point, but we discuss next why this scenario does not seem likely.
 The observations take place 725 km above the exobase where the mean free path of the ions is much larger than the scale height of the atmosphere. At this altitude N2 has become a smaller contributor to the neutral signal observed, while H2 and CH4 have become more dominant contributors due to their lower mass and greater scale height [De La Haye et al., 2007; Cui et al., 2008; Magee et al., 2009]. Cravens et al. [2009a, 2009b, 2010]showed that near 1500 km the time constants for vertical and horizontal ion transport are shorter than the chemical reaction time constants. For the ions that are the terminal steps in the ion-molecule chemistry such as CH5+, HCNH+, and C2H5+, the time to generate these ions in this region is of the order of 105 seconds, while at a speed of 0.8–1.5 km/s they would be transported into the region where INMS observes enhanced densities from the exobase much more quickly (∼103 seconds assuming a constant velocity and radial flow). Therefore, these ions are most likely not created at the location of the INMS observation, and must have been transported from near or below Titan's exobase. We further support this conclusion with the mass spectrum observed at the exobase height with the mass spectrum observed between 2225 and 3034 km shown in Figure 3. The two spectra are similar indicating that the ions observed above the exobase were likely produced near the exobase.
 Given the long chemical time scales at the observation altitude, ions created locally and then picked-up by the flow, will reflect the primary ionization process. Since the lower exosphere is dominated by CH4 and H2 the ions produced would be primarily CH4+ and H2+. Additionally since one or two ion-molecule reactions are required to produce the CH5+, C2H5+ and HCNH+ ions they are not likely to be produced locally. This is especially true for HCNH+, which requires significant densities of CH5+, C2H5+, and HCN to be produced. These conditions are only satisfied in Titan's ionosphere below the exobase.
Wellbrock et al.  and Coates et al.  studied Titan's wake region finding several cases of ionospheric plasma downstream of Titan. They identify regions of ionospheric plasma by the identification of photoelectron peaks at 24.1 eV created by solar (He II line) photons ionizing N2 in Titan's ionosphere. Wellbrock et al.  specifically studied the T40 flyby prior to closest approach. We note that 24.1 eV photoelectrons are observed far beyond closest approach and during the INMS observation presented here, further confirming the ionospheric origin of this plasma.
 The INMS ion signal is observed to peak at a 9° angle with respect to the direction of spacecraft motion suggesting that the ions have a significant velocity vector compared with the nearly stationary ionospheric ions. This further supports that the ions are not produced at the location of the INMS observation, as they must be accelerated after their production to obtain the observed velocity. We therefore conclude that these ions must have been produced below Titan's exobase and transported to the location of the INMS observation.
 In the ionosphere plasma flows in response to forcings related to thermal pressure gradients, magnetic forces, gravity, and ion-neutral collisions all of which can be described to be in balance in the MHD momentum equation.Cravens et al.  and Ulusen et al.  analyzed empirically and through comparison with model results the pressure terms of the MHD momentum equation to assess the dominant terms for various regions and flybys. It is clear from these two studies that a variety of situations are possible that would drive ionospheric plasma flows. Edberg et al. suggested from RPWS-LP measurements that this region exhibited a structured ionospheric outflow. The INMS-observed outflowing plasma is in the same region as the structured outflow ofEdberg et al. , but is further upstream. The INMS observation (T40) occurred more than one year prior to the Edberg et al.  observation (T55–T59) indicating that this region may be a persistent ion outflow source as suggested by the hybrid models of Titan [cf. Simon et al., 2007; Ledvina et al., 2012]. Additionally, there is evidence of low velocity (5–10 km s−1) flows on the anti-Saturn flank of Titan from the Voyager 1 flyby in 1980 [Hartle et al., 1982], however with the Voyager 1 plasma instrument the authors were unable to determine the exact nature of the composition other than it contained a heavy component likely originating in Titan's ionosphere [Hartle, 1985; Sittler et al., 2005]. We suggest that this particular observation is most likely a persistent pressure and field driven outward flow that links the main ionosphere with the more distant wake. Regardless of the process responsible for the ionospheric plasma outflow it is clear from this observation that material born in Titan's main ionosphere is being transported to the more distant wake and possibly into Saturn's magnetosphere.
 The Editor thanks two anonymous reviewers for assisting in the evaluation of this paper.