Photospheric Radius Expansion Bursts Research Articles

NICER views moderate, strong, and extreme photospheric expansion bursts from the ultracompact X-ray binary 4U 1820–30

Type I X-ray bursts in the ultracompact X-ray binary 4U 1820–30 are powered by the unstable thermonuclear burning of hydrogen-deficient material. We report the detection of 15 type I X-ray bursts from 4U 1820–30 observed by NICER between 2017 and 2023. All these bursts occurred in the low state for the persistent flux in the range of 2.5–8 × 10−9 erg s−1 cm−2 in 0.1–250 keV. The burst spectra during the tail can be nicely explained by blackbody model. However, for the first ~5 s after the burst onset, the time-resolved spectra showed strong deviations from the blackbody model. The significant improvement of the fit can be obtained by taking into account of the enhanced persistent emission due to the Poynting–Robterson drag, the extra emission modeled by another blackbody component, or by the reflection from the surrounding accretion disk. The reflection model provides a self-consistent and physically motivated explanation. We find that the accretion disk density changed with 0.5 s delay in response to the burst radiation, which indicates the distortion of the accretion disk during X-ray bursts. From the time-resolved spectroscopy, all bursts showed the characteristic of photospheric radius expansion (PRE). We find one superexpansion burst with the extreme photospheric radius rph > 103 km and blackbody temperature of ~0.2 keV, 13 strong PRE bursts for rph > 102 km, and one moderate PRE burst for rph ~ 55 km.

Insight-HXMT observations of thermonuclear X-ray bursts in 4U 1636−53

ABSTRACT We conducted an analysis of 45 bursts observed from 4U 1636−53. To investigate the mechanism behind the light-curve profiles and the impact of thermonuclear X-ray bursts on the accretion environment in accreting neutron star low-mass X-ray binaries. This analysis employed both light-curve and time-resolved spectroscopy methodologies, with data collected by the Insight-Hard X-ray Modulation Telescope instrument. We found that 30 bursts exhibited similar light-curve profiles and were predominantly in the hard state, and two photospheric radius expansion (PRE) bursts were in the soft state. The light curves of most bursts did not follow a single exponential decay but displayed a dual-exponential behaviour. The initial exponent had a duration of approximately 6 s. We utilized both the standard method and the ‘fa’ method to fit the burst spectra. The majority of the ‘fa’ values exceeded 1, indicating an enhancement of the persistent emission during the burst. Under the two Comptonization components assumption, we suggest that the scattering of burst photons by the inner corona may mainly contribute to the persistent emission enhancement. We also observed an inverse correlation between the maximum fa and the persistent emission flux in the non-PRE burst. This anticorrelation suggests that when the accretion rate is lower, there is a greater enhancement of persistent emission during the burst peak. The prediction based on Poynting–Robertson drag (P–R drag) aligns with this observed anticorrelation.

Insight-HXMT observations on thermonuclear X-ray bursts from 4U 1608–52 in 2022: The accretion rate dependent anisotropy of burst emission

Thermonuclear X-ray bursts occur on the surface of an accreting neutron star (NS), and their characteristics and interplay with the surrounding circumstance could be a clue to understand the nature of the NS and accretion process. For this purpose, Insight-HXMT has performed high cadence observations on the bright thermonuclear X-ray burster–4U 1608–52 during its outburst in July and August 2022; nine bursts were detected, including seven bursts with the photospheric radius expansion (PRE). Time-resolved spectroscopy of the bright PRE bursts reveals that an enhancement of accretion rate or the Comptonization of the burst emission by the corona could reduce the residuals when fitting their spectra with the conventional model–blackbody. The inferred energy increment rate of the burst photon gained from the corona is up to ∼40%, even though the bursts have different peak fluxes and locate at different accretion rates. Moreover, the flux shortage of the rising PRE is observed in the bursts at a high mass accretion rate, but not for the burst with a faint persistent emission, which has been predicted theoretically but first observed in this work. If the flux shortage is due to the disk obscuration, i.e., the burst emission is anisotropic, the phenomenon above could indicate that the anisotropy of the burst emission is accretion rate dependent, which could also be evidence of the truncated disk in the low/hard state.

Type I X-ray bursts’ spectra and fuel composition from the atoll and transient source 4U 1730–22

NICER observed two outbursts from the neutron star low-mass X-ray binary 4U 1730–22 in 2021 and 2022, which showed a similar spectral evolution in the hardness-intensity diagram. Seventeen type I X-ray bursts were identified in both outbursts. The X-ray burst spectra showed clear deviations from the blackbody model, firstly ∼10 s after onset. Adding the enhanced persistent emission due to the Poynting-Robertson drag or the reflection from the accretion disk both significantly improved the fitting results. We found that 12 out of 17 X-ray bursts showed the photospheric radius expansion (PRE) characteristic. Considering the nine PRE bursts out of ten X-ray bursts observed by Insight-HXMT, 78% of bursts from 4U 1730–22 exhibited PRE. According to the burst rise time, the duration, the local accretion rate, and the burst fuel composition estimated from recurrence time, we propose that these PRE bursts were powered by pure helium. From the touchdown flux of PRE bursts, we estimate the source distance of d = 7.54 ± 0.46(X = 0) kpc for a canonical neutron star with MNS = 1.4 M⊙ and RNS = 10 km.

Return of 4U 1730–22 after 49 yr Silence: The Peculiar Burst Properties of the 2021/2022 Outbursts Observed by Insight-HXMT

Staying in quiescence for 49 yr, 4U 1730–22 became active and had two outbursts in 2021 and 2022; 10 thermonuclear X-ray bursts were detected with Insight-HXMT. Among them, the faintest burst showed a double-peaked profile, placing the source as the seventh accreting neutron star (NS) exhibiting double-peaked type I X-ray bursts; the other bursts showed photospheric radius expansion (PRE). The properties of a double-peaked non-PRE burst indicate that it could be related to a stalled burning front. For the five bright PRE bursts, apart from the emission from the neutron star (NS) surface, we find the residuals both in the soft (<3 keV) and hard (>10 keV) X-ray bands. Time-resolved spectroscopy reveals that the excess can be attributed to an enhanced preburst/persistent emission or the Comptonization of the burst emission by the corona. We find, the burst emission shows a rise until the photosphere touches down to the NS surface rather than the theoretical predicted constant Eddington luminosity. The shortage of the burst emission in the early rising phase is beyond the occlusion by the disk. We speculate that the findings above are due to that the obscured part (not only the lower part) of the NS surface is exposed to the line of sight due to the evaporation of the obscured material by the burst emission, or the burst emission is anisotropic (ξ > 1) in the burst early phase. In addition, based on the fluxes of PRE bursts at their touchdown times, we derive a distance estimation as 9.0–12.4 kpc.

AstroSat observation of rapid type-I thermonuclear burst from low-mass X-ray binary GX 3$$+$$1

We report the results of an observation of low-mass X-ray binary GX 3 $$+$$ 1 with AstroSat’s large area X-ray proportional counter (LAXPC) and soft X-ray telescope (SXT) instruments on-board for the first time. We have detected type-I thermonuclear burst ( $$\sim $$ 15 s) present in the LAXPC 20 light curve with a double peak feature at higher energies and our study of the hardness–intensity diagram reveals that the source was in a soft banana state. The pre-burst emission could be described well by a thermally Comptonized model component. The burst spectra is modeled adopting a time-resolved spectroscopic method using a single color blackbody model added to the pre-burst model, to monitor the parametric changes as the burst decays. Based on our time-resolved spectroscopy, we claim that the detected burst is a photospheric radius expansion (PRE) burst. During the PRE phase, the blackbody flux is found to be approximately constant at an averaged value of $$\sim $$ 2.56 in $$10^{-8}$$ ergs s $$^{-1}$$ cm $$^{-2}$$ units. On the basis of literature survey, we infer that AstroSat/LAXPC 20 has detected a burst from GX 3 $$+$$ 1 after more than a decade, which is also a PRE one. Utilizing the obtained burst parameters, we provide a new estimation to the source distance, which is $$\sim $$ $$9.3 \pm 0.4$$ kpc, calculated for an isotropic burst emission. Finally, we discuss and compare our findings with the published literature reports.

Insight-HXMT Observation of 4U 1608–52: Evidence of Interplay between a Thermonuclear Burst and Accretion Environment

A Type I burst could influence the accretion process through radiation pressure and Comptonization both for the accretion disk and the corona/boundary layer of an X-ray binary, and vice versa. We investigate the temporal evolution of a bright photospheric radius expansion (PRE) burst of 4U 1608–52 detected by Insight-HXMT in 1–50 keV, with the aim to study the interplay between the burst and persistent emission. Apart from the emission from the neutron star (NS) surface, we find residuals in both the soft (<3 keV) and hard (>10 keV) X-ray bands. Time-resolved spectroscopy reveals that the excess can be attributed to either an enhanced preburst/persistent emission or the Comptonization of the burst emission by the corona/boundary layer. The Comptonization model is a convolution thermal-Comptonization model (thcomp in XSPEC), and the Comptonization parameters are fixed at the values derived from the persistent emission. We find, during the PRE phase, after the enhanced preburst/persistent emission or the Comptonization of the burst emission is removed, the NS surface emission shows a plateau and then a rise until the photosphere touches down on the NS surface, resulting in a flux peak at that moment. We speculate that the findings above correspond to the lower part of the NS surface that is obscured by the disk being exposed to the line of sight due to the evaporation of inner disk by the burst emission. The consistency between the f a model and convolution thermal-Comptonization model indicates the interplay between thermonuclear bursts and accretion environments. These phenomena do not usually show up in conventional blackbody model fittings, which may be due to the low count rate and narrow energy coverage in previous observations.

NICER observations of the evidence of Poynting-Robertson drag and disk reflection during type I X-ray bursts from 4U 1636–536

Type I X-ray bursts are the result of an unstable thermonuclear burning of accreting matter on the neutron star (NS) surface. The quick release of energetic X-ray photons during such bursts interacts with the surrounding accretion disk, which raises the accretion rate due to Poynting-Robertson drag and, thus, a fraction of the burst emission is reflected. We analyzed two photospheric radius expansion bursts in the NS low-mass X-ray binary 4U 1636–536 that took place in 2017, using data from Neutron star Interior Composition Explorer. The time-resolved burst spectra showed clear deviations from a blackbody model. The spectral fitting can be significantly improved by introducing either the enhanced persistent emission (thefamodel) or the reflection from the accretion disk (therelxillNSmodel). Thefamodel provides a higher blackbody temperature and higher burst flux compared with therelxillNSmodel. The peak fluxes of two bursts from thefamodel, 4.36 × 10−8erg cm−2s−1and 9.10 × 10−8erg cm−2s−1, are slightly higher than the Eddington limits of mixed hydrogen-helium and pure helium bursts from previous observations, respectively. When the disk reflections have been taken into account simultaneously, the peak fluxes are lower to match the preferred values. We find evidence to support the finding that both the Poynting-Robertson drag and disk reflection have been appeared during these two X-ray bursts. Moreover, the disk reflection may contribute ∼20−30% of the total burst emissions.

Thermonuclear X-ray bursts from 4U 1636 − 536 observed withAstroSat

ABSTRACTWe report results obtained from the study of 12 thermonuclear X-ray bursts in six AstroSat observations of a neutron star X-ray binary and well-known X-ray burster, 4U 1636 − 536. Burst oscillations (BOs) at ∼ 581 Hz are observed with 4–5σ confidence in three of these X-ray bursts. The rising phase BOs show a decreasing trend of the fractional rms amplitude at 3σ confidence, by far the strongest evidence of thermonuclear flame spreading observed with AstroSat. During the initial 0.25 s of the rise a very high value ($34.0\pm 6.7{{{\ \rm per\ cent}}}$) is observed. The concave shape of the fractional amplitude profile provides a strong evidence of latitude-dependent flame speeds, possibly due to the effects of the Coriolis force. We observe decay phase oscillations with amplitudes comparable to that observed during the rising phase, plausibly due to the combined effect of both surface modes, as well as the cooling wake. The Doppler shifts due to the rapid rotation of the neutron star might cause hard pulses to precede the soft pulses, resulting in a soft lag. The distance to the source estimated using the photospheric radius expansion bursts is consistent with the known value of ∼6 kpc.

Expanded Atmospheres and Winds in Type I X-Ray Bursts from Accreting Neutron Stars

Abstract We calculate steady-state models of radiation-driven super-Eddington winds and static expanded envelopes of neutron stars caused by high luminosities in type I X-ray bursts. We use flux-limited diffusion to model the transition from optically thick to optically thin, and include effects of general relativity, allowing us to study the photospheric radius close to the star as the hydrostatic atmosphere evolves into a wind. We find that the photospheric radius evolves monotonically from static envelopes (r ph ≲ 50–70 km) to winds (r ph ≈ 100–1000 km). Photospheric radii of less than 100 km, as observed in most photospheric radius expansion bursts, can be explained by static envelopes, but only in a narrow range of luminosity. In most bursts, we would expect the luminosity to increase further, leading to a wind with photospheric radius ≳100 km. In the contraction phase, the expanded envelope solutions show that the photosphere is still ≈1 km above the surface when the effective temperature is only 3% away from its maximum value. This is a possible systematic uncertainty when interpreting the measured Eddington fluxes from bursts at touchdown. We also discuss the applicability of steady-state models to describe the dynamics of bursts. In particular, we show that the sub- to super-Eddington transition during the burst rise is rapid enough that static models are not appropriate. Finally, we analyze the strength of spectral shifts in our models. Expected shifts at the photosphere are dominated by gravitational redshift, and are therefore predicted to be less than a few percent.

Gaia EDR3 parallaxes of type I X-ray bursters and their implications on the models of type I X-ray bursts: A generic approach to the Gaia parallax zero point and its uncertainty

Abstract Light curves of photospheric radius expansion (PRE) bursts, a subset of type I X-ray bursts, have been used as standard candles to estimate the ‘nominal PRE distances’ for 63% of PRE bursters (bursters), assuming PRE burst emission is spherically symmetric. Model-independent geometric parallaxes of bursters provide a valuable chance to test models of PRE bursts (PRE models) and can be provided in some cases by Gaia astrometry of the donor stars in bursters. We searched for counterparts to 115 known bursters in the Gaia Early Data Release 3 and confirmed 4 bursters with Gaia counterparts that have detected ( $\gt\!3\,\sigma$ , prior to zero-point correction) parallaxes. We describe a generic approach to the Gaia parallax zero point as well as its uncertainty using an ensemble of Gaia quasars individually determined for each target. Assuming the spherically symmetric PRE model is correct, we refined the resultant nominal PRE distances of three bursters (i.e. Cen $\textrm{X}-4$ , Cyg $\textrm{X}-2$ , and $4\textrm{U}\,0919-54$ ) and put constraints on their compositions of the nuclear fuel powering the bursts. Finally, we describe a method for testing the correctness of the spherically symmetric PRE model using parallax measurements and provide preliminary results.

A catalogue of intermediate-duration Type I X-ray bursts observed with the INTEGRAL satellite

ABSTRACT We present a catalogue of long-duration bursts observed with the Joint European X-ray Monitor and IBIS/ISGRI instruments onboard the INTEGRAL satellite. The 14 bursts have e-folding times ranging from 55 s to ≈17 min, and are therefore classified as intermediate-duration bursts, caused by the ignition of an unusually thick helium layer. Though seven events have already been reported in literature, we have systematically reanalysed the whole sample. We find three new photospheric radius expansion bursts, which are not reported in the literature, allowing us to provide a new estimate of the distances to these sources. We apply the enhanced persistent emission method (also known as the fa method) on sources with detectable persistent emission prior to a burst, in order to follow the evolution of the accretion rate during the burst. Although we do not get significantly better fits, the evolution of the fa factor shows an indicative behaviour, which we discuss.

Insight-HXMT observation on 4U 1608–52: Evolving spectral properties of a bright type-I X-ray burst

The evidences for the influence of thermonuclear (type-I) X-ray bursts upon the surrounding environments in neutron star low-mass X-ray binaries (LMXB) were detected previously via spectral and timing analyses. Benefitting from a broad energy coverage of Insight-HXMT, we analyze one photospheric radius expansion (PRE) burst, and find an emission excess at soft X-rays. Our spectral analysis shows that, such an excess is not likely relevant to the disk reflection induced by the burst emission and can be attributed to an enhanced pre-burst/persistent emission. We find that the burst and enhanced persistent emissions sum up to exceed Eddington luminosity by ∼ 40 percentages. We speculate that the enhanced emission is from a region beyond the PRE radius, or through the Comptonization of the corona.

NICER Discovers Spectral Lines during Photospheric Radius Expansion Bursts from 4U 1820−30: Evidence for Burst-driven Winds

Abstract We report the discovery with the Neutron Star Interior Composition Explorer (NICER) of narrow emission and absorption lines during photospheric radius expansion (PRE) X-ray bursts from the ultracompact binary 4U 1820−30. NICER observed 4U 1820−30 in 2017 August during a low-flux, hard spectral state, accumulating about 60 ks of exposure. Five thermonuclear X-ray bursts were detected, of which four showed clear signs of PRE. We extracted spectra during the PRE phases and fit each to a model that includes a Comptonized component to describe the accretion-driven emission, and a blackbody for the burst thermal radiation. The temperature and spherical emitting radius of the fitted blackbody are used to assess the strength of PRE in each burst. The two strongest PRE bursts (burst pair 1) had blackbody temperatures of ≈0.6 keV and emitting radii of ≈100 km (at a distance of 8.4 kpc). The other two bursts (burst pair 2) had higher temperatures (≈0.67 keV) and smaller radii (≈75 km). All of the PRE bursts show evidence of narrow line emission near 1 keV. By coadding the PRE phase spectra of burst pairs 1 and, separately, 2, we find, in both coadded spectra, significant, narrow, spectral features near 1.0 (emission), 1.7, and 3.0 keV (both in absorption). Remarkably, all the fitted line centroids in the coadded spectrum of burst pair 1 appear systematically blueshifted by a factor of 1.046 ± 0.006 compared to the centroids of pair 2, strongly indicative of a gravitational shift, a wind-induced blueshift, or more likely some combination of both effects. The observed shifts are consistent with this scenario in that the stronger PRE bursts in pair 1 reach larger photospheric radii, and thus have weaker gravitational redshifts, and they generate faster outflows, yielding higher blueshifts. We discuss possible elemental identifications for the observed features in the context of recent burst-driven wind models.

Simultaneous Constraints on the Mass and Radius of Aql X–1 from Quiescence and X-Ray Burst Observations

Abstract The measurement of neutron star mass and radius is one of the most direct ways to distinguish between various dense matter equations of state. The mass and radius of accreting neutron stars hosted in low-mass X-ray binaries can be constrained by several methods, including photospheric radius expansion from type I X-ray bursts and from quiescent spectra. In this paper, we apply for the first time these two methods simultaneously to constrain the mass and radius of Aql X–1. The quiescent spectra from Chandra and XMM-Newton, and photospheric radius expansion bursts from RXTE are used. The determination of the mass and radius of Aql X–1 is also used to verify the consistency between the two methods and to narrow down the uncertainties of the neutron star mass and radius. It is found that the distance to Aql X–1 should be in the range of 4.0–5.75 kpc, based on the overlapping confidence regions between photospheric radius expansion burst and quiescent spectra methods. In addition, we show that the mass and radius determined for the compact star in Aql X–1 are compatible with strange star equations of state and conventional neutron star models.

The direct cooling tail method for X-ray burst analysis to constrain neutron star masses and radii

Determining neutron star (NS) radii and masses can help to understand the properties of matter at supra-nuclear densities. Thermal emission during thermonuclear X-ray bursts from NSs in low-mass X-ray binaries provides a unique opportunity to study NS parameters, because of the high fluxes, large luminosity variations and the related changes in the spectral properties. The standard cooling tail method uses hot NS atmosphere models to convert the observed spectral evolution during cooling stages of X-ray bursts to the Eddington flux F_Edd and the stellar angular size \Omega. These are then translated to the constraints on the NS mass M and radius R. Here we present the improved, direct cooling tail method that generalises the standard approach. First, we adjust the cooling tail method to account for the bolometric correction to the flux. Then, we fit the observed dependence of the blackbody normalization on flux with a theoretical model directly on the M-R plane by interpolating theoretical dependences to a given gravity, hence ensuring only weakly informative priors for M and R instead of F_Edd and \Omega. The direct cooling method is demonstrated using a photospheric radius expansion burst from SAX J1810.8--2609, which has happened when the system was in the hard state. Comparing to the standard cooling tail method, the confidence regions are shifted by 1 sigma towards larger radii, giving R=11.5-13.0 km at M=1.3-1.8 M_sun for this NS.

THE FERMI–GBM THREE-YEAR X-RAY BURST CATALOG

ABSTRACT The Fermi Gamma-ray Burst Monitor (GBM) is an all-sky gamma-ray monitor well known in the gamma-ray burst (GRB) community. Although GBM excels in detecting the hard, bright extragalactic GRBs, its sensitivity above 8 keV and its all-sky view make it an excellent instrument for the detection of rare, short-lived Galactic transients. In 2010 March, we initiated a systematic search for transients using GBM data. We conclude this phase of the search by presenting a three-year catalog of 1084 X-ray bursts. Using spectral analysis, location, and spatial distributions we classified the 1084 events into 752 thermonuclear X-ray bursts, 267 transient events from accretion flares and X-ray pulses, and 65 untriggered gamma-ray bursts. All thermonuclear bursts have peak blackbody temperatures broadly consistent with photospheric radius expansion (PRE) bursts. We find an average rate of 1.4 PRE bursts per day, integrated over all Galactic bursters within about 10 kpc. These include 33 and 10 bursts from the ultra-compact X-ray binaries 4U 0614+09 and 2S 0918-549, respectively. We discuss these recurrence times and estimate the total mass ejected by PRE bursts in our Galaxy.

Neutron star radii, universal relations, and the role of prior distributions

We investigate constraints on neutron star structure arising from the assumptions that neutron stars have crusts, that recent calculations of pure neutron matter limit the equation of state of neutron star matter near the nuclear saturation density, that the high-density equation of state is limited by causality and the largest high-accuracy neutron star mass measurement, and that general relativity is the correct theory of gravity. We explore the role of prior assumptions by considering two classes of equation of state models. In a first, the intermediate- and high-density behavior of the equation of state is parameterized by piecewise polytropes. In the second class, the high-density behavior of the equation of state is parameterized by piecewise continuous line segments. The smallest density at which high-density matter appears is varied in order to allow for strong phase transitions above the nuclear saturation density. We critically examine correlations among the pressure of matter, radii, maximum masses, the binding energy, the moment of inertia, and the tidal deformability, paying special attention to the sensitivity of these correlations to prior assumptions about the equation of state. It is possible to constrain the radii of $1.4~\mathrm{M}_{\odot}$ neutron stars to a be larger than 10 km, even without consideration of additional astrophysical observations, for example, those from photospheric radius expansion bursts or quiescent low-mass X-ray binaries. We are able to improve the accuracy of known correlations between the moment of inertia and compactness as well as the binding energy and compactness. We also demonstrate the existence of a correlation between the neutron star binding energy and the moment of inertia.

4U 1746–37: An ultra‐low‐mass compact star candidate

AbstractThe Rossi X‐ray Timing Explorer observed three photospheric radius expansion bursts in 4U 1746‐37, all with low touchdown fluxes. We discuss the possibiity of a low‐mass neutron star in 4U 1746‐37 if the touchdown flux corresponds to its Eddington limit. 4U 1746‐37 is a dipping binary system which has a high inclination angle. Two geometric effects, the reflection of the far side accretion disc and the obscuration of the near side accretion disc have also been included. We apply a Monte‐Carlo simulation, with typical values of hydrogen mass fraction and color correction factor, to constrain the mass and radius of the neutron star in 4U 1746‐37. We found that the mass of 4U 1746‐37 is 0.41+0.70–0.30 M® at 99.7 % confidence which is an ultra‐low‐mass compact star candidate. It could be reproduced by a self‐bound compact star, i.e., quark star or quark‐cluster star, from an ccretion‐induced collapse process. (© 2015 WILEY‐VCH Verlag GmbH &amp; Co. KGaA, Weinheim)

X-ray softening during the 2008 outburst of XTE J1810-189

XTE J1810-189 underwent an outburst in 2008, and was observed over $\sim 100$ d by RXTE. Performing a time-resolved spectral analysis on the photospheric radius expansion burst detected on 2008 May 4, we obtain the source distance in the range of 3.5--8.7 kpc for the first time. During its outburst, XTE J1810-189 did not enter into the high/soft state, and both the soft and hard colours decreased with decreasing flux. The fractional rms remained at high values ($\sim 30$ per cent). The RXTE/PCA spectra for 3-25 keV can be described by an absorbed power-law component with an additional Gaussian component, and the derived photon index $\Gamma$ increased from $1.84\pm0.01$ to $2.25\pm0.04$ when the unabsorbed X-ray luminosity in 3-25 keV dropped from $4\times10^{36}$ ergs s$^{-1}$ to $6\times10^{35}$ ergs s$^{-1}$. The relatively high flux, dense observations and broadband spectra allow us to provide strong evidence that the softening behaviour detected in the outburst of XTE J1810-189 originates from the evolution of non-thermal component rather than the thermal component (i.e. neutron star surface emission).