1: %\documentstyle[epsfig]{aastex}
2:
3: \documentclass[useAMS, usenatbib]{mn2e}
4:
5: %\documentclass[8pt]{aastex}
6: %\documentclass[]{emulateapj}
7:
8: %\usepackage{showkeys}
9: %\documentclass{aa}
10: %\usepackage{epsfig}
11: %\usepackage{graphicx}
12: %\usepackage{longtable}
13: \usepackage{rotating}
14: %\documentclass[12pt]{article}
15: %\usepackage{psfig}
16: %\usepackage{palatino, multicol}
17: \usepackage{natbib}
18: \usepackage{latexsym}
19: \usepackage{amssymb}
20: \bibpunct{(}{)}{;}{a}{}{,}
21: %posat per mi:
22: %\usepackage{subfigure}
23:
24: \def\rem#1{{\bf (#1)}}
25: \def\hide#1{}
26: \def \aj {AJ}
27: \def \mnras {MNRAS}
28: \def \apj {ApJ}
29: \def \apjs {ApJS}
30: \def \apjl {ApJL}
31: \def \aap {A\&A}
32: \def \nat {Nature}
33: \def \araa {ARAA}
34: \def \pasp {PASP}
35: \def \aaps {AAPS}
36: \def \prd {PhRvD}
37:
38:
39: \voffset=-0.8in
40:
41: \begin{document}
42:
43: \author[Linares et al.]{\parbox[t]{\textwidth}{\raggedright Manuel
44: Linares$^1$\thanks{linares@uva.nl}, Anna Watts$^1$, Rudy
45: Wijnands$^1$, Paolo Soleri$^1$, Nathalie Degenaar$^1$, Peter A.
46: Curran$^1$, Rhaana L. C. Starling$^2$ and Michiel van der
47: Klis$^1$}\\
48: \vspace*{6pt}\\
49: $^1$ Astronomical Institute ``Anton Pannekoek'', University of
50: Amsterdam and Center for High-Energy Astrophysics,
51: \\ Kruislaan 403,
52: NL-1098 SJ Amsterdam, Netherlands
53: \\ $^2$ Department of Physics and Astronomy, University of Leicester,
54: University Road, Leicester LE1 7RH, UK }
55:
56:
57: \title[Long burst XTE~J1701--407]{The {\it Swift} capture of a long X-ray burst from XTE~J1701--407}
58:
59: %\date{}
60:
61: \maketitle
62:
63: \begin{abstract}
64:
65: XTE~J1701--407 is a new transient X-ray source discovered on June 8th,
66: 2008. More than one month later it showed a rare type of thermonuclear
67: explosion: a long type I X-ray burst. We report herein the results of
68: our study of the spectral and flux evolution during this burst, as
69: well as the analysis of the outburst in which it took place. We find
70: an upper limit on the distance to the source of 6.1 kpc by considering
71: the maximum luminosity reached by the burst. We measure a total
72: fluence of 3.5$\times 10^{-6}$ erg/cm$^2$ throughout the $\sim$20
73: minutes burst duration and a fluence of 2.6$\times 10^{-3}$ erg/cm$^2$
74: during the first two months of the outburst. We show that the flux
75: decay is best fitted by a power law (index $\sim$1.6) along the tail
76: of the burst. Finally, we discuss the implications of the long burst
77: properties, and the presence of a second and shorter burst detected by
78: {\it Swift} ten days later, for the composition of the accreted material and
79: the heating of the burning layer.
80:
81: \end{abstract}
82:
83: \begin{keywords}
84: binaries: close --- X-rays: bursts --- stars: neutron --- X-rays: binaries
85: \end{keywords}
86:
87:
88:
89:
90: \section{Introduction}
91: \label{sec:intro}
92:
93: The bulge of our Galaxy harbours a large fraction of the known
94: population of accreting compact objects. During one of the regular
95: monitoring observations of that region with the proportional counter
96: array (PCA) onboard the Rossi X-ray timing explorer ({\it RXTE}) on June
97: 8th, 2008, a new transient X-ray source was discovered: XTE~J1701--407
98: \citep{Markwardt08}. Three days later a follow-up observation with the
99: X-ray telescope (XRT) onboard {\it Swift} improved the source localization
100: and showed a spectrum that could be fitted with an absorbed power law
101: of index $\sim$2 \citep{Degenaar08}. The transient nature, flux level
102: and spectrum of the source are characteristic of X-ray binaries, but
103: the exact class of the binary (high-mass or low-mass) was unkown and
104: the compact object (black hole or neutron star) remained unidentified
105: for more than one month. On July 17th at 13:29:59 UT the burst alert
106: telescope (BAT) onboard {\it Swift} detected an X-ray flare at a position
107: consistent with that of XTE~J1701--407 \citep{Barthelmy08}. {\it Swift}
108: began slewing to the source 60 seconds after the trigger and the XRT
109: began observing the field 97 seconds after the initial BAT trigger,
110: detecting a fading X-ray source at the position of XTE~J1701--407
111: \citep{Barthelmy08}. \citet{Markwardt08b} suggested that the flare
112: could be caused by a thermonuclear burst based on the early BAT
113: data. By studying the evolution of the XRT spectrum, \citet{Linares08}
114: found a clear cooling curve and confirmed the speculation that this
115: was a thermonuclear event, identifying the source as an accreting
116: neutron star (NS) and the system, in all likelihood, as a low-mass
117: X-ray binary (LMXB). As noted by \citet{Linares08} this particular
118: event, having a total duration of more than 15 minutes, belongs to the
119: rare subclass of long duration bursts \citep{Cumming06}.
120:
121: After the long X-ray burst the source continued in
122: outburst\footnote{To clarify the terminology we stress that in the
123: context of LMXBs an outburst is powered by accretion and lasts for
124: weeks to years whereas a type I X-ray burst, or ``burst'', results
125: from unstable thermonuclear burning on the surface of a NS and lasts
126: for $\sim$10 seconds to $\sim$hours.} and {\it RXTE} kept observing it. A
127: refined {\it Swift}-XRT position \citep{Starling08}, a search for an IR
128: counterpart \citep{Kaplan08} and the discovery of kHz quasi-periodic
129: oscillations \citep[QPOs;][]{Strohmayer08} were all reported during
130: the following two weeks. On July 27th BAT detected another X-ray flare
131: from XTE~J1701--407 \citep{Sakamoto08}. The soft spectrum and duration
132: ($\sim$10~s) of the flare indicated that this was a type I X-ray
133: burst (http://gcn.gsfc.nasa.gov/notices\_s/318166/BA/), and
134: hereafter we refer to it as the short burst (even though from the BAT
135: data the total duration is somewhat uncertain). At the moment of
136: writing the source is still active, more than two months after the
137: start of the outburst.
138:
139: In this Letter we present our analysis of the bursting behaviour of
140: this source. Long bursts are rare, and therefore probe quite unusual
141: burning regimes. Their duration is between that of normal Type I X-ray
142: bursts (durations $\sim 10-100$ s, triggered by unstable burning of
143: H/He) and superbursts (durations $\sim$ hours, triggered by unstable
144: burning of C). There are two scenarios in which such bursts could
145: occur, both requiring the build-up and ignition of a thick layer of
146: He. One possibility is that the system is ultracompact, so that it
147: accretes nearly pure He \citep{Cumming06}. The other possibility is
148: that unstable H burning at low accretion rates builds up He
149: \citep{Peng07, Cooper07}. While several examples of the first type are
150: now known (see for example \citealt{intZand07, Falanga08} and
151: references therein), the second class - which probe the boundaries of
152: H burning stability - have so far proved much rarer
153: \citep{Chenevez07}. In this Letter we will demonstrate that XTE
154: J1701-407 is a good candidate for membership of this second class.
155:
156:
157: \section{Observations and Data Analysis}
158: \label{sec:data}
159:
160: We analysed all the 27 pointed {\it RXTE}-PCA observations of
161: XTE~J1701--407 taken until August 10th, 2008 (ID 93444-01). We also
162: analysed the {\it Swift}-XRT data taken during and after the long
163: burst (on July 17th, obsids 00317205000 and 00317205001). {\it Swift}
164: observed XTE~J1701--407 again on July 27th, triggered by the BAT
165: detection of the short burst (Sec.~\ref{sec:intro}). We also included
166: this XRT observation (obsid 00318166000) in our analysis, which
167: started $\sim$109 seconds after the trigger but did not detect the
168: burst. In the following two sections we provide the details of our
169: analysis. All errors given in Section~\ref{sec:results} correspond to
170: the 1~$\sigma$ confidence level.
171:
172: \subsection{Swift}
173: \label{sec:dataswift}
174:
175: We ran the XRT pipeline (v. 0.12.0) on all the {\it Swift}
176: observations. For the spectral analysis we created exposure maps and
177: ancilliary files and used the latest response matrices (v011) provided
178: by the {\it Swift}-XRT team. We grouped the spectra to a minimum of 20
179: counts per bin, applyied a systematic error of 2.5\% \citep{Campana08}
180: and fitted the 0.5-10~keV spectra within Xspec
181: \citep[][v. 11]{Arnaud96}. In our fits (see models below) we fixed the
182: absorbing column density to the average value found from X-ray fits to
183: the persistent emission: 3.5$\times 10^{22}$ cm$^{-2}$
184: \citep[][]{Degenaar08,Markwardt08}.
185:
186: Observation 00317205000 started on July 17th, 2008 at 13:31:36.9 UT
187: (Sec.~\ref{sec:intro}) and lasted for a total of about one
188: kilosecond. All the XRT data were collected in windowed timing (WT)
189: mode, except for the initial $\sim$8 seconds which were taken in
190: photon counting (PC) mode. We divided the observation into nine
191: contiguous intervals of increasing length in order to compensate for
192: the flux decay and maintain an approximately constant total number of
193: counts in each interval (between $\sim$1500 and $\sim$2500). About 830
194: seconds after the start of the observation a data gap of $\sim$3300
195: seconds occurred, after which the XRT continued to observe the source
196: for another $\sim$35 seconds. This last data segment allows us to
197: estimate the persistent flux after the X-ray burst. We extracted
198: source and background spectra for each interval using circular regions
199: with radii of $\sim$30 and $\sim$15 pixels, respectively, and fitted
200: the resulting spectra using an absorbed black body model (reduced
201: $\chi^2$ between 0.7 and 1.1).
202:
203: We performed fast Fourier transforms on 2s segments of XRT-WT data
204: taken during the long burst in order to search for rapid
205: variability. For that purpose we used the energy range 0.5-10 keV,
206: keeping the original time resolution and thereby sampling the
207: $\sim$0.5-280~Hz frequency range.
208:
209: Both post-burst observations (00317205001 and 00318166000) allow us to
210: characterize the persistent soft X-ray emission of XTE~J1701--407. For
211: the PC data we used an annular extraction region with radii 10-25
212: pixels in order to account for pile-up. The energy fraction enclosed
213: by the resulting annuli was about 20\%, and such energy loss was taken
214: into account when creating the ancilliary files. Finally, including
215: also the WT data, we fitted the energy spectra with an absorbed power
216: law model (reduced $\chi^2$ of 1.1 and 1.2).
217:
218:
219: \subsection{RXTE}
220: \label{sec:datarxte}
221:
222: We analysed all the pointed {\it RXTE} observations of the source made
223: until August 10th. After applying standard filters we extracted energy
224: spectra from Standard 2 data collected by PCU2 and corrected them for
225: background using the latest PCA background models. We generated
226: reponse matrices and fitted the spectra within Xspec in the 3.0-25.0
227: keV band, after applying a 1\% systematic error, with a black body
228: plus power law model corrected for absorption (see
229: Sec.~\ref{sec:dataswift}). The resulting fits had reduced $\chi^2$
230: between 0.5 and 1.4 (for 45 degrees of freedom). We extracted a
231: hardness ratio (``hard colour'' hereafter) for each observation,
232: defined as the count rate in the 9.7-16.0 keV band divided by that in
233: the 6.0-9.7 keV band and normalized by that of the Crab. Furthermore,
234: we extracted a background and deadtime corrected HEXTE (cluster B)
235: energy spectrum from each observation, but found no significant source
236: emission above 50 keV in any of them.
237:
238: \section{Results}
239: \label{sec:results}
240:
241: \subsection{Outburst}
242: \label{sec:outburst}
243:
244: The overall lightcurve is shown in Figure~\ref{fig:lc}, together with
245: the evolution of the hard colour. The outburst shows two peaks at a
246: similar flux, (6.2, 6.8)$\times 10^{-10}$ erg/cm$^2$/s. The BAT
247: transient monitor
248: lightcurves\footnote{http://swift.gsfc.nasa.gov/docs/swift/results/transients/}
249: indicate that the source was active during the three weeks where no
250: {\it RXTE} pointed observations were obtained, at a flux level similar
251: to that measured during the first two weeks of outburst (see
252: Fig.~\ref{fig:lc}). The hard colour traces directly the changes in
253: spectral state along the outburst, with a clear tendency: when the
254: source is bright its spectrum is soft whereas the spectrum is harder
255: at the lowest fluxes. However, it is interesting to note that as is
256: the case for many other NS-LMXBs the correspondence between flux and
257: hardness is not one-to-one \citep{vanderKlis06}. Together with the
258: presence of kHz quasi-periodic oscillations \citep{Strohmayer08}, the
259: peak luminosity and the spectral evolution (Fig.~\ref{fig:lc})
260: show that XTE~J1701--407 is a new member of the atoll
261: source class \citep{HK89}. Both X-ray bursts occur at relatively high
262: fluxes (see Table~\ref{table:bursts}) and soft spectra (low hard
263: colour).
264:
265: \begin{figure}
266: \begin{center}
267: \resizebox{1.0\columnwidth}{!}{\rotatebox{-90}{\includegraphics[]{./lc_wbat.ps}}}
268: \caption{Time evolution of the 2-50 keV unabsorbed flux {\it (top, filled dots)} and the hard colour {\it (bottom)} during the outburst of XTE~J1701--407, as measured by the {\it RXTE}-PCA. Arrows indicate the times of the long (LB) and short (B) X-ray burst. The 15-50 keV flux measured by {\it Swift}-BAT is shown as well {\it (top, grey rectangles)}.}
269: \label{fig:lc}
270: %\epsscale{1.0}
271: \end{center}
272: \end{figure}
273:
274:
275: By interpolating linearly between the available datapoints we
276: integrate the lightcurve and find a total outburst fluence up until
277: August 10th of 2.6$\times 10^{-3}$ erg/cm$^2$. More than half of this
278: energy (1.6$\times 10^{-3}$ erg/cm$^2$) was radiated before the long
279: burst occurred, whereas the fluence between long and short bursts was
280: about 5$\times 10^{-4}$ erg/cm$^2$.
281:
282: From our spectral fits to the post-burst XRT observations
283: (Sec.~\ref{sec:dataswift}) we measure 2-10~keV fluxes of
284: ($4.0\pm0.9)\times 10^{-10}$ erg/cm$^2$/s (July 17th) and
285: ($5.6\pm1.0)\times 10^{-10}$ erg/cm$^2$/s (July 27th), which are
286: consistent with the fluxes in the same band obtained from the nearest
287: (within one day) PCA observations. In both cases the photon index was
288: $\sim$2. Furthermore, we estimate unabsorbed 1-2~keV fluxes of
289: $2.3\times 10^{-10}$ erg/cm$^2$/s (July 17th) and $3.5\times 10^{-10}$
290: erg/cm$^2$/s (July 27th) and add them to the unabsorbed 2-50~keV
291: fluxes measured with the PCA within one day of the bursts ($6.3\times
292: 10^{-10}$ erg/cm$^2$/s for both bursts; see Fig.~\ref{fig:lc}). These
293: 1-50~keV fluxes (Table~\ref{table:bursts}) are the closest we can
294: safely get to the bolometric persistent flux at the times of the
295: bursts, but we note that the uncertainties in the absorbing column
296: density and the extrapolation to the 0.01-1~keV energy range could
297: change the bolometric flux by a factor of $\sim$2 \citep[see
298: e.g. discussion in][]{intZand07}.
299:
300:
301: \subsection{X-ray bursts}
302: \label{sec:burst}
303:
304: From the flux and spectral parameters reported by \citet{Markwardt08b}
305: we estimate a bolometric unabsorbed peak flux of 8.4$\times 10^{-8}$
306: erg/cm$^2$/s for the long burst. By assuming that the peak luminosity
307: was equal to or lower than that typical of photospheric expansion type
308: I X-ray bursts \citep[$\sim$3.8$\times 10^{38}$ erg/s][]{Kuulkers03}
309: we can place an upper limit on the distance to the source of 6.1
310: kpc. On the other hand, if we use the Eddington limit for a mix of H
311: and He ($\sim$1.6$\times 10^{38}$ erg/s for a hydrogen fraction X=0.7)
312: as a limit for the peak luminosity the upper limit on the distance
313: becomes 4.0 kpc. The rise time in the BAT lightcurve was about 50
314: seconds, the peak occurred about ten seconds after the
315: trigger (http://gcn.gsfc.nasa.gov/notices\_s/317205/BA/) and
316: the total duration, assuming that the burst finished when it reached
317: 1\% of the peak flux, was about 21 min. The fluxes in the post-burst
318: XRT observation, taken more than two hours after the peak of the long
319: burst, are similar to the flux we measure in the last interval
320: ($\sim$1 hour after the peak, Sec.~\ref{sec:dataswift}), confirming
321: that the source had returned to the persistent level.
322:
323: Figure~\ref{fig:burst} shows the results of our time-resolved
324: spectroscopy of the tail of the long burst. The flux decays by almost
325: two orders of magnitude and the temperature decreases from
326: 1.72$\pm$0.05~keV to 0.88$\pm$0.02~keV during the burst tail \citep[see
327: also][]{Linares08}. XRT did not observe the peak of the burst, and no
328: spectral variations were seen in BAT data of that period
329: \citep{Markwardt08b}. Therefore the question of whether or not the
330: burst showed photospheric radius expansion cannot be addressed. Going
331: back to our XRT measurements, the radius of the black body appears to
332: increase between $\sim$200~s and $\sim$500~s after the peak
333: (Fig.~\ref{fig:burst}). However, the large error affecting this
334: parameter, the uncertainties on the underlying model (NS atmosphere in
335: a dramatically-non-steady state) and the absence of an associated drop
336: in the black body temperature imply that this cannot be taken as
337: concrete evidence of a physical increase in emitting area.
338:
339: \begin{figure}
340: \begin{center}
341: \resizebox{0.6\columnwidth}{!}{\rotatebox{0}{\includegraphics[]{./burst.ps}}}
342: \caption{Time evolution of the bolometric flux {\it(top)}, black body temperature {\it(middle)} and black body radius {\it(bottom)} along the long X-ray burst. Filled dots correspond to our {\it Swift}-XRT measurements whereas empty squares show {\it Swift}-BAT results \citep{Markwardt08b}. The horizontal dotted line on the top panel indicates the persistent flux. Error bars in the x axis represent the averaged intervals. The distance upper limit of 6.1 kpc has been used to obtain the black body radius.}
343: \label{fig:burst}
344: %\epsscale{1.0}
345: \end{center}
346: \end{figure}
347:
348:
349:
350: We fit the flux decay of the long burst measured by the XRT and find
351: that a simple power law with an index of -1.62$\pm$0.04 and
352: normalization (3.0$\pm$0.6)$\times 10^{-5}$ erg/cm$^2$/s gives a
353: remarkably good description of the data (reduced $\chi^2$ of 1.2;
354: see Fig.~\ref{fig:decay}). A simple or double exponential decay do not
355: fit the data satisfactorily (reduced $\chi^2$ of 22 and 7,
356: respectively). We note that power law cooling is consistent with the
357: expectations from models for long bursts \citep{Cumming04}.
358:
359: The bolometric fluence measured by XRT during the long burst, after
360: subtracting the persistent flux level, is 2.2$\times 10^{-6}$
361: erg/cm$^2$. We integrate the flux decay along the data gap
362: (Sec.~\ref{sec:dataswift}) and estimate a fluence of 4.4$\times
363: 10^{-7}$ erg/cm$^2$ during that period. Adding the reported BAT
364: fluence \citep{Markwardt08b} we find a total burst fluence of
365: 3.5$\times 10^{-6}$ erg/cm$^2$, which using the 6.1 kpc upper limit on
366: the distance corresponds to a maximum radiated energy of
367: 1.6$\times 10^{40}$ erg.
368:
369: We searched the XRT power spectra obtained during the tail of the long
370: burst (Sec.~\ref{sec:dataswift}) and found no strong pulsations or
371: QPOs. This non-detection is not surprising given i) the distribution
372: of burst oscillation frequencies \citep{Galloway06} combined with our
373: Nyquist frequency ($\sim$280 Hz) and ii) the relatively low count rate
374: collected by the XRT: for the maximum rate of $\sim$90 c/s, a
375: 3$\sigma$ detection of a 2~Hz wide QPO in a 2s FFT would require a
376: fractional rms of at least $\sim$ 25\% \citep[][]{vanderKlis95b}, much
377: higher than the values usually measured \citep{Galloway06}.
378:
379:
380: \section{Discussion}
381: \label{sec:discussion}
382:
383:
384: Theoretical modelling of long bursts suggests that
385: they involve the build-up and ignition of a thick layer of He. One
386: possibility is that the binary is ultracompact, so that the accreted
387: fuel is almost pure He. At low accretion rates, and in the absence of
388: heating from steady H burning, the temperature remains low and ignition
389: can be delayed until a thick layer of He has built up
390: \citep{intZand05, Cumming06}. Although He burning is quick, the
391: cooling time is long because of the thickness of the layer in which
392: the heat is deposited. Long bursts can also be triggered in systems
393: accreting a mix of H and He if the source accretes at low rates, close
394: to the point where H burning stabilises \citep{Fujimoto81, Peng07,
395: Cooper07}. In this case weak H flashes can build up He, which ignites
396: sporadically, resulting in a long burst. Layer thickness again
397: results in a long cooling time: but the presence of H in the burning
398: mix may also prolong nuclear energy generation.
399:
400:
401: \begin{figure}
402: \begin{center}
403: \resizebox{0.6\columnwidth}{!}{\rotatebox{0}{\includegraphics[]{./decay.ps}}}
404: \caption{Fit to the long burst flux decay with a power law of index $\sim$1.6, together with the values of the baseline-subtracted bolometric flux.}
405: \label{fig:decay}
406: %\epsscale{1.0}
407: \end{center}
408: \end{figure}
409:
410:
411: Using the measured fluxes $F$ from Table \ref{table:bursts} and the upper limit on the distance $d$ that we
412: obtained from the long burst, we infer an upper limit on the global accretion rate $\dot{M} \approx
413: 4\pi d^2 R F/GM$ of
414: 3.7$\times 10^{-10}$ $M_\odot$ yr$^{-1}$ (for a neutron star mass $M =
415: 1.4 M_\odot$ and radius $R = 10$ km) around the times of the two bursts.
416: Assuming isotropy, this
417: corresponds to an upper limit on the local accretion rate $\dot{m}$ of 1.9$\times 10^{3}$ g/cm$^2$/s,
418: (2.5\% of the Eddington rate for an Eddington luminosity of 1.6$\times 10^{38}$ erg/s).
419: This accretion rate is
420: close to the rate where H burning is expected to transition
421: to stability \citep{Fujimoto81, Peng07, Cooper07}, so we need to examine both scenarios for the
422: generation of the long burst.
423:
424:
425: The BAT rise time of the long burst, $\approx 50$ s, is much longer
426: than most of the pure He bursts examined by \citet{intZand07} and
427: \citet{Falanga08}. This points to it being a mixed H/He type burst,
428: triggered by H ignition, although we note that burst lightcurves can
429: change considerably depending on the energy range where they are observed
430: \citep[see][]{Chelovekov06,Molkov05}.
431:
432: We next consider the ignition conditions. The long burst had a total
433: energy output of at most $E_b = 1.6\times 10^{40}$ ergs (6.1 kpc
434: distance). The corresponding ignition column depth $y = E_b
435: (1+z)/4\pi R^2 Q_\mathrm{nuc}$, where $z$ is the redshift, $R$ the
436: neutron star radius and $Q_\mathrm{nuc} = 1.6 + 4 X$ MeV/nucleon the
437: nuclear energy release, given an H fraction X at ignition
438: \citep{Galloway06}. Assuming $z=0.31$ (the redshift for a 1.4
439: $M_\odot$, 10 km radius neutron star) we derive an ignition depth $y =
440: 3.9\times 10^8$ g/cm$^2$ for solar abundances (X=0.7) and $y =
441: 1.1\times 10^9 $ g/cm$^2$ for pure He (X=0). Ignition of pure He at
442: this column depth would require a heat flux from the deep crust of
443: $\gtrsim$ 2 MeV/nucleon at 1\% Eddington accretion rates (Figure 22,
444: \citealt{Cumming06}). This is higher than the crustal energy release
445: of $\sim$ 1.5 MeV/nucleon calculated by
446: \citet{haensel1990,haensel2003}, but close to the value of $\sim$ 1.9
447: MeV/nucleon recently found by \citet{gupta2007,haensel2008}. Recent
448: results by \citet{horowitz2008} demonstrate that even larger amounts
449: of heat may be released in the crust of an accreting neutron
450: star. However, the high heat requirements certainly put some strong
451: constraints on the pure He accretion scenario. Using the maximum local accretion
452: rate calculated above (1.9$\times 10^{3}$ g/cm$^2$/s), these ignition conditions
453: predict minimum recurrence times $\Delta t_\mathrm{rec} = y(1+z)/\dot{m}$ of
454: $\sim$10 days (X = 0) and $\sim$3 days (X = 0.7). The data provide
455: only weak constraints on the time without bursts that elapsed before
456: the long burst (we estimate that the source was in the field of view
457: of BAT for about 20\% of the two months analysed herein), but these
458: numbers are at least compatible with the time since the start of the
459: outburst.
460:
461:
462: \begin{table}
463: \center
464: \footnotesize
465: \caption{Properties of both type I X-ray bursts detected from XTE~J1701--407}
466: \begin{minipage}{\textwidth}
467: \begin{tabular}{ l c c}
468: \hline\hline
469: & Long & Short \\
470: \hline
471: BAT rise time (s) & 50 & 5 \\
472: Duration\footnote{From start of rise to 1\% of peak flux.} (min) & 21 & ? \\
473: Peak flux\footnote{Bolometric} ($10^{-8}$ erg/s/cm$^2$) & 8.4 & $\sim$4.9 \\
474: Total fluence\footnote{Including fluence in BAT \citep{Markwardt08b} and \\ during XRT data gap.} ($10^{-6}$ erg/cm$^2$) & 3.5 & ? \\
475: Persistent flux\footnote{Unabsorbed 1-50~keV} ($10^{-10}$ erg/s/cm$^2$) & 8.6$\pm$0.9 & 9.8$\pm$1.0 \\
476: \hline\hline
477: \end{tabular}
478: \end{minipage}
479: \label{table:bursts}
480: \end{table}
481:
482:
483: The presence of a short, weaker burst at similar accretion rates is
484: perhaps the strongest piece of evidence pointing to mixed H/He fuel.
485: The short burst has a relatively slow rise time, $\approx 5$ s,
486: similar to that expected for mixed H/He bursts. The fact that it has
487: a lower peak flux than the long burst also implies some H content: if
488: the system were a pure He accretor we would expect both bursts to
489: reach similar peak fluxes since He bursts, whether long or short, are
490: expected to exhibit photospheric radius expansion \citep{Cumming04b}.
491: In the H-triggered burst scenario it is possible to have bursts with
492: quite different properties at very similar accretion rates, as H
493: burning transitions to stability \citep{Cooper07}.
494:
495: XTE~J1701--407 increases the number of known NS-LMXBs in the
496: interesting low accretion rate bursting regime. The balance of
497: evidence suggests that it may be a good probe of thermonuclear burning
498: near the boundary of stable H ignition. However, further observations,
499: in particular better constraints on recurrence times and burst
500: energetics, are required to confirm this picture. Identification of
501: the companion star and determination of the orbital parameters may
502: also be able to confirm whether or not the system is ultracompact.
503:
504: \section*{Acknowledgements}
505:
506: ML would like to thank C. Markwardt and A. Cumming for stimulating
507: discussions and N. Rea, E. Mart{\' i}nez and Ll. Guasch for their
508: encouraging comments. RLCS acknowledges support from STFC.
509:
510: \bibliographystyle{mn2e}
511: %\bibliography{biblio}
512: \begin{thebibliography}{}
513:
514: \bibitem[\protect\citeauthoryear{{Arnaud}}{{Arnaud}}{1996}]{Arnaud96}
515: {Arnaud} K.~A., 1996, in {Jacoby} G.~H., {Barnes} J., eds, Astronomical Data
516: Analysis Software and Systems V Vol.~101 of Astronomical Society of the
517: Pacific Conference Series, {XSPEC: The First Ten Years}.
518: pp 17--+
519:
520: \bibitem[\protect\citeauthoryear{{Barthelmy} et~al.,}{{Barthelmy}
521: et~al.}{2008}]{Barthelmy08}
522: {Barthelmy} S., et~al., 2008, GCN, 7985, 1
523:
524: \bibitem[\protect\citeauthoryear{{Campana}, {Panagia}, {Lazzati}, {Beardmore},
525: {Cusumano}, {Godet}, {Chincarini}, {Covino}, {Della Valle}, {Guidorzi},
526: {Malesani}, {Moretti}, {Perna}, {Romano} \& {Tagliaferri}}{{Campana}
527: et~al.}{2008}]{Campana08}
528: {Campana} S., {Panagia} N., {Lazzati} D., {Beardmore} A.~P., {Cusumano} G.,
529: {Godet} O., {Chincarini} G., {Covino} S., {Della Valle} M., {Guidorzi}
530: C., {Malesani} D., {Moretti} A., {Perna} R., {Romano} P.,
531: {Tagliaferri} G., 2008, \apjl, 683, L9
532:
533: \bibitem[\protect\citeauthoryear{{Chelovekov}, {Grebenev} \&
534: {Sunyaev}}{{Chelovekov} et~al.}{2006}]{Chelovekov06}
535: {Chelovekov} I.~V., {Grebenev} S.~A., {Sunyaev} R.~A., 2006, Astronomy
536: Letters, 32, 456
537:
538: \bibitem[\protect\citeauthoryear{Chenevez et~al.,}{Chenevez
539: et~al.}{2007}]{Chenevez07}
540: Chenevez J., et~al., 2007, \aap, 469, L27
541:
542: \bibitem[\protect\citeauthoryear{Cooper \& Narayan}{Cooper \&
543: Narayan}{2007}]{Cooper07}
544: Cooper R.~L., Narayan R., 2007, \apj, 661, 468
545:
546: \bibitem[\protect\citeauthoryear{Cumming}{Cumming}{2004}]{Cumming04b}
547: Cumming A., 2004, Nuc. Phys. B. Proc. Supp., 132, 435
548:
549: \bibitem[\protect\citeauthoryear{Cumming \& Macbeth}{Cumming \&
550: Macbeth}{2004}]{Cumming04}
551: Cumming A., Macbeth J., 2004, \apj, 603, L37
552:
553: \bibitem[\protect\citeauthoryear{Cumming, Macbeth, in't Zand \& Page}{Cumming
554: et~al.}{2006}]{Cumming06}
555: Cumming A., Macbeth J., in't Zand J. J.~M., Page D., 2006, \apj, 646, 429
556:
557: \bibitem[\protect\citeauthoryear{{Degenaar} \& {Wijnands}}{{Degenaar} \&
558: {Wijnands}}{2008}]{Degenaar08}
559: {Degenaar} N., {Wijnands} R., 2008, The Astronomer's Telegram, 1572, 1
560:
561: \bibitem[\protect\citeauthoryear{Falanga, Chenevez, Cumming, Kuulkers, Trap \&
562: Goldwurm}{Falanga et~al.}{2008}]{Falanga08}
563: Falanga M., Chenevez J., Cumming A., Kuulkers E., Trap G., Goldwurm A.,
564: 2008, \aap, 484, 43
565:
566: \bibitem[\protect\citeauthoryear{Fujimoto, Hanawa \& Miyaji}{Fujimoto
567: et~al.}{1981}]{Fujimoto81}
568: Fujimoto M.~Y., Hanawa T., Miyaji S., 1981, \apj, 247, 267
569:
570: \bibitem[\protect\citeauthoryear{{Galloway}, {Muno}, {Hartman}, {Psaltis} \&
571: {Chakrabarty}}{{Galloway} et~al.}{2008}]{Galloway06}
572: {Galloway} D.~K., {Muno} M.~P., {Hartman} J.~M., {Psaltis} D.,
573: {Chakrabarty} D., 2008, accepted by ApJS
574:
575: \bibitem[\protect\citeauthoryear{{Gupta}, {Brown}, {Schatz}, {M{\"o}ller} \&
576: {Kratz}}{{Gupta} et~al.}{2007}]{gupta2007}
577: {Gupta} S., {Brown} E.~F., {Schatz} H., {M{\"o}ller} P., {Kratz} K.-L.,
578: 2007, \apj, 662, 1188
579:
580: \bibitem[\protect\citeauthoryear{{Haensel} \& {Zdunik}}{{Haensel} \&
581: {Zdunik}}{1990}]{haensel1990}
582: {Haensel} P., {Zdunik} J.~L., 1990, \aap, 227, 431
583:
584: \bibitem[\protect\citeauthoryear{{Haensel} \& {Zdunik}}{{Haensel} \&
585: {Zdunik}}{2003}]{haensel2003}
586: {Haensel} P., {Zdunik} J.~L., 2003, \aap, 404, L33
587:
588: \bibitem[\protect\citeauthoryear{{Haensel} \& {Zdunik}}{{Haensel} \&
589: {Zdunik}}{2008}]{haensel2008}
590: {Haensel} P., {Zdunik} J.~L., 2008, \aap, 480, 459
591:
592: \bibitem[\protect\citeauthoryear{{Hasinger} \& {van der Klis}}{{Hasinger} \&
593: {van der Klis}}{1989}]{HK89}
594: {Hasinger} G., {van der Klis} M., 1989, \aap, 225, 79
595:
596: \bibitem[\protect\citeauthoryear{{Horowitz}, {Dussan} \& {Berry}}{{Horowitz}
597: et~al.}{2008}]{horowitz2008}
598: {Horowitz} C.~J., {Dussan} H., {Berry} D.~K., 2008, Physical Review C, 77,
599: 045807
600:
601: \bibitem[\protect\citeauthoryear{in't Zand, Cumming, van~der Sluys, Verbunt \&
602: Pols}{in't Zand et~al.}{2005}]{intZand05}
603: in't Zand J. J.~M., Cumming A., van~der Sluys M.~V., Verbunt F., Pols
604: O.~R., 2005, \aap, 441, 675
605:
606: \bibitem[\protect\citeauthoryear{in't Zand, Jonker \& Markwardt}{in't Zand
607: et~al.}{2007}]{intZand07}
608: in't Zand J. J.~M., Jonker P.~G., Markwardt C.~B., 2007, \aap, 465, 953
609:
610: \bibitem[\protect\citeauthoryear{{Kaplan} \& {Chakrabarty}}{{Kaplan} \&
611: {Chakrabarty}}{2008}]{Kaplan08}
612: {Kaplan} D.~L., {Chakrabarty} D., 2008, The Astronomer's Telegram, 1630, 1
613:
614: \bibitem[\protect\citeauthoryear{{Kuulkers}, {den Hartog}, {in't Zand},
615: {Verbunt}, {Harris} \& {Cocchi}}{{Kuulkers} et~al.}{2003}]{Kuulkers03}
616: {Kuulkers} E., {den Hartog} P.~R., {in't Zand} J.~J.~M., {Verbunt} F.~W.~M.,
617: {Harris} W.~E., {Cocchi} M., 2003, \aap, 399, 663
618:
619: \bibitem[\protect\citeauthoryear{{Linares}, {Soleri}, {Curran}, {Wijnands},
620: {Degenaar}, {van der Klis}, {Starling} \& {Markwardt}}{{Linares}
621: et~al.}{2008}]{Linares08}
622: {Linares} M., {Soleri} P., {Curran} P., {Wijnands} R., {Degenaar} N., {van
623: der Klis} M., {Starling} R., {Markwardt} C., 2008, The Astronomer's
624: Telegram, 1618, 1
625:
626: \bibitem[\protect\citeauthoryear{{Markwardt}, {Pereira} \& {Swank}}{{Markwardt}
627: et~al.}{2008a}]{Markwardt08}
628: {Markwardt} C.~B., {Pereira} D., {Swank} J.~H., 2008a, The Astronomer's
629: Telegram, 1569, 1
630:
631: \bibitem[\protect\citeauthoryear{{Markwardt}, {Cummings} \&
632: {Krimm}}{{Markwardt} et~al.}{2008b}]{Markwardt08b}
633: {Markwardt} C.~B., {Cummings} J., {Krimm} H., 2008b, The Astronomer's
634: Telegram, 1616, 1
635:
636: \bibitem[\protect\citeauthoryear{Molkov, Revnivtsev, Lutovinov \&
637: Sunyaev}{Molkov et~al.}{2005}]{Molkov05}
638: Molkov S., Revnivtsev M., Lutovinov A., Sunyaev R.~A., 2005, \aap, 434,
639: 1069
640:
641: \bibitem[\protect\citeauthoryear{Peng, Brown \& Truran}{Peng
642: et~al.}{2007}]{Peng07}
643: Peng F., Brown E.~F., Truran J.~W., 2007, \apj, 654, 1022
644:
645: \bibitem[\protect\citeauthoryear{{Sakamoto} et~al.,}{{Sakamoto}
646: et~al.}{2008}]{Sakamoto08}
647: {Sakamoto} T., et~al., 2008, GCN, 8034, 1
648:
649: \bibitem[\protect\citeauthoryear{{Starling} \& {Evans}}{{Starling} \&
650: {Evans}}{2008}]{Starling08}
651: {Starling} R., {Evans} P., 2008, The Astronomer's Telegram, 1621, 1
652:
653: \bibitem[\protect\citeauthoryear{{Strohmayer}, {Markwardt} \&
654: {Swank}}{{Strohmayer} et~al.}{2008}]{Strohmayer08}
655: {Strohmayer} T.~E., {Markwardt} C.~B., {Swank} J.~H., 2008, The
656: Astronomer's Telegram, 1635, 1
657:
658: \bibitem[\protect\citeauthoryear{{van der Klis}}{{van der
659: Klis}}{1995}]{vanderKlis95b}
660: {van der Klis} M., 1995, Procs. of the NATO Advanced Study Institute on
661: the Lives of Neutron Stars, Kemer, Turkey, 1993. Eds. Alpar, Kiziloglu, van Paradijs; Publ.
662: Kluwer Academic
663:
664: \bibitem[\protect\citeauthoryear{{van der Klis}}{{van der
665: Klis}}{2006}]{vanderKlis06}
666: {van der Klis} M., 2006, in "Compact Stellar X-ray Sources", ed. W. H. G.
667: Lewin \& M. van der Klis (Cambridge Univ. Press) (astro-ph/0410551), pp
668: 39--112
669:
670: \end{thebibliography}
671:
672:
673: \end{document}
674: