Loitering Outflow FeLoBAL Quasars

• Loitering outflow FeLoBAL quasars are rare systems where dense, low-ionization gas lingers either near the nucleus (parsec scales) or persists at host-galaxy (kpc) distances.
• Spectroscopic analyses using techniques like SimBAL and photoionization modeling reveal key parameters—radius, velocity, column density, and covering fraction—that distinguish nuclear loiterers from extended absorbers.
• These systems offer practical insights into outflow dynamics and AGN feedback, with some regimes showing kinetic powers in the feedback-relevant range while others remain dynamically subdued.

Searching arXiv for recent and foundational papers on FeLoBAL quasars and “loitering” outflows. Search 1: "FeLoBAL loitering outflow SimBAL 2022". Loitering outflow FeLoBAL quasars are FeLoBAL systems in which dense, low-ionization BAL gas is inferred either to remain very close to the nucleus with unusually low velocity or to persist at kiloparsec scales after leaving the launching region. FeLoBALs themselves are BAL quasars with low-ionization absorption and prominent Fe II and/or Fe III absorption, and they are rare: roughly $\sim 1\%$ of the BAL population and $\sim 0.3\%$ of quasars in flux-limited samples (McGraw et al., 2013, McGraw et al., 2015). In one well-defined usage, Choi et al. identify a “loitering outflow” class with $\log R<1$ pc, $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$, and $v_{\rm width,FeII*}<2000\ {\rm km\,s^{-1}}$ (Choi et al., 2022). In another usage, SDSS J1321−0041 and SDSS J0932+0840 are described as long-lived FeLoBAL outflows persisting at $\sim$kpc radii, with $R=2.5^{+1.0}_{-1.4}$ kpc and $R=0.7^{+0.9}_{-0.4}$ kpc, respectively (Byun et al., 2023, Sharma et al., 2024). This suggests that the term is used for two related but distinct regimes: torus-scale stalled or slowly moving absorbers, and large-radius persistent FeLoBAL phases.

1. Terminology and observational scope

FeLoBAL quasars are distinguished by broad absorption in low-ionization ions such as Mg II and Al III together with Fe II and often Fe III, in addition to the high-ionization BAL species typical of BAL quasars (McGraw et al., 2013). The “loitering outflow” label is therefore a refinement within an already extreme BAL subclass, not a replacement for the FeLoBAL definition.

The recent literature uses the term in two main ways. Choi et al. reserve it for compact nuclear outflows characterized by small radii, low line-of-sight velocities, and high column density (Choi et al., 2022). By contrast, the analyses of SDSS J1321−0041 and SDSS J0932+0840 use “loitering” for dense, low-ionization gas that has reached host-galaxy scales yet still persists with measurable column density and modest velocity (Byun et al., 2023, Sharma et al., 2024).

Usage in the literature	Characteristic properties	Representative analyses
Nuclear loiterers	$\log R<1$ pc; $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$; high $\sim 0.3\%$0, $\sim 0.3\%$1, and $\sim 0.3\%$2	SimBAL sample (Choi et al., 2022)
Large-radius persistent FeLoBALs	$\sim 0.3\%$3–2.5 kpc; low or intermediate velocity; dense low-ionization gas at host-galaxy scales	SDSS J0932+0840 (Sharma et al., 2024), SDSS J1321−0041 (Byun et al., 2023)

This terminological duality is central. A loitering FeLoBAL is not defined by one radius alone; it is defined by a dynamical state in which unusually dense, low-ionization gas remains observable for long enough, and at modest enough velocity relative to other BAL winds, that its structure and environmental coupling can be measured.

2. Spectroscopic signatures and diagnostic methods

Loitering FeLoBALs are identified through complex low-ionization spectra. Commonly detected ions include Mg II, Al II, Al III, Si II, Fe II, Fe III, and, in the best cases, multiple excited levels of Fe II and Si II (Byun et al., 2023). In the SimBAL loitering objects, strong He I* $\sim 0.3\%$4, $\sim 0.3\%$5, Balmer absorption, and overlapping Fe II troughs are also prominent (Choi et al., 2022). Balmer absorption is especially informative because it requires very high total column, high density, and a sustained population in $\sim 0.3\%$6 (Leighly et al., 9 Sep 2025).

The standard quantitative workflow combines apparent-optical-depth column measurements with photoionization and level-population modeling. For normalized spectra,

$\sim 0.3\%$7

and integrated ionic columns are obtained from

$\sim 0.3\%$8

after conversion from optical depth in velocity space (Byun et al., 2023). Cloudy grids then constrain $\sim 0.3\%$9, while excited-state ratios constrain density through collisional-equilibrium calculations with CHIANTI or related atomic data (Sharma et al., 2024).

Partial covering is pervasive. In the Balmer-absorption survey of fourteen FeLoBALQs, eight objects require partial covering to reconcile the observed Balmer optical-depth ratios, and several spectra are best fit when the continuum is fully covered while the emission-line region is only partially covered (Leighly et al., 9 Sep 2025). FBQS J1151+3822 provides a concrete example: line-of-sight covering fractions are $\log R<1$0 for strong Fe II, $\log R<1$1 for He I*, and $\log R<1$2 for Mg II (Lucy et al., 2014). Such values show that loitering FeLoBALs are not homogeneous screens but structured, multiphase absorbers.

Variability is the other major diagnostic. In a twelve-object FeLoBAL monitoring program, robust BAL variability at $\log R<1$3 was detected in 3/12 sources over rest-frame timescales of $\log R<1$4 to $\log R<1$5 yr, in Fe II, Fe II*, Mg II, Mg I, and Ni II* troughs (McGraw et al., 2015). These changes can be interpreted either as ionization changes or as gas traversing the line of sight, and both interpretations are used in the literature depending on the object and the degree of saturation.

3. Nuclear loiterers at parsec scales

The clearest formal definition of loitering FeLoBALs comes from the SimBAL analysis of 50 low-redshift FeLoBAL quasars. Eleven BAL components in eleven objects satisfy $\log R<1$6, $\log R<1$7, and $\log R<1$8 (Choi et al., 2022). These absorbers occupy the high-ionization, high-density end of the FeLoBAL distribution, with typical $\log R<1$9 to $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$0, $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$1–7.8 $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$2, and covering-fraction-corrected $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$3–23.5. Most lie at $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$4–3 pc, inside the dusty torus region inferred for these luminosities, and they often require modified partial-covering models in which the BLR is unabsorbed while only part of the continuum is covered (Choi et al., 2022).

Variability studies independently place some FeLoBAL gas at similar scales. McGraw et al. report a representative upper limit of $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$5 pc from a typical crossing timescale $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$6 yr and transverse speeds $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$7 (McGraw et al., 2013). A later multi-epoch study of the same general class constrains three variable FeLoBALs to $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$8, 7, and 60 pc under a transverse-motion interpretation (McGraw et al., 2015).

Detailed single-object analyses sharpen the physical picture. FBQS J1151+3822 has $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$9, $v_{\rm width,FeII*}<2000\ {\rm km\,s^{-1}}$0, $v_{\rm width,FeII*}<2000\ {\rm km\,s^{-1}}$1, and $v_{\rm width,FeII*}<2000\ {\rm km\,s^{-1}}$2 to 127 pc; its newly emerged narrow lines at $v_{\rm width,FeII*}<2000\ {\rm km\,s^{-1}}$3 are interpreted as dense cores condensing inside the main outflow (Lucy et al., 2014). FBQS J1408+3054 underwent dramatic FeLoBAL weakening, and the crossing interpretation implies a transverse velocity between 2600 and $v_{\rm width,FeII*}<2000\ {\rm km\,s^{-1}}$4 and a location between approximately 5800 and 46,000 Schwarzschild radii, or 1.7 to 14 pc (Hall et al., 2010). Mrk 231 adds a geometrically complex nuclear case: its FeLoBAL region is placed within $v_{\rm width,FeII*}<2000\ {\rm km\,s^{-1}}$5–10 pc, with strong optical FeLoBAL troughs, no unambiguous far-UV BALs, and a faint Ly$v_{\rm width,FeII*}<2000\ {\rm km\,s^{-1}}$6 line centered at $v_{\rm width,FeII*}<2000\ {\rm km\,s^{-1}}$7, requiring a patchy, partially covering BAL region close to the engine (Veilleux et al., 2012).

These objects do not imply a static absorber. Rather, they show that a high-column, low-ionization outflow can remain at torus or inner-nuclear scales for years to centuries while evolving through covering-factor changes, ionization changes, and internal substructure.

4. Kpc-scale persistent FeLoBALs

FeLoBALs can also loiter at host-galaxy radii. SDSS J1321−0041 is a high-resolution UVES case with strong Fe II, Si II, and C II BALs. Its best-fit parameters are $v_{\rm width,FeII*}<2000\ {\rm km\,s^{-1}}$8, $v_{\rm width,FeII*}<2000\ {\rm km\,s^{-1}}$9, $\sim$0, and $\sim$1 kpc, with $\sim$2 and $\sim$3 (Byun et al., 2023). The authors explicitly describe it as dense gas that is no longer near the launching region but persists at large radii.

SDSS J0932+0840 presents a slower kpc-scale case. Its narrow S2 component at $\sim$4 has $\sim$5, $\sim$6, $\sim$7, $\sim$8, and $\sim$9 kpc (Sharma et al., 2024). Its dynamical time is

$R=2.5^{+1.0}_{-1.4}$0

and the absorber is interpreted as a kpc-scale, relatively slow FeLoBAL component consistent with long-lived gas persisting in the inner host galaxy (Sharma et al., 2024).

Theoretical work makes this large-radius regime physically plausible. Faucher-Giguère, Quataert, and Murray argue that the extreme $R=2.5^{+1.0}_{-1.4}$1, kpc-scale distances, and short cloud-destruction times implied by some FeLoBAL analyses are more naturally explained if the observed FeLoBAL absorbers are formed in situ in radiative shocks when a quasar blast wave impacts moderately dense interstellar clumps (Faucher-Giguere et al., 2011). Bautista and Dunn then show that year-scale variability in the ionizing continuum can drive supersonic ionization fronts, cooling fronts, and fragmentation in kpc-scale FeLoBAL clouds, and that the measured velocity differences among components in QSO 2359−1241 and SDSS J0318−0600 can be reproduced if strong magnetic fields of $R=2.5^{+1.0}_{-1.4}$2 mG are present within the clouds (Bautista et al., 2010).

This suggests that large-radius loitering does not imply a monolithic shell. It can instead denote dense, cool clumps embedded in a larger shocked flow, with the observed FeLoBAL phase tracing only one part of a multi-phase outflow.

5. Energetics and feedback relevance

The feedback significance of loitering FeLoBALs is highly heterogeneous. Many studies estimate energetics with a partial-shell relation such as

$R=2.5^{+1.0}_{-1.4}$3

with $R=2.5^{+1.0}_{-1.4}$4 often set to 0.2 and $R=2.5^{+1.0}_{-1.4}$5 (Byun et al., 2023, Sharma et al., 2024). Whether a loitering FeLoBAL is feedback-relevant then depends on $R=2.5^{+1.0}_{-1.4}$6, $R=2.5^{+1.0}_{-1.4}$7, and especially $R=2.5^{+1.0}_{-1.4}$8.

At the weak end, the SimBAL loitering class is generally not feedback-relevant. Its mass outflow rates are frequently $R=2.5^{+1.0}_{-1.4}$9, kinetic luminosities are typically $R=0.7^{+0.9}_{-0.4}$0–43, and $R=0.7^{+0.9}_{-0.4}$1 for virtually all loitering outflows (Choi et al., 2022). SDSS J0932+0840 is similar: $R=0.7^{+0.9}_{-0.4}$2, $R=0.7^{+0.9}_{-0.4}$3, and $R=0.7^{+0.9}_{-0.4}$4, explicitly below the level required for significant AGN feedback (Sharma et al., 2024).

At the strong end, SDSS J1321−0041 reaches

$R=0.7^{+0.9}_{-0.4}$5

above or near canonical $R=0.7^{+0.9}_{-0.4}$6–5% thresholds, and is concluded to contribute to AGN feedback (Byun et al., 2023). FBQS J1151+3822 spans $R=0.7^{+0.9}_{-0.4}$7 to $R=0.7^{+0.9}_{-0.4}$8, showing that parsec- to tens-of-parsecs FeLoBALs can also enter the feedback-relevant regime (Lucy et al., 2014).

Integrated IR studies support a broader feedback connection. In a sample of 31 FeLoBAL QSOs, all objects have $R=0.7^{+0.9}_{-0.4}$9, the AGN contributes 75% of the total IR luminosity on average, and the Mg II absorption-strength parameter $\log R<1$0 anticorrelates with the starburst fraction $\log R<1$1: all systems with $\log R<1$2 have $\log R<1$3, leading to the conclusion that radiatively driven outflows curtail obscured star formation to less than $\log R<1$4 of the total IR luminosity (Farrah et al., 2011).

X-ray studies add an inner, highly absorbed phase. Suzaku observations of three FeLoBALs imply intrinsic columns $\log R<1$5, $\log R<1$6, and $\log R<1$7, and for plausible radii and velocities the inferred kinetic feedback efficiency is $\log R<1$8–6.9% (Morabito et al., 2010). The observational record therefore separates two issues that are often conflated: loitering is a dynamical description, whereas feedback relevance is an energetic one.

6. Host galaxies, controversies, and broader connections

The host-galaxy literature weakens a simple equation between FeLoBALs and recent major mergers. HST F160W imaging of ten FeLoBAL hosts at $\log R<1$9 finds only 1/10 clear major merger in the full sample, 3/10 disturbed hosts, and no strong enhancement relative to luminous blue quasars; the conclusion is that FeLoBAL hosts are not dominated by recent major mergers (Villforth et al., 2018). This does not exclude an evolutionary interpretation in individual objects, but it argues against identifying the entire class with a single post-merger blowout stage.

The location problem is equally non-unique. Variability studies place some FeLoBAL gas within $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$0 pc or within $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$1, 7, and 60 pc for individual variable systems (McGraw et al., 2013, McGraw et al., 2015). Detailed photoionization analyses place other systems at $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$2–2.5 kpc (Byun et al., 2023, Sharma et al., 2024). Hall et al. therefore conclude from FBQS J1408+3054 that not all FeLoBAL outflows can be associated with galaxy-scale outflows in ultraluminous infrared galaxies transitioning to unobscured quasars (Hall et al., 2010). Lucy et al. add that the apparent rarity of small-radius, high-density FeLoBALs may be a symptom of selection bias in studies using density-sensitive lines (Lucy et al., 2014). This suggests that FeLoBALs are intrinsically multi-scale.

Another active issue is whether low-ionization emission can arise in the wind itself. In variable FeLoBALs with vanishing absorption, Mg II and UV Fe II emission are observed to be blueshifted by thousands of $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$3 relative to H$|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$4, and the proposed interpretation is that these lines arise in the outflowing winds normally seen only in absorption (Rafiee et al., 2016). Such cases blur the distinction between absorber and emitter and reinforce the picture of a stratified, partly emitting BAL outflow.

Recent Balmer-absorption work extends the loitering framework further. In a sample of fourteen FeLoBALQs with Balmer absorption, nearly all have the low outflow speeds characterizing loitering outflow FeLoBAL quasars, and the authors note suggestive similarities to Little Red Dots, including Balmer absorption, steep reddening with a scattered blue continuum, weak hot dust emission, and X-ray weakness (Leighly et al., 9 Sep 2025). A plausible implication is that loitering FeLoBAL phenomenology is not confined to one redshift or one observational selection, but may instead mark a recurrent configuration of dense, low-ionization gas near or within the broader AGN outflow cycle.

Loitering outflow FeLoBAL quasars therefore denote a physically diverse but observationally coherent set of systems in which dense, metal-rich, low-ionization BAL gas persists longer, and at lower effective velocity, than standard caricatures of a rapidly accelerating disk wind would suggest. Some are torus-scale, high-density, low-power nuclear absorbers; some are parsec-scale winds caught through variability; some are host-galaxy-scale FeLoBAL phases with dynamical times of order $|v_{\rm off,FeII*}|<2000\ {\rm km\,s^{-1}}$5 yr and, in the most extreme cases, feedback-level kinetic power. The main encyclopedic lesson of the current literature is not uniformity but stratification: FeLoBAL outflows occupy parsec to kiloparsec scales, can alternate between absorption and emission signatures, and can be either weakly coupled or strongly coupled to AGN feedback depending on where, and in what phase, the loitering gas is observed.