GRB 211227A as a peculiar long gamma-ray burst from compact star merger

Long-duration gamma-ray bursts (GRBs) associated with supernovae (SNe) are believed to originate from massive star core-collapse events, whereas short-duration GRBs that are related to compact star mergers are expected to be accompanied by kilonovae. GRB 211227A, which lasted about 84 s, had an initial short/hard spike followed by a series of soft gamma-ray extended emission at redshift $z=$0.228. We performed follow-up observations of the optical emission using BOOTES, LCOGT, and the Lijiang 2.4m telescope, but we detected no associated supernova signature, even down to very stringent limits at such a low redshift. We observed the host galaxy within a large error-circle and roughly estimate the physical offset of GRB 211227A as $20.47\pm14.47$ kpc from the galaxy center. These properties are similar to those of GRB 060614, and suggest that the progenitor of GRB 211227A is not favored to be associated with the death of massive stars. Hence, we propose that GRB 211227A originates from a compact star merger. Calculating pseudo-kilonova emission for this case by adopting the typical parameters, we find that any associated pseudo-kilonova is too faint to be detected. If this is the case, it explains naturally the characteristics of the prompt emission, the lack of SN and kilonova emission, and the large physical offset from the galaxy center.

Phenomenally, GRBs are divided as long-and shortduration with a division line at the observed duration T 90 ∼2 s (Kouveliotou et al. 1993). From the theoretical point of view, long-duration GRBs are powered by a relativistic jet that breaks out of the envelope of a massive star when it undergoes core-collapse. There is about 0.1M ⊙ of nickel that can also be created via explosive nucleosynthesis during this collapse. Therefore, a bright optical/infrared transient called a supernova (SN; Woosley & Bloom 2006;Hjorth & Bloom 2012) should be produced when the 56 Ni decays into cobalt. Naturally, some long-duration GRBs are associated with SN when the distance is not large enough (Galama et al. 1998;Kippen et al. 1998;Hjorth et al. 2003;Stanek et al. 2003;Malesani et al. 2004;Della Valle et al. 2006;Mazzali et al. 2008;Bufano et al. 2012;Xu et al. 2013;Cano et al. 2017;Lü et al. 2018). On the other hand, the short-duration GRBs are believed to originated from compact star mergers. The leading candidates are neutron star−neutron star (NS−NS) and neutron star−black hole (NS−BH) systems. The NS−NS merger may result in a magnetar as remnant (Rosswog et al. 2000;Dai et al. 2006;Metzger et al. 2010;Yu et al. 2013;Zhang 2013;Lasky et al. 2014;Lü et al. 2015) or black hole (Rosswog et al. 2014). If this is the case, a mildly isotropic, sub-relativistic ejecta with 10 −4 − 10 −2 M ⊙ and 0.1−0.3c, can be ejected during the merger to heat ejecta (Rosswog et al. 2014;Hotokezaka et al. 2013). So that, an optical/infrared transient that Lü et al. is powered by radioactive decay from r-process radioactive materials may be detected from any direction if the flux is high enough (Li & Paczýnski 1998;Kulkarni 2005;Metzger et al. 2010;Barnes & Kasen 2013), and the transient is known as a macronova, kilonova, or merger-nova (Kulkarni 2005;Metzger et al. 2010;Yu et al. 2013).
Observationally, the first GRB-SN association event was discovered as the under−luminous GRB 980425 and the Type Ic SN 1998bw at redshift z =0.0085 (Galama et al. 1998;Kippen et al. 1998;Pian et al. 1998;Sadler et al. 1998). Afterwards, a handful of long-duration GRBs associated with spectroscopically−identified type Ib/c SNe were detected (Kovacevic et al. 2014;Cano et al. 2016;Lü et al. 2018). By comparing with the first GRB−SN association, the first kilonova emission event from a compact star merger was discovered in GRB 130603B with an excess near−IR emission matching the predictions for r-process emission (Tanvir et al. 2013;Berger et al. 2013;Fan et al. 2013;Fong et al. 2014). After that, several short GRBs were claimed to be associated with a kilonova or mergernova, e.g., GRB 050709 (Jin et al. 2016), GRB 060614 (Yang et al. 2015), GRB 070809 (Jin et al. 2020), GRB 080503 (Gao et al. 2015), GRB 160821B (Kasliwal et al. 2017;Troja et al. 2019;Lamb et al. 2019), and GRB 150101B (Troja et al. 2018). Gao et al (2017) carried out a complete search for magnetar−powered mergernovae from a sample of Swift GRBs, and found that three magnetar−powered mergernova candidates are associated with short GRBs (050724, 061006 and 070714B) from late optical observations. Rastinejad et al. (2021) presented a comprehensive optical and near-infrared catalog to search for possible kilonova emission and constrain its ejecta mass (also see Yuan et al. 2021).
A particularly interesting case, GRB 060614 which is a nearby long-duration GRB at z = 0.125 (Gehrels et al. 2006;Gal-Yam et al. 2006;Fynbo et al. 2006;Della Valle et al. 2006), needs to be mentioned again. The light curve of GRB 060614 is characterized by a short/hard spike (with a duration ∼ 5 s) followed by a series of soft gamma-ray extended emission 9 with a duration ∼ 100 s. Phenomenologically, it definitely belongs to the long-duration GRB population and a bright SN was expected. It should have been detected at such a low redshift, but a SN association was not detected even with a deep search. These facts are consistent with the compact star merger scenario, but they are not the direct evidence (Gehrels et al. 2006;Zhang et al. 2007a). There was an ongoing debate on the physical origin of this case until 2015. Possible evidence of a kilonova from SN-less long-duration GRB 060614 was claimed (Yang et al. 2015). They discovered a near-infrared bump (only two data points) that is significantly above the regular decaying afterglow, and therefore claimed it to be a signature of the kilonova component.
GRB 211227A is potentially associated with a galaxy at redshift z =0.228 (Beardmore et al. 2021;Malesani et al. 2021). The light curve of its prompt emission is very 9 The observed extended emission is softer than that of initial short/hard spike of short GRBs and gamma-ray emission of typical long GRBs. The extended emission detected in the BAT band seems to be simply the internal plateau emission when the emission is bright and hard enough (Lü et al. 2015). similar to that of GRB 060614, which also has no SN association at this low redshift. However, the difference with GRB 060614 is that we do not find any kilonova signature with GRB 211227A. One question is what is the progenitor of GRB 211227A, massive star collapse or compact star merger? In this paper, we systematically analyze the observational data of both prompt emission, afterglow, and host galaxy (in §2). Then, comparisons to GRB 211227A and GRB 060614 are shown in §3. In §4, we attempt to investigate the possible origin by calculating the SN and kilonova emissions. The conclusions are drawn in §5 with some additional discussion. Throughout the paper, a concordance cosmology with parameters H 0 = 71 km s −1 Mpc −1 , Ω M = 0.30, and Ω Λ = 0.70 is adopted.

THE OBSERVATIONS AND DATA ANALYSIS
2.1. Swift BAT observations GRB 211227A triggered the Burst Alert Telescope (BAT) at 23:32:06 UT on 27 December 2021 (Beardmore et al. 2021). We downloaded the BAT data from the Swift website 10 , and used the standard HEASOFT tools (version 6.28) to process the BAT data. For more details on the analysis, please refer to Sakamoto et al. (2008); Zhang et al. (2009);and Lü et al. (2020). We extract the light curves in different energy bands with a 128 ms time-bin ( Figure 1). The light curve shows a complex structure with a total duration of about T 90 =84 s, an initial short/hard spike (with a duration ∼ 4 s) followed by a series of soft gamma-ray extended emission with a duration ∼ 80 s.
We also extract the spectrum by invoking Xspec for fitting, and the background is extracted by choosing two time-intervals with 80 s before and after the burst. Due to the narrow energy-band of Swift/BAT, the timeintegrated spectrum of the GRB 211227A prompt emission is well fitted by a simple power-law model (N ∝ E −Γ ) with an index Γ = 1.53 ± 0.03. Moreover, we also separate the light curve into two time slices: the initial hard spike (0.1s-4s) and the long-lasting extended emission (4s-84s). The spectra of both time slices can be fitted by a power-law model with Γ 1 = 1.67 ± 0.08 and Γ 2 = 1.5 ± 0.03, respectively (see Figure 2). The bottom panel of Figure 1 shows a strong temporal evolution of the spectrum, with Γ ∼ 1.5 at the beginning and Γ ∼ 2 near the end. The fluence in the 15-150 keV band is 8.2 ± 0.2 × 10 −6 erg cm −2 , which corresponds to isotropic energy E iso ∼ 1.14 × 10 51 ergs by adopting z = 0.228 in this energy band.

Swift XRT observations
The Swift X-ray telescope (XRT) began observing the field at 73.5 seconds after the BAT trigger. We made use of the public data from the Swift archive 11 (Perri et al. 2021). The X-ray light curve in the early time seems to be a broken power-law decay with indices α 1 = 0.2±0.08, α 2 = 3.10 ± 0.31, and break time t b = 103 ± 12 s (see Figure 4). We also extract the time-resolved spectra of the initial X-ray tail with a power law model, and the spectrum shows a strong temporal evolution that tracked hard-to-soft. The time-integrated spectrum of the GRB 211227A In X-ray emission is fitted by a simple powerlaw model with an index Γ = 1.11±0.07, and the column density of hydrogen N H is (2.2 ± 0.5) × 10 21 cm −2 .

BOOTES follow-up observations
The Burst Observer and Optical Transient Exploring System (BOOTES) followed GRB 211227A with two 60cm robotic telescopes at BOOTES-2/TELMA station in La Mayora (Malaga, Spain) and BOOTES-4/MET station at the Lijiang Astronomical Observatory (Yunnan, China). The BOOTES-2/TELMA telescope performed two epoch observations at 2021-12-27.99UT and 2021-12-28.93UT and the BOOTES-4/MET telescope took four epoch observations at 2021-12-28.69UT, 2021-12-29.61UT, 2022-01-03.75UT, and 2022-01-04.69UT, respectively. A series of images were obtained using clear and Sloan-i filters with exposure of 10s, 60s and 90s (Hu et al. 2021). The burst's optical afterglow is not detected in the stacked images of each epoch which are calibrated via nearby comparison stars from the SDSS catalogue for the i-filter data or through the transformation equation for the clear filter data. The obtained upper limits with in 3σ are listed in Table1, and the value of extinction (Av) is 0.062.

Lijiang 2.4m Optical Observations
In order to observe the potential optical emission, on 3 January 2022 (∼6.8 days after trigger), we observed the field of GRB 211227A with the Lijiang 2.4m telescope, which is located at the Lijiang Observatory of Yunnan Observatories, Chinese Academy of Science (Wang et al. 2019;Xin et al. 2020). The photometric observations were performed using the Johnson R-band filter with a total exposure of 1200 s at an airmass of 1.2 and a seeing of ∼1.7 under good weather conditions. Three nearby faint stars are in the field of view (FOV) which can provide calibration of the target. However, we did not detect any source to a depth of R>21.5 mag within the XRT enhanced error box. On 7 January 2022 (∼10.7 days after trigger), we observed this field again, with a total exposure time to 1800 s. Again, no signal was detected, and only an upper limit with R>21.8 mag was obtained.

Other telescopes follow-up observations
Beside the BOOTES and Lijiang 2.4m, there were several ground-optical telescopes that performed follow-up observations of the source after the GRB 211227A trigger, such as the Nanshan/NEXT-0.6m located at Nanshan, Xinjiang, China (Fu et al. 2021), the LCO 1-m Sinistro instrument at the Teide, Tenerife site (Strausbaugh & Cucchiara 2021), the 2m robotic Liverpool telescope (Perley 2021), the CAHA 2.2m telescope at the Calar Alto Observatory, Almeria, Spain (Kann et al. 2021), and the Gemini/GMOS-S (O'Connor et al. 2022). However, none of them could detect any source within upper limits.

Host galaxy Observation
We observed the localization of GRB 211227A by using the 1m telescope at the Las Cumbres Observatory Global Telescope (LCOGT) at the Teide Tenerife site.
The observation started at 01:28:37 UT on 28 December 2021 (from 6991s to 8599s after the BAT trigger), and consisted of 5×300 s integrations in the R band. Our combined image is centered at 7795s after the trigger. No afterglow is detected in the XRT error circle in our stacked image down to a limiting magnitude of R> 23.81 mag, calculated with respect to the USNO-B.1 catalog. However, there exists a known source at the border of the Swift/XRT error circle at coordinates: R.A. (J2000)=08:48:35.975 and Dec (J2000)= -02:44:06.93, which has R ∼ 19.13 ± 0.09 mag from our images, classified as a galaxy with a photo-redshift of z = 0.244±0.089 from SDSS ( Figure 3). Moreover, Malesani et al. (2021) reported that 5 emission lines are observed in the putative host galaxy (e.g., [O II], Hbeta, [O III], Halpha, [N II], and [S II]), and they proposed that a possible host galaxy of GRB 211227A with redshift z=0.228 was detected by the X-shooter spectrograph. The redshift z = 0.244 ± 0.089 what we measured is consistent with that reported in Malesani et al. (2021) with z =0.228 via the X-shooter spectrograph, and the position of this source for LCOGT observations is the same with GCN report in Malesani et al. (2021). In our calculations, we adopt z =0.228 which is the spectroscopic redshift. In addition, we also calculate the physical distance from the center of GRB 211227A to the host galaxy center as 20.47±14.47 kpc if we assume that the galaxy of z =0.228 is intrinsic host of GRB 211227A.

GRBS
From the observational point of view, the duration of GRB 211227A is much longer than that of shortduration GRBs, but is consistent with that of typical long-duration GRBs. Even only considering its initial hard spike lasting ∼4 s, is still characterized this burst as a category of long-duration GRBs. However, based on the characteristics of the GRB 211227A prompt emission, it is very similar to that of GRB 060614 which had an initial short/hard spike with ∼5 s followed by a series of soft gamma-ray extended emission lasting ∼102 s (Gal-Yam et al. 2006;Zhang et al. 2007a). Figure  1 shows the complete comparison of the prompt emission light curves in different energy bands. The profile of the light curves of GRB 211227A and GRB 060614 are consistent with each other, and both of them are characterized by a short/hard spike followed by a series of soft gamma-ray extended emission with a duration ∼ 100 s. Moreover, we also present their spectral evolution, and find that a strong temporal evolution is obvious in both. The evolution behavior of the spectrum is from Γ ∼ 1.5 at the beginning to Γ ∼ 2 near the end. Moreover, the isotropic energy of GRB 211227A is E iso ∼ 1.14 × 10 51 ergs, which is also close to that of GRB 060614 with E iso ∼ 7.76 × 10 50 ergs.
In terms of the X-ray afterglow, the X-ray emission of GRB 211227A is not significant difference with other long-duration GRBs that presented several segments (Zhang et al. 2006). The spectrum shows a strong temporal evolution that tracked hard-to-soft, which is consistent with the behavior of X-ray tail emission in most long-and short-duration GRBs (Zhang et al. 2007b). Figure 4 shows the X-ray emission of GRB 211227A in comparison to that of GRB 060614 (left) and other short Lü et al.
GRBs with extended emission (right). We find that the temporal decay index of the X-ray tail before 10 3 s is consistent with that of GRB 060614 (f ∝ t −3 ; Zhang et al. 2007b), and shows a similar temporal evolution that tracked from hard-to-soft. Moreover, we also collect all of the X-ray light curves of the short GRBs with extended emission and plateau emission in X-ray observed by Swift (Lü et al. 2015) 12 , and compare with the X-ray emission of GRB 211227A. We find that the behavior of the early X-ray emission of GRB 211227A is also similar to that of other short GRBs with extended emission. Moreover, Figure 3 shows the comparison of the physical offset of GRB 211227A with that of other long and short GRBs. We find that the physical offset is larger than that of most long-duration GRBs, but is consistent with the distribution of other short GRBs within a larger error range. In general, long-duration GRBs are believed to originate from the deaths of massive stars, and their host galaxies are typically irregular galaxies with intense star formation (Woosley 1993;Fruchter et al. 2006). In contrast, short-duration GRBs are associated with nearby early-type galaxies with little star formation and are related to compact star mergers with a large offset from the center of the host galaxy (Berger et al. 2005;Fong et al. 2010). Moreover, from the statistical point of view, Leibler & Berger (2010) found that the two GRB host populations remain distinct based on the stellar mass distribution and population ages of galaxies. However, such a cozy picture was destroyed by nearby long-duration GRB 060614 without a SN association (Gehrels et al, 2006;Gal-Yam et al, 2006;Fynbo et al, 2006;Della Valle et al, 2006) and short-duration GRB 200826A associated with a SN (Zhang et al. 2021;Ahumada et al. 2021). Interestingly, GRB 211227A shares some similar properties with GRB 060614, and has a large physical offset from the galaxy center. One question is what is the physical origin of GRB 211227A? In this section, we attempt to find some clues to reveal the physical origin of this case.

Massive star core-collapse scenario
The "smoking gun" signature of a long-duration GRB from a massive star core-collapse is the detection of the associated SN in the optical band (Galama et al. 1998;Kippen et al. 1998;Hjorth et al. 2003;Stanek et al. 2003;Malesani et al. 2004;Della Valle et al. 2006;Mazzali et al. 2008;Bufano et al. 2012;Xu et al. 2013;Cano et al. 2017;Lü et al. 2018). Being a long-duration GRB 211227A at a low redshift z =0.228, it is expected that a SN component should be detected around ten days after the trigger, yet it is surprising that deep searches of an underlying SN give null results. This raises interesting questions regarding whether the distance is too large for detecting a less energetic event. Therefore, we shift several SNe to the redshift z =0.228 to see how bright they are. Figure 5 shows light curves of SN 1998bw, SN 2001ke, SN 2013dx, and SN 2016jca associated with some long-duration GRBs at z =0.228. We find that the upper limit of the luminosity of the SN (also the limiting magnitude of the Lijiang 2.4m telescope with the limit luminosity as (4.3 − 5.6) × 10 42 erg s −1 ), if any, is several times fainter than SN 1998bw and fainter than other Type Ic SN associated with GRBs at z =0.228 (Galama et al. 1998;Della Valle et al. 2003;Xu et al. 2013;Cano et al. 2017). It suggests that lack of a SN associated with GRB 211227A is not caused by too large of a distance, but more likely physical reasons. In order to compare those SNe at z =0.228 with that of higher redshift, we re-plot those SNe if we put them into z =1. It is much dimmer than the upper limits of observations. On the other hand, based on the upper limit of observations, one can roughly estimate the redshift z = 0.285 of SNe that can be ruled out. Moreover, the measurement of a large physical offset is also inconsistent with that of other long GRBs. From the prompt emission analysis, on the other hand, GRB 211227A is consistent with the properties of GRB 060614. Those facts suggest that it does not likely match with the physical properties expected in massive star core-collapse. If so, the lack of a SN signature is a natural expectation. However, the origin of massive star collapse of GRB 211227A is still to be ruled out if it occur in the host galaxy with a higher redshift.

Compact star mergers scenario
The direct detection of GW170817 and its electromagnetic counterpart (e.g., GRB 170817A and AT2017gfo), which was the first identified multimessenger gravitational-wave and electromagnetic signal, originated in the merger of a binary NS system (Abbott et al. 2017;Covino et al. 2017;Goldstein et al. 2017;Savchenko et al. 2017;Zhang et al. 2018). The observations of kilonova AT2017gfo were comprehensive, long lasting, and multi-wavelength compared to previous kilonova candidates (Coulter et al. 2017;Arcavi et al. 2017;Lipunov et al. 2017;Tanvir et al. 2017;Valenti et al. 2017;Ai et al. 2018;Metzger 2019;Rossi et al. 2020). A roughly constraint of kilonova parameters can be realized via a multi-wavelength fit with the kilonova model Li et al. 2018;Ai et al. 2018). By adopting the same parameters (M ejc ∼ 0.03M ⊙ , β ∼ 0.25c, and κ ∼ 0.97 cm 2 g −1 ) with kilonova AT2017gfo, one can also calculate the kilonova emission of GRB 211227A in K-, R-, and U-bands at redshift z =0.228 (see Figure 7). The peak magnitude is not much of an improvement, and it remains fainter than the observed limitation. Yang et al. (2015) reported the discovery of nearinfrared bump that is inconsistent with the afterglow emission of GRB 060614, but arises from a kilonova. They invoke the kilonova model to fit the excess in the near-infrared band with the ejecta mass (M ejc ∼ 0.1M ⊙ ), velocity (β ∼ 0.2c), and opacity (κ ∼ 10 cm 2 g −1 ). In order to compare the kilonova emission between GRB 211227A and GRB 060614, we calculate the possible kilonova emission by adopting the same parameters with that of GRB 060614 even with the large uncertainty. Figure 6 shows the kilonova emission in K-, R-, and U-bands at redshift z =0.228 by considering only one energy source. Here, the ejecta mass range (0.01 − 0.1)M ⊙ ) is adopted in our calculations. The peak magnitude of kilonova emission is much lower than the observed limitation of the Lijiang 2.4m and current other optical telescopes. If so, the lack of a kilonova signature is the natural expectation for a less energetic event.
Inspired by GRB 170817A and GRB 060614, we propose that the long-duration GRB 211227A is from a more energetic merger event. Moreover, the large physical offset (∼20.47 kpc) of this case that is larger than that of most long GRBs, is also supported this hypothesis even still consistent with serval long GRBs within a large error. In general, robust associations of a fainter-thansupernova optical/IR transient (called kilonova) with some GRBs suggest that they are likely related to NS-NS or NS-BH mergers (Tanvir et al. 2013;Berger 2014;Yang et al. 2015;Abbott et al. 2017;Goldstein et al. 2017;Savchenko et al. 2017;Troja et al. 2017;Zhang et al. 2018). Such a transient is powered by the rprocess (rapid neutron capture) and radioactive decay of the synthesized heavy elements (e.g., Li & Paczýnski 1998;Freiburghaus et al. 1999;Hotokezaka et al. 2013;Coulter et al. 2017;Arcavi et al. 2017;Lipunov et al. 2017;Tanvir et al. 2017;Valenti et al. 2017). If GRB 211227A is from a compact star merger, one can roughly calculate how bright it is by adopting some typical parameters.
For a NS-NS merger, additional energy should be injected into the ejecta if the post-merger remnant is a NS (Metzger et al. 2010;Yu et al. 2013;Gao et al. 2017;Yu et al. 2018;Ai et al. 2018;Yuan et al. 2021). We also invoke a model of hybrid energy sources (r-process and energy injection from the NS) to calculate the kilonova emission. Lü et al. (2015) proposed that both extended emission and internal plateau can be produced by magnetar spin-down 13 . The observed EE and X-ray tail of GRB 211227A seems to come from the contribution of a magnetar spin-down. If this is the case, we adopt the same parameters as above in the r-process, and L 0 = 10 48 erg s −1 and t = 105 s as the initial luminosity and time scale of energy injection from NS. We find that the peak magnitude is also not improved even when considering hybrid energy sources, because the time-scale of energy injection is too short to be contributed to the kilonova. Hence, the progenitor of GRB 211227A seems to come from compact star mergers, which can naturally explain the lack of SN and kilonova emission, the characteristic of prompt emission, the initial X-ray emission, and the large physical offset.

CONCLUSION AND DISCUSSION
GRB 211227A was observed by Swift to have a duration of ∼ 84 seconds at redshift z =0.228, but the light curve is characterized by an initial short/hard spike (with a duration ∼ 4 s) followed by a series of soft gamma-ray extended emission with a duration ∼ 80 s. Both the prompt emission and the X-ray initial tail of this case show a strong temporal evolution that tracked from hardto-soft, and those behaviours are very similar to that of GRB 060614. Several optical telescopes made follow-up observations of the afterglow and host galaxy, but did not detect any afterglow emission except for upper limitations. We also applied for the Lijiang 2.4m optical telescope to observe this event twice, but obtained only upper limits. Based on the LCOGT observations, we can roughly estimate the physical offset from the galaxy center as 20.47 ± 14.47 kpc.
At such low redshift, it is expected that a SN component should be detected around ten days after the trigger if we believe the massive star core-collapse origin, yet it is surprising that deep searches of an underlying SN give null results. We shift several SN events that are associated with long-duration GRBs into the redshift z =0.228, and find that the upper limit of the luminosity of the SN, if any, is several times fainter than SN 1998bw, and fainter than other Type Ic SN associated with GRBs at z =0.228. The lack of an associated SN and the large physical offset of this case suggest that the progenitors of GRB 211227A do not seem to originate from a massive star core-collapse. If so, the lack of a SN signature and the large physical offset are a natural expectation.
Alternatively, we propose that the long-duration GRB 211227A is from a compact star merger. If this is the case, we also calculate the kilonova emission by adopting the same parameters with that of GRB 060614 and GRB 170817A, and find that the kilonova emission is too faint to be detected. Hence, the progenitor of GRB 211227A seems to come from compact star mergers, which can naturally interpret the lack of SN and kilonova emissions, the characteristics of the prompt emission, and the large physical offset. Moreover, Kann et al. (2011) used a large sample of GRB afterglow data to compare the optical afterglows of Type I (originated from compact star merger) and Type II GRBs (related to death of massive star), and found that the optical afterglow of Type I GRBs are intrinsically lower than that of Type II GRBs. The lack of optical afterglow of GRB 211227A also support its origin in a compact star merger.
On the other hand, GRB211227A is a possible optically-dark burst. Melandri et al. (2012) studied the properties of the population of completely optically-dark GRBs that observed by Swift, and presented the relationship between optical flux and X-ray flux that observed at t = 11 hours for completely sample of optically-dark GRBs. In order to test this possibility, we also calculate the upper limits of both optical and X-ray flux observed at t = 11 hours, and find that the observed both optical and X-ray flux at t = 11 hours are much lower than that of optically-duck bursts identified in Melandri et al. (2012). So that, the possibility that GRB 211227A is an optically-dark GRB cannot be ruled out. If this is the case, GRB 211227A should be generated in much denser environments in the host galaxy. Moreover, another possibility is that the GRB 211227A is not host at z=0.228, but located at a higher redshift with a faint and underlying galaxy which we do not observe by the deepest upper limit of the field. Bloom et al (2002) and Lyman et al. (2017) pointed out that one can calculate the probability of chance coincidence of possible host galaxy. Following the method of those two paper, we also roughly estimate the probability of chance coincidence P ch = 0.038, which is less than 0.1 (P ch ≥ 0.1 no obvious host; Bloom et al (2002)). So that, the galaxy at z=0.228 seems to be the host of GRB 211227A.
In order to investigate the progenitor of a GRB 211227A-like event, more information on the host galaxy is essential. We therefore encourage intense multiband, more sensitive optical follow-up observations of GRB 211227A-like events to catch electromagnetic signals (e.g., kilonova or SN signatures) and the properties of the host galaxy in the future. If possible, the observed GRB 211227A-like event with a GW signal will also provide a good probe to study progenitors.
We are very grateful to thank the referee for constructive comments and suggestions to improve this manuscript. We also thank WeiKang Zheng and Meng-Hua Chen for helpful comments. We acknowledge the use of the public data from the Swift data archive. This work is supported by the National Natural Science Foundation of China (grant Nos. 11922301, 12133003, and U1938201), the Guangxi Science Foundation (grant Nos. 2017GXNSFFA198008, and AD17129006), the Program of Bagui Young Scholars Program (LHJ), the teaching reform project of Guangxi Higher education(2019JGZ102), and special funding for Guangxi distinguished professors (Bagui Yingcai and Bagui Xuezhe). AJCT acknowledgdes support from the Spanish Ministry project PID2020-118491GB-I00 and the collaboration with the rest of IAA-CSIC ARAE group and the UMA Unidad Asociada to CSIC colleagues. MCG ackowledges support from the Ramón y Cajal Fellowship RYC2019-026465-I. YDH acknowledges support under the additional funding from the RYC2019-026465-I.  -Swift/BAT light curves of GRB 211227A in different energy bands with a 128 ms time-bin (red solid lines) and the spectral evolution (red diamonds). In order to compare with GRB 060614, we also plot the light curve (black solid lines) and spectral evolution (black points) in the same panels.       -Calculated kilonova emission of GRB 211227A at z =0.228 by considering one energy source with M ejc ∼ 0.03M ⊙ , β ∼ 0.25c, and κ ∼ 0.97 cm 2 g −1 which are taken from kilonova AT2017gfo .