SPITZER PHOTOMETRY OF WISE-SELECTED BROWN DWARF AND HYPER-LUMINOUS INFRARED GALAXY CANDIDATES

In the historical nomenclature of stellar classification, astronomers have classified stars according to their spectra, finally settling on the following spectral types, OBAFGKM, with O representing the hottest stars and M the coolest (Morgan et al. 1943). With the discovery of numerous (sub)stellar objects cooler than the coldest M dwarfs, the need for new spectral types became evident. To describe and characterize this new population, two new spectral types were introduced to the stellar taxonomy, L and T dwarfs (see, e.g., Kirkpatrick 2005 for an in depth review), while a new type, "Y," is used for even colder objects, as suggested by Kirkpatrick et al. (1999) and Kirkpatrick (2000).

These low-mass, low-temperature astrophysical sources are considered to be the lowest-mass population of star-like objects. Those with masses below ∼0.075 M_☉ are referred to as "brown dwarfs"(Kumar 1963; Hayashi & Nakano 1963) and have core temperatures insufficient to sustain hydrogen fusion. Brown dwarfs are important objects for several reasons. They provide benchmarks with which to test our theories of gas chemistry and cloud formation in low-temperature atmospheres (Burgasser 2008). As analogs of extra-solar planets, brown dwarfs are easier to study because they are typically isolated. They represent in situ time capsules of star formation, since their mass is never recycled back into the interstellar medium, therefore preserving information on metallicity enrichment over the lifetime of the Galaxy (Burgasser 2008). They provide important constraints on both the functional form of the stellar mass function and the low-mass limit of star formation (Kirkpatrick et al. 2012; Metchev et al. 2008; Reylé et al. 2010; Burningham et al. 2010). Though predicted to exist in the 1960s, brown dwarfs remained elusive for decades until surveys such as the Two Micron All-Sky Survey (2MASS; Skrutskie et al. 2006), the Sloan Digital Sky Survey (York et al. 2000), and the Deep Near-Infrared Survey of the Southern Sky (DENIS; Epchtein et al. 1997) revealed them in large numbers.

Recently, researchers have shifted their focus toward the coldest brown dwarfs, which are one of the two primary science objectives of theWide-field Infrared Survey Explorer (WISE; Wright et al. 2010). WISE launched on 2009 December 14 and began surveying the sky on 2010 January 7 at wavelengths of 3.4, 4.6, 12, and 22 μm, hereafter referred to as W1, W2, W3, and W4, respectively. It completed its first full pass of the sky on 2010 July 17 and the WISE All-Sky Data Release was issued on 2012 March 14.¹³ The survey reaches 5σ point source sensitivities in unconfused regions to better than 0.08, 0.11, 1 and 6 mJy, respectively (Wright et al. 2010). The W1 and W2 filters were specifically designed to probe the deep 3.3 μm CH₄ absorption band in the spectra of brown dwarfs and the region relatively free of opacity at ∼4.6 μm, making their W1 − W2 colors very red and allowing cool brown dwarfs to be readily identifiable.

Hundreds of potentially ultra-cool brown dwarf candidates have been identified using the WISE All-Sky Data. A detailed description of this search is described in Kirkpatrick et al. (2011, 2012), where, along with Mainzer et al. (2011), Wright et al. (2011), Burgasser et al. (2011), Cushing et al. (2011), Gelino et al. (2011), and G. Mace et al. (in preparation), follow-up and preliminary analyses have been presented. With the relatively shallow depth of WISE, a majority of these ultra-cool brown dwarf candidates either have very faint W1 detections or were simply not detected in W1, resulting in W1 − W2 color limits. The Spitzer Infrared Array Camera (IRAC) 3.6 and 4.5 μm bands (Fazio et al. 2004), sometimes referred to as ch1 and ch2, are very similar to the W1 and W2 filters in WISE. This makes IRAC an excellent instrument for deeper follow-up observations (see Figure 2 in Mainzer et al. 2011).

In this paper, we present the photometric properties from IRAC follow-up of a large sample (∼1500) of WISE-selected brown dwarf candidates (see Section 4.1). Of these, 182 have been spectroscopically classified, with classifications ranging from a single M dwarf to 12 examples of the latest and coldest spectral type, Y (Kirkpatrick 2008; Cushing et al. 2011; Kirkpatrick et al. 2012). We also present new, ultra-cool brown dwarf candidates discovered using the Spitzer data alone in an effort to identify fainter, widely separated companions to the WISE sources. As a control sample, we also identify field brown dwarf candidates from a Spitzer campaign to follow-up WISE-selected hyper-luminous infrared galaxy (HyLIRG) candidates (see Section 4.2), as this enables us to assess whether the Spitzer-selected brown dwarf companion candidates are more likely to be truly associated with the WISE-selected brown dwarf candidates. We release a catalog for 906 WISE-selected HyLIRG candidates. In addition to these two extremely rare populations of astrophysical sources, we release a photometric catalog of ∼7.33 × 10⁵ sources detected in the 1000's of Spitzer Astronomical Observation Requests (AORs) in these follow-up campaigns. In Section 2, we briefly describe the observations. We discuss the photometry and source detection in Section 3. In Section 4, we describe the catalogs presented in this release. We discuss the new brown dwarf candidates discovered by this analysis in Section 5 and summarize the paper in Section 6. All magnitudes are given in the Vega system unless otherwise noted.

Warm Spitzer observations were carried out at 3.6 and 4.5 μm under Spitzer programs 70062, 70162, 80033, and 80109. Both the brown dwarf and HyLIRG candidate field observations used a 5-point dither pattern with 30 s exposures per pointing in each IRAC band. IRAC has a 5' × 5' field of view with 12 pixels. For this analysis, we utilized the post-BCD Spitzer pipeline images, which have been resampled onto 06 pixels. The AORs were executed between 2010 June and 2012 May and comprise a total of 1564 brown dwarf and 919 extragalactic follow-up observations. These observations have been described in detail in Kirkpatrick et al. (2011) and Eisenhardt et al. (2012). The images for this analysis were reduced with two different versions of the Spitzer reduction pipeline, 18.18 and 19.1, with the only change between the two versions being an update to the masking process which includes knowledge of latents going back to previous AOR's. None of these changes affect our reduction and processing techniques.

Source detection and photometry were carried out using SExtractor (Bertin & Arnouts 1996). We constructed a ch2 selected catalog by using the dual-image mode capabilities of SExtractor. In dual-image mode, sources are detected and their centroids and apertures are defined in one image and subsequently the photometry is measured in another image using those predetermined apertures and source centers. As in Eisenhardt et al. (2010), the choice of using ch2 to construct this catalog was motivated by the fact that cool brown dwarfs are generally brighter in ch2 than ch1. Photometry was measured in 60 diameter apertures and corrected using the calibration values determined by the IRAC instrument team, −0.133 and −0.113 for ch1 and ch2, respectively. We use the IRAC zero points provided by the Spitzer Science Center (SSC), which are 18.789 and 18.316 Vega magnitudes for ch1 and ch2, respectively. SExtractor was configured to define a source as a set of 20 or more connected pixels, each lying 1.0σ above the background. We utilize the post-BCD coverage maps to provide an exposure flag (COVCH#) for all sources in the full catalog, with COVCH# equal to the number of frames going into the post-BCD stack at that position. We also provide the SExtractor parameter CLASS_STAR to aid in separating point sources from extended sources. We compared the SExtractor aperture magnitudes for our brown dwarf candidates to an independent measurement of aperture photometry as presented in Kirkpatrick et al. (2011, 2012). This independent photometry tool was written in IDL using both public scripts from IDL Astronomy User's Library¹⁴ and proprietary code created specifically for this task. In short, this IDL code centroids on the known position of the brown dwarf candidate, obtains the aperture photometry of the source, and applies the appropriate aperture corrections as provided by the SSC. We find a mean offset Δ ch2 ∼ 0.001 mag and a standard deviation sigma (Δ ch2) ∼ 0.04 mag between the two codes, indicating that both are behaving reasonably.

There are three separate catalogs presented in this paper. The first (WISE-Spitzer_BD_photometric_catalog_V1.0.fits in the online journal), described in Section 4.1 and Table 1, is a list of Spitzer and WISE photometry for 1510 WISE-selected brown dwarf candidates targeted in Spitzer programs 70062 and 80109. The second (WISE-Spitzer_EXGAL_photometric_catalog_V1.0.fits in the online journal), described in Section 4.2 and Table 2, is a list of Spitzer and WISE photometry for 906 WISE-selected HyLIRG candidates (hereafter "W1W2-dropouts"; Eisenhardt et al. 2012) that were the targets for Spitzer programs 70162 and 80033. The third (WISE-Spitzer_combined_catalog_V1.0.fits in the online journal), described in Section 4.3 and Table 3, is a list of all (COV ⩾ 3) sources found on the Spitzer IRAC images for the three programs above.

4.1. The Brown Dwarf Catalog

Brown dwarf candidates were selected from the WISE source databases using the color criteria discussed in Kirkpatrick et al. (2011, 2012). To summarize, our main selection used a color cut chosen to select cold brown dwarfs with types ⩾T5. A color cut of W1 − W2 > 1.5 mag was used with WISE internal source lists from early processing runs and W1 − W2 > 2.0 mag was used for later processing, specifically using the combination of the WISE All-Sky Source Catalog and WISE All-Sky Reject Table that are contained in the WISE All-Sky Data Release (Cutri et al. 2012). We also performed another cut with a much more relaxed color criterion (W1 − W2 > 0.4 mag) to select bright (W2 S/N > 30), nearby brown dwarfs with types ⩾L5. Other constraints, including the lack of a positional match in earlier-epoch all-sky surveys (2MASS, DSS1, DSS2), cuts on the W2 − W3 color, stipulations on the reduced χ² value from point-spread function photometric fitting, etc., were also imposed, as discussed in Kirkpatrick et al. (2011, 2012).

Most of the selected sources were scheduled for Spitzer ch1 and ch2 observations as part of programs 70062 and 80109. As of late 2012 May, a total of 1564 unique candidates had been observed in both IRAC bands. We were not able to identify a Spitzer counterpart for 54 of these candidates, which were therefore deemed to be spurious sources in WISE. Further inspection revealed that these spurious sources fell into two categories: (1) cosmic rays that bled through to the final images in the WISE co-addition process due to multiple cosmic-ray hits and low frame coverage and (2) WISE sources within nebulous patches (or nebulae themselves) rather than nearby brown dwarfs. For the remaining 1510 candidates, however, Spitzer counterparts were easily detected in both ch1 and ch2 bands. These 1510 sources comprise the WISE-selected Brown Dwarf Photometric Catalog, as described in Table 1.

The WISE and Spitzer photometry for each source in this catalog is graphically illustrated in Figure 1. This figure plots the Spitzer/IRAC ch1 − ch2 color as a function of WISE W1 − W2 color. The 1510 sources from the catalog are shown by black dots. A total of 166 of these have spectroscopic confirmation from Mainzer et al. (2011), Burgasser et al. (2011), Kirkpatrick et al. (2011, 2012), Cushing et al. (2011), and G. Mace et al. (in preparation), and are plotted with colored symbols, as indicated in the legend of the plot. The coldest brown dwarfs, the Y dwarfs, populate the reddest locus. The location of our latest type Y dwarf, WISE J182831.08+265037.7, which was targeted in a different Spitzer program (program 551; PI: Mainzer), is shown by the magenta octagon. As further illustrated in Figure 8 of Kirkpatrick et al. (2012), the Spitzer ch1 − ch2 color for WISE 1828+2650, typed as ⩾Y2, is not as red as objects typed slightly earlier, at Y1.

Zoom In Zoom Out Reset image size

Many excellent brown dwarf candidates in the catalog and on this diagram still lack spectroscopic data. Of the five objects in Figure 1 with colors of ch1 − ch2 > 2.8 mag, three are promising Y dwarf candidates—WISE J064723.23−623235.5 (ch1 − ch2 = 2.87, W1 − W2 > 3.77), WISE J082507.35+280548.5 (ch1 − ch2 = 2.96, W1 − W2 > 3.75), and WISE J220905.73+271144.00 (ch1 − ch2 = 2.98, W1 − W2 > 3.68). All three are extremely faint in the J and H passbands, so we are awaiting spectroscopic confirmation using the WFC3 grisms on board the Hubble Space Telescope. The other two sources, WISE 035358.23+375458.5 (ch1 − ch2 = 3.00, W1 − W2 = 3.52) and WISE J141127.45−612925.6 (ch1 − ch2 = 3.42, W1 − W2 > 3.35), are located within nebular patches on the Spitzer images and appear to be reddened or embedded objects rather than brown dwarfs.

These last two sources highlight an important point: spectroscopic confirmation of sources is vital to establishing candidates as bona fide brown dwarfs. Sixteen of our targets, shown by the yellow dots in Figure 1, have been shown spectroscopically not to be brown dwarfs (G. Mace et al. in preparation). These scatter widely over the diagram, proving that there is contamination at all colors. One interesting feature can be seen in the lower left quadrant of the diagram where our L and early-T dwarf candidates lie. These objects have bright WISE and Spitzer magnitudes, so the colors of these objects are well measured. These high-quality measurements enable us to split objects with W1 − W2 < 1.5 mag and ch1 − ch2 < 1.0 mag into two tracks. The lower track is replete with spectroscopically confirmed L and early-T dwarfs and primarily have 2MASS detections; the upper track has fewer confirming observations and generally no 2MASS detections, but objects with spectra here are typically not found to be brown dwarfs. Users of this catalog are cautioned that the sources here are candidates only, our selections were fairly liberal, so contamination by other types of sources is expected.

4.2. The W1W2-dropout Catalog

4.3. The Complete Photometric Catalog

We provide a catalog of the ∼7.33 × 10⁵ sources found in the Spitzer fields (COVCH1 ⩾ 3 and COVCH2 ⩾ 3) whose properties are described in Table 3. We provide WISE and 2MASS photometry from the WISE All-Sky source catalog for sources that fall within 2'' of a Spitzer source. In addition, we provide optical photometry for Spitzer sources that fall within 2'' of a source in the USNO-B catalog. Table 4 summarizes the statistics that resulted from matching between these catalogs for two different samples. Sample A comprises sources with the deepest imaging and most reliable photometric measurements available (COVCH1 = 5 and COVCH2 = 5). Sample B represents sources with slightly shallower imaging (COVCH1 ⩾ 3 and COVCH2 ⩾ 3) and thus slightly lower photometric reliability, though still reasonably good enough for most scientific purposes.

Table 4. Photometric Catalog Statistics

Sample	Coverage	No. of w/Spitzer	No. of w/WISE	No. of w/2MASS	No. of w/USNO-B
A	=5	3.41 × 105	1.00 × 105	3.90 × 104	7.75 × 104
B	⩾3	7.33 × 105	2.02 × 105	8.22 × 104	1.63 × 105

Download table as:  ASCIITypeset image

To characterize the photometric sample, we provide color–color and color–magnitude diagrams in Figure 2. The left plot of Figure 2 is similar to Figure 1, but includes all sources from Sample A having WISE measurements (a total of ∼10⁵ sources) rather than just the WISE brown dwarf candidates. We also show these colors for the 166 spectroscopically confirmed brown dwarfs, plotted with colored symbols as indicated in the legend of the plot. Few field sources in the Spitzer images have colors as extreme as the latest T and Y dwarfs.

Zoom In Zoom Out Reset image size

In the center plot of Figure 2, we show ch1 − ch2 versus ch2 mag. The limiting magnitude is ch2 ∼ 18.0 mag, which is ∼2.5 mag fainter then the WISE-selected candidates and comparable to the depth of the IRAC Shallow Survey (Eisenhardt et al. 2004), which had similar exposure time. We see that the majority of the WISE-selected brown dwarf candidates occupy a diffuse region in this color–magnitude domain, mainly with ch2 < 15.5 and ch1 − ch2 > 0.7.

It is challenging to separate Galactic sources from extragalactic sources using the ch1 and ch2 filters in the IRAC camera alone. In the rightmost plot of Figure 2, we show ch1 − ch2 versus ch2 − W3. Here, we can see the benefit of including photometric measurements at 12 μm. The additional color information helps separate the Galactic and extragalactic red (ch1 − ch2 > 1) populations with cool brown dwarfs having ch2 − W3 < 3, while the extragalactic population has ch2 − W3 > 3. A similar separation is shown in Figure 1 of Eisenhardt et al. (2010).

We present new brown dwarf candidates discovered from a blind search using the following search criteria: ch1 − ch2 ⩾ 1.5, ch2 ⩽ 17.0 mag and coverage equal to 5 in both IRAC channels (sample A from Table 4). As noted by Eisenhardt et al. (2010), this choice of color is expected for sources with spectral types later than T7 and the magnitude cut allows for only robustly measured sources to be inspected. This search resulted in 666 candidates from the brown dwarf fields and 247 candidates from the extragalactic fields. We then used the basic calibrated data (BCDs) to inspect the individual frames to remove any spurious sources or sources which appeared to be contaminated by cosmic rays. After also removing sources which were the WISE-selected targets, we were left with 13 new ultra-cool brown dwarf candidates: seven in the WISE brown dwarf fields and six in the WISE HyLIRG fields. Table 5 lists these 13 candidates, and Multi-wavelength finder charts (1' × 1') for them are presented in Figure 3. The finder charts show that all of these candidates lack detections at shorter wavelengths and 12 μm, as expected for brown dwarf candidates. However, the Spitzer images are relatively deep compared to the observations at shorter (DSS, 2MASS) and longer wavelengths (WISE). The difference in depth between these sets of observations makes it difficult to exclude certain types of extra-galactic objects such as AGNs.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 5. New Brown Dwarf Candidates

Origin	Candidate		Candidate		Primary		Typea	Typeb	destc	destd	Separatione
	R.A. (J2000)	Decl. (J2000)	ch2	ch1 − ch2	ch2	ch1 − ch2			(pc)	(pc)
BD	02:39:59.8	+70:40:20.4	16.42 ± 0.04	1.91 ± 0.16	14.14 ± 0.01	0.78 ± 0.01	T8	T5	40.6	21.8	3627
EXGAL	03:02:21.1	02:16:09.8	16.80 ± 0.05	1.54 ± 0.18
EXGAL	04:12:09.3	69:45:24.1	16.94 ± 0.06	1.65 ± 0.22
BD	05:24:39.3	+46:46:31.2	15.57 ± 0.02	2.29 ± 0.11	14.52 ± 0.01	1.03 ± 0.01	T9	T6	20.5	21.5	8326
EXGAL	08:21:51.3	23:45:17.1	16.63 ± 0.04	2.45 ± 0.32
BD	16:29:50.0	+25:45:25.9	16.45 ± 0.03	1.68 ± 0.12	15.44 ± 0.01	1.06 ± 0.03	T7.5	T6	44.5	32.8	13913
BD	16:53:42.8	−16:00:10.0	16.91 ± 0.08	1.69 ± 0.33	14.61 ± 0.01	1.00 ± 0.02	T7.5	T5.5	54.92	24.1	4498
EXGAL	20:38:38.1	−17:49:49.1	15.32 ± 0.01	2.61 ± 0.12
EXGAL	21:41:45.5	01:20:14.5	15.95 ± 0.02	2.27 ± 0.19
EXGAL	03:14:52.0	−53:52:41.9	16.81 ± 0.04	2.54 ± 0.40
BD	03:04:26.6	+43:31:10.9	16.77 ± 0.05	1.93 ± 0.22	14.90 ± 0.01	1.22 ± 0.02	T8	T7	47.6	23.0	2138
BD	18:22:38.6	−67:50:03.9	16.66 ± 0.04	2.06 ± 0.22	15.12 ± 0.01	1.09 ± 0.02	T9	T6.5	33.9	26.7	9627
BDf	13:59:42.7	−23:32:35.6	16.56 ± 0.04	1.61 ± 0.16			T7

Notes. ^aEstimated brown dwarf type for new candidate. ^bEstimated brown dwarf type for primary. ^cEstimated distance to the new candidate, in parsecs. ^dEstimated distance to the primary, in parsecs. ^eDistance between new candidate and the primary, in arcseconds. ^fprimary Spitzer target for this observation was not detected in the imaging, thus no information is given for the primary target.

Download table as:  ASCIITypeset image

For comparison, Eisenhardt et al. (2010) found 32 similar brown dwarf candidates in the deeper Spitzer Deep Wide Field Survey (SDWFS; Ashby et al. 2009) covering 10 deg², but selecting candidates to a fainter limit, ch2 ⩽ 18.5. Of these 32 SDWFS brown dwarf candidates, 7 had ch2 ⩽ 17.0. After accounting for the coverage depth of 5 criterion, the 2483 IRAC pointings searched here cover a comparable area to SDWFS, suggesting a somewhat higher surface density of brown dwarfs candidates than in SDWFS. All of the SDWFS candidates with ch2 ⩽ 17.0 were among those rejected as probable AGNs because they had ch2 − ch4 > 2. An analogous AGN rejection criterion using the ch2 − W3 color is not possible here because the W3 depth is not adequate, so it is plausible that many of the 13 brown dwarf candidates identified here are AGNs. If so, then it is possible that the somewhat higher surface density is attributable to AGN clustering.

Determining whether any of the new candidates could potentially be widely separated companions can be achieved in two ways. First, we check the surface density of our brown dwarf candidates in the control sample versus that in the brown dwarf fields. We find a surface density of (1.5 ± 0.6 deg⁻²) in the control sample and (1.0 ± 0.4 deg⁻²) in the brown dwarf fields. The errors are calculated based on simple Poisson statistics. The fact that the control field has a larger surface density suggests that the candidates found are unlikely to be physically associated companions. Second, we can determine the spectrophotometric distances to both the primary WISE-selected source and the new Spitzer-selected candidate. We estimate the spectral type by using Figure 11 from Kirkpatrick et al. (2011) and the derived ch1 − ch2 color. We then estimate the photometric distances listed in Table 5 using the apparent magnitude and the absolute magnitude, where the latter is estimated using the absolute magnitude versus spectral type relationship in Figure 13 of Kirkpatrick et al. (2012).

Most of the new candidates do not appear to be physically associated with the WISE-selected targets, but are more distant background sources. Assuming that the WISE-selected candidates are indeed brown dwarfs, they reside at ∼20 pc, while the majority of the new Spitzer-selected candidates are at ≳40 pc. There is only one case for which both the primary and the candidate may be physically associated, J052439.3+464631.2, where both sources have estimated distances of ∼20 pc. The estimated spectral types for these two sources are T6 and T9. The angular separation of 8326 for this system corresponds to a physical separation of at least 1600 AU at 20 pc.

The probability of discovering widely separated companions rests on the probability that such systems exist, which is itself a complicated function of mass, age, environment, and formation mechanisms. An in depth review on multiplicity in very low mass systems is given in Burgasser et al. (2007) and references therein. These brown dwarf primaries have low mass and thus small gravitational potential wells. This is well summarized in Figure 6 of Burgasser et al. (2007), where they plot the separation of binaries as a function of total mass for a wide variety of known systems. Brown dwarf binaries tend to have small separations with higher binding energies. Systems with lower binding energies tend to be disrupted, primarily due to outside perturbers or gravitational instabilities, which translates to almost no old brown dwarf binaries having been found with separations greater than 20 AU. There are only a handful of low-mass binaries with separations greater than 20 AU, three are relatively young and thus have not had enough time for collisional disruption. Another system is DENIS 0551-4443AB (Billères et al. 2005), which is thought to be relatively old and is the kind of widely separated system for which we were searching. Billères et al. (2005) suggest that such a system is fragile, and it would not have survived a close encounter with a third body. Furthermore, its existence demonstrates that some very low mass stars/brown dwarfs form without ejection from a multiple system, or any other strong dynamical interaction. Other widely separated multiple systems have been discovered since Burgasser et al. (2007), including the NLTT20346 + 2MASS J0850359+105716 system (Faherty et al. 2011). Is it possible that J052439.3+464631.2 is also an old, widely separated binary like these?

The nature of J052439.3+464631.2 can be probed further by considering the environment, magnitudes, and colors of the individual objects in this system. Using the WISE All-Sky Image Archive, we investigate a region ∼5' on a side centered on the new candidate and make the following observations. First, this system resides near the Galactic Plane which makes it more difficult to interpret given the high source density and complex nebulosity. The pair resides at the edge of a 12 μm bright nebula, thus suggesting that rather than a widely separated T6–T9 binary, these are more likely reddened or embedded objects and not brown dwarfs. Another possibility is that they could be physically associated T6 and T9 brown dwarfs, though not in a binary system, i.e., forming in the same stellar cocoon and around the same time. Spectroscopic observations of these candidates should be made to disentangle these scenarios.

Other than J052439.3+464631.2, this search yielded no plausible widely separated companions, therefore suggesting that very low mass (>T7) brown dwarf binaries with separations of 20 to 2000 AU are extremely rare.

This paper summarizes initial Spitzer follow-up and analysis of some of the rarest astrophysical objects discovered to date. We present, for the first time, a very large sample of WISE-selected ultra-cool brown dwarf candidates. Of these, 184 have spectroscopic data with ∼92% of them confirmed as bona fide ultra-cool brown dwarfs and of which ∼83% have spectral types later than T5. We also present 13 new, though more distant, ultra-cool (>T7) brown dwarf candidates discovered using the Spitzer data. Only one possible widely separated companion system was found, suggesting that widely separated, cold brown dwarf binaries are extremely rare. To date, the WISE team has only followed up a small percentage (∼12%) of the total WISE-selected sample, providing follow-up opportunities by the astrophysical community on a host of interesting discoveries yet to be made. Given the wide range of (sub)stellar spectral types it probes, the sample will allow for a multitude of future statistical studies. In addition, this sample will provide excellent follow-up sources for the James Webb Space Telescope.

There are several avenues of investigation yet to be fully explored which we summarize as follows. Detailed proper motion measurements can be made with the current data available (WISE + Spitzer), though long term monitoring will allow for full kinematics and robust distances to be determined. Several large parallax programs are currently underway monitoring a large number of the spectroscopically verified sources, though many remain to be monitored. Long term monitoring will also provide critical insight and constraints on variability in brown dwarfs, two examples being the study of atmospheric meteorology and possible detection of transits by planetary companions. Using the difference between the WISE and Spitzer filter profiles will allow one to explore differences in the underlying spectra at these wavelengths. With no all-sky deep surveys planned at these wavelengths in the foreseeable future, this is the quintessential brown dwarf sample because it comprises the closest and brightest brown dwarfs possible.

In addition to presenting these two very rare populations of astrophysical sources, we release a catalog of ∼7.33 × 10⁵ objects detected in the Spitzer observations, including multi-wavelength (2MASS, USNO-B, and WISE) photometry for a subsample. Providing multi-wavelength photometry for hundreds of thousands of astrophysical sources will allow for a number of scientific studies to be conducted, from fully characterizing the mid-IR properties of the coldest brown dwarfs, to a more complete characterization of the most IR luminous galaxies in the universe.

This publication makes use of data products from the Wide-Field Infrared Survey Explorer, which is a joint project of the University of California, Los Angeles, and the Jet Propulsion Laboratory/California Institute of Technology, funded by the National Aeronautics and Space Administration (NASA). This publication also makes use of observations made with the Spitzer Space Telescope, which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under a contract with NASA. This work is also based in part on observations made with the NASA/ESA Hubble Space Telescope, obtained at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. Some of the spectroscopic classifications presented herein were obtained at the W. M. Keck Observatory, which is operated as a scientific partnership among the California Institute of Technology, the University of California and the National Aeronautics and Space Administration. This publication also makes use of data products from 2MASS. 2MASS is a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by the National Aeronautics and Space Administration and the National Science Foundation. This research has made use of the NASA/IPAC Infrared Science Archive (IRSA), which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with NASA.