1: \documentclass[apj]{emulateapj}
2: %\usepackage{psfig}
3:
4: \newcommand{\etal}{et al.}
5: \newcommand{\loe}{\stackrel{<}{\sim}}
6: \newcommand{\goe}{\stackrel{>}{\sim}}
7: \newcommand\rcw{\hbox{RCW\hspace{1.5pt}103}}
8: \newcommand\xte{{\it RXTE\/}}
9: \newcommand\XTE{{\it RXTE\/}}
10: \newcommand\einstein{{\it Einstein}}
11: \newcommand\ginga{{\it Ginga}}
12: \newcommand\GINGA{{\it Ginga}}
13: \newcommand\Ginga{{\it Ginga}}
14: \newcommand\sax{{\it Beppo}SAX\/}
15: \newcommand\SAX{{\it Beppo}SAX\/}
16: \newcommand\ASCA{{\it ASCA\/}}
17: \newcommand\EGRET{{\it EGRET\/}}
18: \newcommand\egret{{\it egret\/}}
19: \newcommand\asca{{\it ASCA\/}}
20: \newcommand\ROSAT{{\it ROSAT\/}}
21: \newcommand\rosat{{\it ROSAT\/}}
22: \newcommand\chandra{{\it Chandra}}
23: \newcommand\Chandra{{\it Chandra}}
24: \newcommand\xmm{{\it XMM-Newton}}
25: \newcommand\XMM{{\it XMM-Newton}}
26:
27: \def\s{\phantom{1}}
28: \def\xray{\hbox{X-ray}}
29: \def\nh{\hbox{$N_{\rm H}$}}
30:
31: \def\snr{Kes~79}
32: \def\psr{\rm{PSR J1852$+$0040}}
33: \def\src{\rm{CXOU~J185238.6$+$004020}}
34: \def\one{\rm{1E~1207.4$-$5209}}
35: \newcommand\degree{ .\!^{\circ} }
36: \def\simlt{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$<$}}}}
37: \def\simgt{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$>$}}}}
38:
39: \slugcomment{Received 2007 April 10; accepted 2007 May 7}
40:
41: \shorttitle{X-ray Timing of PSR J1852+0040}
42: \shortauthors{Halpern et al.}
43:
44: \begin{document}
45:
46: \title{X-ray Timing of PSR J1852+0040 in Kesteven~79: \\
47: Evidence of Neutron Stars Weakly Magnetized at Birth}
48:
49: \author{J. P. Halpern\altaffilmark{1}, E. V. Gotthelf\altaffilmark{1},
50: F. Camilo\altaffilmark{1}, and F. D. Seward\altaffilmark{2}}
51:
52: \altaffiltext{1}{Columbia Astrophysics Laboratory, Columbia University,
53: New York, NY 10027}
54: \altaffiltext{2}{Smithsonian Astrophysical Observatory, Cambridge, MA 02138}
55:
56: \begin{abstract}
57: The 105-ms X-ray pulsar J1852+0040 is the central
58: compact object (CCO) in supernova remnant \snr. We report a
59: sensitive upper limit on its radio flux density of $12\,\mu$Jy at
60: 2~GHz using the NRAO Green Bank Telescope.
61: Timing using the {\it Newton X-Ray Multi-Mirror Mission} (\xmm)
62: and the \chandra\ {\it X-ray Observatory} over a 2.4~yr span
63: reveals no significant change in its spin period.
64: The $2\sigma$ upper limit on the period derivative leads,
65: in the dipole spin-down formalism, to an energy loss rate
66: $\dot E < 7 \times 10^{33}$~ergs~s$^{-1}$,
67: surface magnetic field strength $B_p < 1.5 \times 10^{11}$~G,
68: and characteristic age $\tau_c \equiv P/2\dot P> 8$~Myr.
69: This value of $\tau_c$
70: exceeds the age of the SNR by 3 orders of magnitude,
71: implying that the pulsar was born spinning
72: at its current period. However, the X-ray luminosity of \psr,
73: $L_{\rm bol} \approx 3 \times 10^{33}\,(d/7.1\ {\rm kpc})^2$ ergs~s$^{-1}$,
74: is a large fraction of $\dot E$, which challenges the
75: rotation-powered assumption. Instead,
76: its high blackbody temperature $kT_{\rm BB} = 0.46\pm 0.04$ keV,
77: small blackbody radius $R_{\rm BB} \approx 0.8$~km, and large pulsed fraction
78: $f_p \approx 80\%$, may be evidence of accretion onto a polar cap, possibly
79: from a fallback disk made of supernova debris.
80: If $B_p < 10^{10}$~G, an accretion disk can
81: penetrate the light cylinder and interact with the magnetosphere
82: while resulting torques on the neutron star remain within the
83: observed limits. A weak $B$-field is also inferred in another CCO,
84: the 424-ms pulsar \one, from its steady
85: spin and soft X-ray absorption lines.
86: We propose this origin of radio-quiet CCOs:
87: the magnetic field, derived from a turbulent
88: dynamo, is weaker if the NS is formed spinning slowly,
89: which enables it to accrete SN debris. Accretion excludes
90: neutron stars born with both $B_p < 10^{11}$~G
91: and $P > 0.1$~s from radio pulsar surveys,
92: where $B_p < 10^{11}$~G is not encountered except
93: among very old ($\tau_c > 40$~Myr) or recycled pulsars.
94: Finally, such a CCO, if born in SN~1987A,
95: could explain the non-detection of a pulsar there.
96: \end{abstract}
97:
98: \keywords{ISM: individual (Kesteven~79, SN~1987A) --- pulsars: individual
99: (\one, \src, \psr) --- stars: neutron}
100:
101: \section {Introduction}
102:
103: A compact X-ray source, \src,
104: was found in the center of the supernova remnant
105: (SNR) Kes~79 by \cite{sew03}. Using \xmm\ in 2004 October,
106: we discovered 105-ms X-ray pulsations from this CCO \citep[Paper 1]{got05},
107: also named \psr. Its pulsed fraction was
108: as high as 86\% and it had an apparently thermal X-ray spectrum.
109: From the discovery observations, only an upper limit on its period
110: derivative could be determined, $\dot P < 7 \times 10^{-14}$~s~s$^{-1}$,
111: which left the energetics of the NS and the mechanism of its
112: X-ray emission ambiguous. The two most plausible models explored in Paper 1,
113: a rotation-powered pulsar, and accretion from fallback material,
114: each encountered some difficulties.
115:
116: Here, we present three new observations,
117: two from \xmm\ and one from \chandra, that over a 2.4 year span refine
118: the spectral and timing properties of \psr. In addition, we report
119: the negative result of a search for radio pulsations in a deep GBT pointing.
120: From the assembled X-ray timing of \psr, a sensitive new upper limit is
121: obtained on its period derivative, suggesting that the underlying cause
122: of its unusual properties is a weak magnetic field,
123: $B_p < 1.5 \times 10^{11}$~G.
124: Unlike canonical radio pulsars with $B_p \sim 10^{12-13}$~G, a young NS
125: that is weakly magnetized and spinning slowly at birth is able
126: to accrete from material that is thought to remain following a
127: supernova explosion. We discuss this as a model for the properties
128: of radio-quiet CCOs, including the similar pulsar \one.
129:
130: \section{\xmm\ and \chandra\ Observations}
131:
132: The two observations previously reported in Paper~1 were obtained with
133: \xmm\ on 2004 October 18 and 23. Two more observations, reported
134: here, were made on 2006 October 8 and 2007 March 20
135: using the same instrument modes, and exposure times of 30~ks each.
136: The pn CCD of the European Photon Imaging Camera (EPIC;
137: \citealt{tur03}), was operated in ``small window'' mode
138: with a $4\farcm3 \times 4\farcm3$ field-of-view (FOV) and 5.7~ms time
139: resolution to resolve the pulsations of \psr.
140: The EPIC~MOS CCDs were exposed in
141: ``full frame'' mode with a $30^{\prime}$ diameter FOV and time
142: resolution of 2.7~s. For all three cameras the medium
143: density filter was used. We processed all four observations
144: with Science Analysis System
145: version xmmsas\_20060628\_1801-7.0.0, using the same techniques as in Paper~1.
146:
147: A 32~ks \chandra\ observation was performed on 2006 November~23
148: using the Advanced Camera for Imaging and Spectroscopy (ACIS)
149: operated in continuous-clocking (CC) mode to provide a time resolution
150: of 2.85~ms. The back-illuminated ACIS-S3 CCD was used, and
151: the Scientific Instrument Module (SIM) was offset in the $-Z$
152: direction to place the pulsar $0^{\prime}\!.85$ from the edge of the
153: CCD to reduce contamination from the thermal remnant. To achieve
154: the fast timing in CC mode, one spatial dimension of the CCD image is
155: sacrificed by integrating along the column direction during each CCD
156: readout cycle. The geometry of the observation is illustrated in
157: Figure~\ref{image}, overlaid on an image of \snr\ previously
158: obtained with ACIS-I in timed exposure mode on 2001 July 31 \citep{sew03}.
159: We also reanalyzed the 2001 observation for the spectral properties
160: of the pulsar. During both observations the CCD
161: background count rate was stable and no time filtering was necessary.
162: All photon arrival times were adjusted by the standard processing to
163: account for the spacecraft dithering, and SIM offset in the case of
164: the CC-mode observation. Reduction and
165: analysis of the \chandra\ data are all based on the standard software
166: package CIAO (v3.4) and CALDB (v3.3).
167:
168: \begin{figure}[t]
169: \centerline{
170: \hfill
171: \includegraphics[width=0.95\linewidth,angle=270]{f1.eps}
172: \hfill
173: }
174: \caption{Geometry of the 2006 November 23
175: \chandra\ CC-mode observation of \psr,
176: showing the outline of the S3 CCD and the columns used for source and
177: background extraction, overlaid on the 2001 July 31
178: ACIS-I observation of \citet{sew03}.
179: (This broadband image of \snr\ is square-root
180: scaled to emphasize faint emission features and background.)
181: The counts were summed along the length of the
182: columns, oriented at position angle $241^{\circ}\!.5$
183: east of north. Both spectral and timing data for the pulsar were
184: extracted from the six central columns. The
185: background spectrum was obtained from nine neighboring columns on either
186: side ({\it flanking lines}), separated from the source by a
187: two-column gap.
188: }
189: \label{image}
190: \end{figure}
191:
192: \subsection{Timing Analysis}
193:
194: \begin{deluxetable*}{llccccc}[h]
195: \tablecolumns{7}
196: \tablewidth{0pt}
197: \tablecaption{X-ray Timing of \psr}
198: \tablehead{
199: \colhead{Mission} & \colhead{Date} & \colhead{Epoch} & \colhead{Duration} &
200: \colhead{Rate\tablenotemark{a}} & \colhead{Period\tablenotemark{b}} &
201: \colhead{$f_p$\tablenotemark{c}} \\
202: & \colhead{(UT)} & \colhead{(MJD)} & \colhead{(s)} & \colhead{(s$^{-1}$)}
203: & \colhead{(ms)} & \colhead{(\%)}\\
204: }
205: \startdata
206: {\it XMM} & 2004 Oct 18 & 53,296.001 & 30,587 & 0.041(2) & 104.912643(18) & 86(16)\\
207: {\it XMM} & 2004 Oct 23 & 53,301.985 & 30,515 & 0.046(2) & 104.912600(27) & 61(16)\\
208: {\it XMM} & 2006 Oct 08 & 54,016.245 & 30,243 & 0.045(2) & 104.912593(20) & 77(15)\\
209: \chandra\ & 2006 Nov 23 & 54,062.257 & 32,165 & 0.022(2) & 104.912612(19) & 100(16)\\
210: {\it XMM} & 2007 Mar 20 & 54,179.878 & 30,506 & 0.043(2) & 104.912609(19) & 71(15)\\
211: \enddata
212: \tablenotetext{a}{\footnotesize For \xmm, background and dead-time corrected
213: count rate in a $15^{\prime\prime}$ radius aperture.
214: Statistical ($\sqrt{N}$) uncertainty
215: in the last digit is given in parentheses.}
216: \tablenotetext{b}{\footnotesize Period derived from a $Z^2_1$ test. Period
217: uncertainty is $1\sigma$ computed by the Monte Carlo method described in
218: \citet{got99}.}
219: \tablenotetext{c}{\footnotesize Pulsed fraction defined as
220: $f_p \equiv N({\rm pulsed})/N({\rm total})$. Unpulsed level set at lowest
221: point of 20-bin folded lightcurve.}
222: \label{timetable}
223: \end{deluxetable*}
224:
225: \begin{deluxetable*}{lcccccc}
226: %\rotate
227: \tablecolumns{7}
228: \tablewidth{0pt}
229: \tablecaption{X-ray Spectral Fits for \psr \label{spectable}}
230: \tablehead{
231: \colhead{} & \colhead{2001 Jul 31} & \colhead{2004 Oct 18} & \colhead{2004 Oct 23} & \colhead{2006 Oct 08} &
232: \colhead{2006 Nov23} & \colhead{2007 Mar 20}\\
233: \colhead{Parameter} & \colhead{\chandra} & \colhead{{\it XMM}} & \colhead{{\it XMM}} & \colhead{{\it XMM}} & \colhead{\chandra} &
234: \colhead{{\it XMM}}
235: }
236: \startdata
237: \multispan7{\hfill \hbox{Blackbody~Model}\hfill \vspace{5pt}}\\
238: \tableline
239: $N_{\rm H}$ ($10^{22}$ cm$^{-2}$) & $1.1 \pm 0.3$ & $1.4 \pm 0.3$ & $1.4 \pm 0.3$ & $1.4\pm0.3$ & $1.6\pm 0.3$
240: & $1.3\pm 0.3$ \\
241: $kT_{\rm BB}$ (keV) & $0.50 \pm 0.04$ & $0.44\pm 0.04$ & $0.45\pm 0.03$ & $0.46\pm0.04$ & $0.41\pm 0.04$
242: & $0.46\pm 0.04$ \\
243: $R_{\rm BB}$ (km) & $0.62\pm0.05$ & $0.92 \pm 0.09$ & $0.79 \pm 0.09$ & $0.74\pm0.08$ & $0.96\pm0.04$
244: & $0.69\pm0.09 $ \\
245: $A_{\rm BB}$ ($10^{10}$~cm$^2$) & $4.8\pm0.7$ & $10.7 \pm 2.1$ & $7.8 \pm 1.7$ & $6.8 \pm 1.5$ & $11.5\pm4.4$
246: & $6.0\pm1.5$ \\
247: $F_{\rm BB} (10^{-13}$~cgs)\tablenotemark{a} & $2.0\pm0.3$ & $1.9 \pm 0.3$ & $2.0 \pm 0.3$ & $2.0\pm 0.3$ & $1.7\pm0.4$
248: & $1.9\pm0.3$ \\
249: $L_{\rm BB, bol} (10^{33}$~cgs)\tablenotemark{b}& $3.2\pm 0.7$ & $4.3 \pm 1.1$ & $3.3 \pm 1.0$ & $3.2\pm 1.0$ & $3.5\pm1.7$
250: & $2.9\pm1.1$ \\
251: $\chi^2_{\nu}({\nu})$ & 0.71(18) & 0.85(68) & 0.67(70) & 0.86(69) & 0.88(22)
252: & 0.67(69) \\
253: \cutinhead{Power-law~Model}
254: $N_{\rm H}$ ($10^{22}$ cm$^{-2}$) & $2.6^{+0.7}_{-0.4}$ & $3.2^{+0.6}_{-0.6}$ & $3.1^{+0.6}_{-0.5}$ & $3.2^{+0.6}_{-0.5}$ & $3.4^{+0.8}_{-0.6}$ & $3.2^{+0.7}_{-0.6}$ \\
255: $\Gamma$ & $4.1^{+0.7}_{-0.4}$ & $4.9^{+0.7}_{-0.6}$ & $4.9^{+0.6}_{-0.5}$ & $4.8^{+0.6}_{-0.5}$ & $5.3^{+0.8}_{-0.7}$ & $4.9^{+0.6}_{-0.6}$ \\
256: $F_{\rm PL} (10^{-13}$~cgs)\tablenotemark{a} & $2.0\pm 0.6$ & $1.9 \pm 0.43$ & $2.0 \pm 0.4$ & $2.1\pm0.5$ & $1.8\pm0.4$
257: & $1.9\pm0.4$ \\
258: $L_{\rm PL} (10^{33}$~cgs)\tablenotemark{c} & $1.2\pm 0.4$ & $1.1 \pm 0.2$ & $1.2 \pm 0.2$ & $1.2\pm0.3 $ & $1.1\pm0.3$
259: & $1.2\pm0.3$ \\
260: $\chi^2_{\nu}({\nu})$ & 0.94(18) & 0.82(68) & 0.68(70) & 0.78(69) & 0.81(22)
261: & 0.71(69) \\
262: \enddata
263: \tablecomments{\footnotesize Uncertainties are 68\% confidence intervals for two interesting parameters. Luminosities are computed for $d=7.1$~kpc.}
264: \tablenotetext{a}{\footnotesize Absorbed flux in the $1-5$ keV band in units of ergs~cm$^{-2}$~s$^{-1}$.}
265: \tablenotetext{b}{\footnotesize Unabsorbed, bolometric blackbody luminosity in units of ergs~s$^{-1}$.}
266: \tablenotetext{c}{\footnotesize Absorbed luminosity in the $1-5$ keV band in units of ergs~s$^{-1}$. This corrects an error in Table~2 of Paper~1.}
267: \end{deluxetable*}
268:
269: \begin{figure}[t]
270: \centerline{
271: \hfill
272: \includegraphics[width=0.75\textwidth,angle=270]{f2.eps}
273: \hfill
274: }
275: \vspace{0.3cm}
276: \caption{Folded light curves of \psr\ in the $1-5$ keV band using
277: \xmm\ EPIC~pn on 2006 October 8 ({\it top}),
278: \chandra\ ACIS-S3 in CC mode on 2006 November 23 ({\it middle}),
279: and \xmm\ EPIC~pn on 2007 March 20 ({\it bottom}).
280: The estimated background level is indicated by the dashed line.
281: Phase zero is arbitrary.}
282: \label{timeplot}
283: \end{figure}
284:
285: Photon arrival times were converted to the solar system
286: barycenter using the \chandra\ derived source coordinates
287: from Paper~1, R.A. = $18^{\rm h}52^{\rm m}38\fs57$,
288: decl. = $+00\arcdeg40\arcmin19\farcs8$ (J2000.0),
289: and signal strength was maximized by selecting the $1-5$~keV energy
290: band. From the \xmm\ observation of 2006 October 8,
291: a $Z^2_1$ test \citep{buc83} localizes the period to
292: $P = 104.912593(20)$~ms with a peak statistic of $Z^2_1 = 88$.
293: The resulting light curve, which has pulsed fraction
294: $f_p = 77\% \pm 15\%$ after accounting for SNR background,
295: is shown in Figure~\ref{timeplot}. Here, we define the pulsed fraction as
296: $f_p \equiv N({\rm pulsed})/N({\rm total})$,
297: where we choose the minimum of the folded light curve as
298: the unpulsed level. The observation on 2007 March 20 yielded
299: consistent parameters, $P = 104.912609(19)$~ms with $Z^2_1 = 83$
300: and $f_p = 71\% \pm 15\%$. These new values of
301: $f_p$ are intermediate between
302: the 2004 \xmm\ observations, which yielded
303: $f_p = 86\% \pm 16\%$ and $61\% \pm 16\%$, respectively,
304: weakening the marginal evidence of variability of
305: $f_p$ that was noted in Paper~1.
306: Similar analysis of the \chandra\ observation
307: of 2006 Novemeber 23 yields a formally identical period,
308: $P = 104.912612(19)$~ms with $Z^2_1 = 94$, and $f_p$ consistent
309: with 100\%. The pulsed fraction is less reliable in CC-mode data
310: due to systematics of estimating the contamination from SNR background.
311: Table~\ref{timetable} is a summary of all of the available
312: timing data on \psr.
313:
314: While the new observations extend the sampled
315: time span by a factor of 150 since Paper~1,
316: there is still no significant detection of a period derivative.
317: A $\chi^2$ fit to the five measurements of $P$ yields the
318: formally negative $<\dot P> = (-3.4 \pm 2.7) \times 10^{-16}$~s~s$^{-1}$
319: that is, however, consistent with zero at the $\approx 1\sigma$ level.
320: We are unable to link any of the observations with a
321: definite cycle count in order to improve on this measurement.
322: If we adopt the $2\sigma$ upper limit $\dot P < 2.0 \times 10^{-16}$,
323: the dipole spin-down formalism for an isolated pulsar implies
324: an energy loss rate $\dot E = -I\Omega\dot\Omega =
325: 4\pi^2I\dot P/P^3 < 7 \times 10^{33}$~ergs~s$^{-1}$,
326: surface magnetic field strength
327: $B_p = 3.2\times10^{19}\sqrt{P\dot P} < 1.5 \times 10^{11}$~G,
328: and characteristic age $\tau_c \equiv P/2\dot P > 8$~Myr.
329: %At the $3\sigma$ level, the corresponding limits are
330: %$\dot P < 3.9 \times 10^{-16}$~s~s$^{-1}$,
331: %$\dot E < 1.3 \times 10^{34}$~ergs~s$^{-1}$,
332: %$B_p < 2 \times 10^{11}$~G, and
333: %$\tau_c = P/2\dot P > 4$~Myr. At either level of significance,
334: These derived quantities pose problems for a rotation-powered
335: X-ray source, as discussed in \S 4.1.
336:
337: \subsection{Spectral Analysis}
338:
339: \begin{figure}[t]
340: \centerline{
341: \hfill
342: \includegraphics[width=0.34\textwidth,angle=270]{f3a.eps}
343: \hfill
344: }
345: \vspace{0.2cm}
346: \centerline{
347: \hfill
348: \includegraphics[width=0.34\textwidth,angle=270]{f3b.eps}
349: \hfill
350: }
351: \vspace{0.2cm}
352: \centerline{
353: \hfill
354: \includegraphics[width=0.34\textwidth,angle=270]{f3c.eps}
355: \hfill
356: }
357: \caption{X-ray spectra of \psr\ in Kes~79.
358: {\it Top}: The \xmm\ spectrum on 2006 October 8. Data from the EPIC~pn
359: ({\it black, upper}) and EPIC MOS ({\it red, lower}) with
360: the best-fit blackbody model using parameters in Table~\ref{spectable}.
361: The residuals from the fits are in units of $\sigma$.
362: {\it Middle}: Spectrum on 2006 November 23 from the \chandra\ ACIS-S3
363: in CC mode.
364: {\it Bottom}: The \xmm\ spectrum on 2007 March 20.
365: [{\it See the electronic version of the Journal for a color
366: version of this figure.}]}
367: \label{specfig}
368: \end{figure}
369:
370: Following the analysis method in Paper~1 for each \xmm\ observation,
371: spectra from the two EPIC MOS cameras (MOS1, MOS2)
372: were co-added. The EPIC~MOS and EPIC~pn spectra
373: were then fitted simultaneously using the {\tt XSPEC} (v12.21)
374: package. The results of spectal
375: fits using either a blackbody or an absorbed power-law model
376: are presented in Table~\ref{spectable}. An acceptable
377: $\chi^2$ statistic is obtained using either model.
378: However, the blackbody is preferred over
379: the power law based on its derived column density of $\nh =
380: (1.4\pm0.3) \times 10^{22}$~cm$^{-2}$, which is consistent with that
381: found for the SNR \citep[$\nh \approx 1.6 \times
382: 10^{22}$~cm$^{-2}$,][]{sun04}. The fitted column density for the
383: power-law model, $\nh = 3.2^{+0.6}_{-0.5} \times 10^{22}$~cm$^{-2}$,
384: is significantly larger than the integrated 21 cm Galactic value of
385: $2\times 10^{22}$~cm$^{-2}$ in this direction \citep{dic90}.
386: The best-fit blackbody
387: model yields temperature $kT_{\rm BB} = 0.46\pm0.04$~keV (see
388: Fig.~\ref{specfig}). Adopting a distance $d=7.1$~kpc from \cite{cas98},
389: the bolometric blackbody luminosity
390: $L_{\rm BB, bol} \approx 3.0\times10^{33}$~ergs~s$^{-1}$,
391: corresponding to blackbody area
392: $A_{\rm BB} \approx 7 \times 10^{10}\,(d/7.1\ {\rm kpc})^2$~cm$^2$
393: or $\approx 0.4\%$ of the NS surface.
394: These results are consistent with the spectral parameters from the previous
395: \xmm\ observations in 2004 October, described in Paper~1 and summarized
396: here in Table~\ref{spectable}, indicating that
397: the flux and spectrum have remained steady.
398:
399: The \chandra\ spectra from the 2001 and 2006 observations
400: were each prepared by grouping a minimum of 40 counts
401: per channel, and fitted using the {\tt XSPEC} software. The
402: standard spectral response matrices were generated for the target
403: location on the CCD and all spectra were corrected for the effects of
404: charge-transfer inefficiency; however, the spectral gain is not
405: calibrated well in CC-mode. The resulting spectral parameters from
406: the \chandra\ observations, listed in Table~\ref{spectable},
407: are similar to the ones from \xmm.
408: The blackbody fits to the new \xmm\ and \chandra\ spectra
409: are shown in Figure~\ref{specfig}.
410:
411: \section{Search for a Radio Pulsar}
412:
413: On 2002 October 29 we observed the location of \src\ with the ATNF Parkes
414: telescope in NSW, Australia. At that time the period of the pulsar was not
415: yet known, so we employed standard pulsar searching techniques.
416: We used the central beam of the multibeam
417: receiver at a frequency of 1374\,MHz, with 96 channels across a bandwidth
418: of 288\,MHz in each of two polarizations. The integration time was
419: 6.8\,hr, during which total-power samples were obtained every 0.73\,ms
420: and recorded for off-line analysis. We searched the
421: dispersion measure range 0--2500\,cm$^{-3}$\,pc (twice the maximum
422: Galactic DM predicted for this line of sight by the Cordes \& Lazio
423: 2000 electron density model; for the distance of 7.1\,kpc,
424: the predicted DM is 440\,cm$^{-3}$\,pc). The search
425: followed closely that described in more detail in \citet{cam06}.
426: No promising candidate periods were identified.
427:
428: Following the X-ray discovery of the pulsar period, we made a deeper search
429: using the GBT. We did this on 2005 June 10,
430: observing the pulsar for 11.1\,hr with the Spigot spectrometer \citep{kap05},
431: sampling 768 frequency channels across a bandwidth of 600\,MHz
432: centered on 1950\,MHz every 81.92\,$\mu$s. In addition to performing
433: a standard blind search, we also folded the data at a number of DMs
434: for periods close to the X-ray value
435: to increase the sensitivity, but again found no candidates.
436: We used the PRESTO software package to analyze these data \citep{ran01,ran02}.
437:
438: Applying the standard modification to the radiometer equation, assuming a
439: quasi-sinusoidal pulse shape
440: (as in X-rays; see Fig.~\ref{timeplot}), and accounting
441: for a sky temperature at this location of 5\,K, we were sensitive to a
442: pulsar with period 105\,ms having a period-averaged flux density at 2\,GHz
443: $\ga 12\,\mu$Jy. Converting this to the more common 1.4\,GHz pulsar search
444: frequency, using a typical spectral index of --1.6 \citep{lor95},
445: results in $S_{1.4} \la 0.02$\,mJy. For a distance of $\approx 7$\,kpc,
446: this corresponds to a pseudo-luminosity limit of $L_{1.4} \equiv S_{1.4}
447: d^2 \la 1$\,mJy\,kpc$^2$. Only one young radio pulsar has a smaller
448: radio luminosity than this \citep[the pulsar in 3C58;][]{cam02}.
449:
450: Based on these results, it is therefore somewhat unlikely that
451: PSR~J1852+0040 is presently a radio pulsar beaming toward the Earth.
452: More likely, it is either beaming away \citep[observed radio pulsars with
453: similar periods, or ages $\la 10$\,kyr, have beaming fractions of $\sim
454: 0.3$; see][]{tau98}, or not an active radio pulsar at all.
455:
456: \clearpage
457: \section{Discussion}
458:
459: \subsection{Rotation-Powered Pulsar?}
460:
461: Potential obstacles to the interpretation
462: of \psr\ as a rotation-powered pulsar were identified
463: in Paper~1. The newly reduced upper limit on
464: its $\dot P$, by more than 2 orders of magnitude,
465: turns those difficulties into major objections.
466: First, the X-ray luminosity of \psr,
467: $L_{\rm bol} \approx 3 \times 10^{33}\,(d/7.1\ {\rm kpc})^2$ ergs~s$^{-1}$,
468: is comparable to the $2\sigma$ upper limit on its spin-down power,
469: $\dot E < 7 \times 10^{33}$~ergs~s$^{-1}$.
470: %and it exceeds the $2\sigma$ limit
471: %$\dot E < 2.6 \times 10^{33}$~ergs~s$^{-1}$.
472: Thus, it may be difficult to power the X-ray source with
473: spin-down energy.
474: Since the X-ray spectrum is consistent with thermal emission,
475: what about residual cooling of the neutron star?
476: While the X-ray luminosity of \psr\ is consistent with minimal
477: NS cooling curves for an age of $10^{3-4}$~yr \citep{pag04}, its
478: blackbody temperature implies an emitting area that is just $\approx
479: 0.4\%$ of the NS surface. Even taking into account that, in a real
480: NS atmosphere, a blackbody fit overestimates the effective
481: temperature and underestimates the area, the highly
482: modulated X-ray pulse must come from a small hot spot.
483: Although anisotropic conduction in a magnetized NS can enforce
484: temperature gradients on the surface, such models \citep{gep04,per06}
485: do not show spots as small and as hot as observed here. Instead,
486: an external source of localized heating may be needed.
487:
488: The open-field-line polar cap is the
489: most likely target area for external heating of the NS surface,
490: with the energy ultimately drawn
491: from rotation. The canonical area of the polar cap is $A_{\rm pc} =
492: {2\pi^2 R^3 / {P c}} \approx 1 \times 10^{10}$~cm$^2$; this is
493: $\sim 14\%$ of the area implied by the blackbody fit to the X-ray spectrum.
494: Among theories of polar-cap heating, a maximum thermal luminosity
495: is predicted by \cite{wan98} from the outer-gap model for
496: $\gamma$-ray pulsars. In this analysis, the X-ray luminosity of the
497: hot polar cap is limited
498: by the Goldreich-Julian $e^{\pm}$ current flow of $\dot N_0 \approx 2
499: \times 10^{31}\,(P/0.105\,{\rm s})^{-2}\,(B_p/10^{11}\,{\rm
500: G})$~s$^{-1}$ depositing an average energy per particle $E_f
501: \approx 4.3$~ergs. The maximum luminosity is $L_{\rm bol} \approx f
502: E_f \dot N_0 < 4 \times 10^{31}$~ergs~s$^{-1}$. Here, the fraction $f$
503: of the current reaching the surface is set to its estimated maximum
504: value of ${1\over 2}$.
505: Polar-cap heating models of \citet{har01,har02} predict even less X-ray
506: luminosity than \citet{wan98}. Therefore, current theories fall
507: short of predicting the temperature and
508: luminosity of the X-ray emission from \psr\ in the context of
509: a spinning-down neutron star.
510:
511: After the unexplained X-ray properties,
512: a lesser problem for \psr\ as an isolated pulsar is its
513: characteristic age $\tau_c \equiv P/2\dot P > 8$~Myr, which
514: exceeds by 3 orders of magnitude the SNR age, estimated dynamically
515: as $5.4-7.5$~kyr \citep{sun04}. This would require the pulsar
516: to have been born at its current period to be associated with the SNR.
517: While not fundamentally contradicting the rotation-powered hypothesis,
518: this result is a definite example in support of a recent
519: population analysis that favors a wide distribution of radio
520: pulsar birth periods \citep{fau06}, even though \psr\ is not itself a
521: radio pulsar. Furthermore, as magnetic field is generated by a
522: turbulent dynamo whose strength depends on the rotation rate
523: of the proto-neutron star \citep{tho93}, it is natural that
524: long-period pulsars would have the weaker $B$-fields at birth,
525: and the model of \citet{bon06} shows this. \citet{bon06} also
526: find that a slowly rotating NS should have its $B$-field confined
527: to small-scale regions, with the global dipole that is responsible for
528: spin-down absent.
529:
530: Finally, we note that there are no young radio pulsars known with
531: $B_p < 10^{11}$~G. All such weakly magnetized radio pulsars are either
532: very old, with $\tau_c > 40$~Myr \citep{man05}\footnote{ATNF Pulsar Catalogue,
533: version 1.29, http://www.atnf.csiro.au/research/pulsar/psrcat/\hfil},
534: or are recycled, the result of accretion
535: from a binary companion. \cite{fau06} conclude that there is no
536: evidence for magnetic field decay during the radio emitting life of
537: a pulsar, so the observered distribution of $B_p$ should resemble its
538: birth distribution. Among the $B_p$ distribution of ordinary
539: (not recycled) radio pulsars,
540: \psr\ is certainly in the bottom 10\%, and maybe lower. \psr\
541: may become a radio pulsar in the far future, but its weak $B$-field may
542: be preventing radio emission now by permitting accretion. While
543: there is no evidence of a close binary companion in the timing of \psr\
544: (see Paper~1), it is possible that low-level accretion of SN debris
545: prevents it and other radio-quiet CCOs
546: from becoming radio pulsars for thousands or even millions of years.
547: We next consider whether the hypothesis of accretion from such a
548: fallback disk \citep[e.g.,][]{alp01,shi03,eks05}
549: better explains the X-ray timing and spectral
550: properties of \psr.
551:
552: \subsection{Fallback Accretion?}
553:
554: The fact that $\dot P$ is so small leads
555: us to review the viability of accretion of supernova debris
556: through a disk. The X-ray luminosity of \psr\ can be powered
557: by accretion of $\dot m \approx 3 \times 10^{13}$ g~s$^{-1}$,
558: or only $\approx 0.1$ lunar masses of supernova debris over
559: the past 7.5~kyr.
560: The main barrier to disk accretion is the speed-of-light cylinder,
561: of radius $r_{\ell} = cP/2\pi = 5 \times 10^8$~cm.
562: If an accretion disk cannot penetrate the light cylinder, the NS
563: cannot interact with the disk, and it behaves as a isolated pulsar.
564: But if $B_p$ is as small as $10^{10}$~G, accretion at a rate
565: $\dot M \geq 10^{13}$~g~s$^{-1}$ can penetrate the light cylinder,
566: since the magnetospheric radius is then
567: \begin{equation}
568: r_m = 3.2 \times 10^8\,\mu_{28}^{4/7}\,\dot M^{-2/7}_{13}
569: \,\left({M \over M_{\odot}}\right)^{-1/7}\ {\rm cm},
570: \end{equation}
571: where the magnetic moment $\mu = B_p\,R^3/2 \approx 10^{28}B_{p,10}$ G~cm$^3$.
572: If so, the system is in the propeller regime,
573: in which matter flung out from
574: $r_m$ at a rate $\dot M$ takes angular momentum from the NS,
575: causing it to spin down. The propeller spin-down rate is estimated as
576: \begin{equation}
577: \dot P \approx 2\,\dot M\,r_m^2\,I^{-1}\,P\,
578: \left(1- {P \over P_{\rm eq}}\right)
579: \end{equation}
580: \citep[e.g.,][]{men99,eks05}. Here $I \approx 10^{45}$ g~cm$^2$
581: is the NS moment of inertia, and
582: $P_{\rm eq} = 3.2\,\mu_{28}^{6/7}\,\dot M_{13}^{-3/7}\,
583: (M/M_{\odot})^{-5/7}$~s is the equilibrium, or minimum
584: period for disk accretion that is reached when $r_m \leq r_{\rm co}$.
585: The corotation radius is
586: $r_{\rm co} = [GM(P/2\pi)^2]^{1/3} = 3.7 \times 10^7$~cm.
587:
588: While we have not measured a significant
589: $\dot P$ in order to determine the needed $\dot M$, we can adopt
590: the $2\sigma$ upper limit, $\dot P < 2.0 \times 10^{-16}$,
591: as an estimate of $\dot P$.
592: Combining equations (1) and (2), we have in the propeller scenario
593: \begin{displaymath}
594: \dot P \approx 2.2 \times 10^{-16}\,\mu_{28}^{8/7}\,\dot M_{13}^{3/7}\,
595: \left({M \over M_{\odot}}\right)^{-2/7}\,I_{45}^{-1}\,
596: \left({P \over 0.105\ {\rm s}}\right)\break
597: \end{displaymath}
598: \begin{equation}
599: \left(1- {P \over P_{\rm eq}}\right).
600: \end{equation}
601: We must distinguish here between $\dot M$, the matter expelled that is
602: responsible for the torque on the NS,
603: and $\dot m \approx 3 \times 10^{13}$~g~s$^{-1}$,
604: the matter accreted, which is responsible for the X-ray emission
605: from the surface, presumably at a magnetic pole of the neutron star.
606: For the propeller model to be self-consistent,
607: $\dot M$ must be $> \dot m$, which is possible according to
608: equation (3) as long as $B_p < 10^{10}$~G. Even so,
609: $\dot M$ does not contribute significantly to the X-ray luminosity because
610: of the much weaker gravitational potential at the
611: magnetospheric radius $r_m$ from which it is expelled.
612: If $B_p$ is as small as $7 \times 10^8$~G, then \psr\ is
613: not a propeller, but a ``slow rotator,''
614: with $r_m \leq r_{\rm co}$ and $P_{\rm eq} \leq P$.
615: In this spin-up regime, it would be even more difficult to
616: measure $\dot P$, since
617: \begin{equation}
618: \dot P \approx -1.3 \times 10^{-18}\,\mu_{27}^{2/7}\,\dot m_{13}^{6/7}\,
619: \left({M \over M_{\odot}}\right)^{3/7}\,
620: \left(P \over 0.105\ {\rm s}\right)^2
621: \end{equation}
622: \citep{gos79}.
623:
624: These equations also apply during the prior evolution of the pulsar
625: and show that, even if the accretion rate was orders of
626: magnitude higher in the past,
627: the spin-up and spin-down times $P/\dot P$ are much longer than the age
628: of the SNR. There has not been enough time for propeller accretion
629: to have spun the pulsar down from a much smaller $P$, so even in
630: this model it was born at essentially its present $P$.
631:
632: Thus, we regard either propeller spindown or accretion in a
633: weak magnetic field as consistent with
634: the X-ray spectral and timing properties of \psr.
635: If accreting, its emitting region could be larger than
636: the open-field-line polar cap because it would be connected
637: to magnetic field lines extending to $r_m$, which is $<r_{\ell}$.
638: Neither equation (3) nor equation (4) requires there to be much
639: torque noise at the level that present data can measure.
640: While detected flickering is
641: an indicator of accretion in X-ray binaries even in quiescence
642: \citep{rut07}, we do not have strong evidence of variability
643: of \psr. But it is not clear that the processes in binaries can
644: be extrapolated to accretion at such low rates from a fossil disk.
645: Finally, because of the large distance and interstellar extinction to
646: \psr, as well as its crowded optical field, existing optical data do
647: not rule out the presence of a fallback accretion disk around \psr\
648: (see Paper~1 for details).
649:
650: \subsection{Comparison with 1E 1207.4$-$5209 and Other CCOs}
651:
652: Difficulties in understanding the timing behavior of
653: another CCO, the 424-ms pulsar \one\
654: \citep{zav04,woo06}, were resolved recently
655: by \citet{got07}, who showed that it is a steady
656: rotator with $\dot E < 1.5 \times 10^{32}$ ergs~s$^{-1}$
657: and $B_p < 3.5 \times 10^{11}$~G. This makes \one\
658: very similar to \psr\ in its physical parameters.
659: Accretion from a fossil disk of supernova debris
660: \citep{alp01,shi03,eks05,liu06} was one of the
661: theories considered by \citet{zav04} to explain
662: the now defunct timing irregularities of \one.
663: But accretion may still be needed to explain the
664: radio-quiet nature of \one,
665: and the details of its X-ray spectrum. However,
666: unlike \psr, upper limits on optical/IR emission from \one\
667: are comparable to what is expected from a geometrically
668: thin, optically thick disk accreting at the rate required to
669: account for its X-ray luminosity \citep{zav04,wan07}.
670: Therefore, it may prove necessary to invoke a
671: radiatively inefficient flow, or perhaps even accretion of
672: solid particles \citep{cor06}, in order to proceed
673: with this theory for \one.
674: We propose that this may be
675: the first phase in the life of those neutron stars born
676: rotating slowly, and with weak magnetic fields.
677:
678: A unique phenomenon displayed by \one\ is the set of
679: broad absorption lines in its soft X-ray spectrum.
680: They are centered at $0.7$~keV and $1.4$~keV \citep{san02,mer02},
681: and possibly at 2.1~keV and 2.8~keV \citep{big03,del04}, although the
682: reality of the two higher-energy features has been
683: disputed \citep{mor05}.
684: Proposed absorption mechanisms include electron cyclotron
685: in a weak ($8 \times 10^{10}$~G) magnetic field \citep{big03,del04},
686: atomic features from singly ionized helium in a strong
687: ($2 \times 10^{14}$~G) field \citep{san02,pav05},
688: and oxygen or neon in a normal ($10^{12}$~G)
689: field \citep{hai02,mor06}. A detailed critique of these
690: models is beyond the scope of this paper. We only remark that:
691: (1) local values of $B$ measured from X-ray spectral features may
692: differ from the global dipole $B$ that is measured by timing,
693: especially if the dipole component is weak \citep{bon06}, and (2)
694: the attractive feature of the electron cyclotron model for \one,
695: the equal energy spacing of its lines, is in harmony with
696: the present work on \psr\ which, by different arguments,
697: requires a weak surface magnetic field.
698:
699: \citet{xu03} argued that electron cyclotron harmonics
700: of roughly equal depth could be produced because,
701: even though the oscillator strength of the
702: harmonic is much less than that of the fundamental, resonant
703: cyclotron radiative transfer could equalize the line strengths.
704: \citet{liu06} also explained how objections to the cyclotron model
705: involving the relative oscillator strengths
706: can be overcome by assuming that the
707: fundamental is optically thick while the harmonic is optically
708: thin, resulting in comparable equivalent widths.
709: They also proposed a specific geometry that accounts
710: for the strength of the lines, which also accommodates the
711: small pulsed fraction observed from \one.
712: This involves a column rising along the magnetic axis and
713: viewed from the side. Finally, \citet{liu06} remarked that such a
714: column may be more easily maintained in an accretion scenario.
715:
716: While the lower-quality spectra of \psr\ do not
717: yet show evidence of absorption features, cyclotron lines could
718: be weak if either the column density in which they are formed
719: or the viewing geometry is not favorable, or especially if the magnetic
720: field is so small as to shift them below the observable X-ray band.
721: The fitted interstellar column density to \psr,
722: $\nh = 1.4 \times 10^{22}$~cm$^{-2}$, is much larger than that
723: to \one, $\nh = (3-10) \times 10^{20}$~cm$^{-2}$ \citep{mer02}.
724: Therefore, the soft X-ray region in which the strongest lines
725: in \one\ are seen is partly suppressed in the case of \psr.
726: Finally, it is possible that the X-ray spectrum of \psr\ is not a pure
727: blackbody as fitted, but is affected by unmodelled,
728: broad cyclotron absorption at the low-energy end.
729:
730: The half dozen CCOs \citep[for a review, see][]{pav04} are similar
731: in their X-ray luminosities and temperatures. Therefore,
732: they may be compatible with the model that
733: we propose for \psr\ and \one. The only piece of
734: evidence that sharply contradicts such unification
735: is the apparent historical outburst of the CCO in the
736: Cas~A SNR that was inferred from infrared light echos
737: detected with the {\it Spitzer Space Telescope} \citep{kra05}.
738: The geometry and sharpness of the echos implies that beamed emission
739: from the central source flared for less than a few weeks about
740: 50 years ago, and the radiation reprocessed into thermal emission
741: from dust require an equivalent isotropic flare of
742: $\sim 2 \times 10^{46}$~ergs. Because such energetic
743: outbursts are seen on similar timescales from soft gamma-ray
744: repeaters, \citet{kra05} proposed that the Cas~A CCO is a
745: quiescent magnetar, at the other extreme of
746: magnetic field strength from \psr. A convincing determination
747: of the magnetic field strengths of the Cas~A CCO and others
748: will have to await discovery of their pulsations.
749:
750: \subsection{Application to SN 1987A}
751:
752: It has long been recognized that the non-detection of the expected
753: pulsar in the remnant of SN~1987A can be explained if the
754: NS was born with a weak magnetic field or a long rotation
755: period \citep{oge04}, and that several CCOs,
756: including the one in Cas~A, would be undetected if placed in
757: SN~1987A \citep{gra05}. However, it was not established
758: until now that a CCO can have essentially the same spin-down
759: luminosity at birth that it has at an age of 400 or $10^4$~yr,
760: i.e., that it is born with both a weak $B$-field and a long $P$.
761: This new result means that a common type of neutron
762: star can emit less than the observed limits from SN~1987A
763: even if 100\% of its spin-down power is reprocessed into
764: IR emission by dust in the surrounding
765: SN ejecta. The relevant luminosity limits to be satisfied
766: for a point source inside the ring of SN~1987A
767: are $L_x(2-10\,{\rm keV}) < 1.5 \times 10^{34}$ ergs~s$^{-1}$
768: corrected for extinction
769: \citep{par04}, $L(2900-9650$~\AA) $< 8 \times 10^{33}$ ergs~s$^{-1}$
770: corrected for dust absorption \citep{gra05},
771: and $L(10\,\mu{\rm m}) < 1.5 \times 10^{36}$ ergs~s$^{-1}$ for
772: dust emitting at $T \approx 90-100$~K \citep{bou04}. This mid-IR
773: luminosity can be accounted for by radioactive decay of $^{44}$Ti,
774: and therefore represents a very conservative upper limit on the
775: spin-down power of an embedded pulsar.
776: At an age of $10-20$~yr, a cooling NS need
777: emit only $\approx 3 \times 10^{34}$ ergs~s$^{-1}$ of
778: soft X-rays at a temperature of $2.5 \times 10^6$~K \citep{yak02},
779: some of which is absorbed by SN ejecta or ISM in the LMC.
780: The upper limit on spin-down power from \psr,
781: $\dot E < 7 \times 10^{33}$ ergs~s$^{-1}$,
782: is a challenge to detect in SN~1987A,
783: and $\dot E < 1.5 \times 10^{32}$ ergs~s$^{-1}$
784: from \one\ would be impossible. So we conclude that
785: a CCO is a promising model for the unseen NS in SN~1987A.
786:
787: \section{Conclusions and Future Work}
788:
789: We obtained three new X-ray timing points on \psr\ that, in combination
790: with two previous observations, reveal no significant change in spin
791: period on time scales ranging from 1 week to 2.4 yr. In the dipole
792: spin-down formalism, this implies $\dot E < 7 \times 10^{33}$ ergs~s$^{-1}$
793: and $B_p < 1.5 \times 10^{11}$~G.
794: Such a low $B$-field is not seen in young
795: radio pulsars, and while it does not demand a special explanation,
796: it does favor accretion as a source of the small,
797: hot thermal region that is responsible for the highly modulated
798: X-ray pulsations,
799: which may exceed the spin-down power of the isolated NS,
800: and strains any plausible interior cooling or external heating model.
801:
802: We showed that the observered spin period, 105 ms, in combination with
803: a weak field, $B_p \sim 10^{10}$~G, would allow disk
804: accretion in the propeller regime, which may be a more satisfactory
805: model overall. If $B_p$ is even weaker, $< 10^9$~G, \psr\ could
806: be a ``slow'' accretor.
807: %Accretion from a SN debris disk is one of the explanations
808: %suggested previously for the continued erratic timing behavior
809: %of the other CCO pulsar, \one.
810: Considering also the properties of the CCO \one,
811: we speculated that this
812: class may have weak magnetic fields as a result of slow
813: natal rotation, which enables them to accrete. While we do
814: not decisively favor accretion over a rotation-powered pulsar,
815: we emphasize that for \psr\ a weak magnetic field is an unavoidable
816: feature of either interpretation. Such low-$B$ births may
817: be very common if they also occurred in Cas~A and/or SN~1987A.
818:
819: A radio pulsar detection of \psr\ would demonstrate cleanly
820: that it is rotation powered, but our negative results
821: so far are inconclusive because there are several reasons
822: why a rotation-powered pulsar may not be detected in radio.
823: Continued X-ray monitoring will provide a more sensitive
824: test for spin-down or accretion torques, perhaps eventually
825: in a definitive manner, and more evidence of
826: whether or not the X-ray flux and pulse profile are variable.
827: It is now important to obtain a phase-connected
828: timing solution, the most efficient and sensitive method of measuring
829: a small $\dot P$. If obtained soon, a well-timed series of
830: observations should allow a coherent solution to be fitted
831: retroactively back to 2004. Also, considering the possibility
832: that a weak magnetic field and accretion may be responsible for the
833: absorption lines in the soft X-ray spectrum of \one, a deeper X-ray
834: spectral study of \psr\ may be revealing. Finally, in order to
835: test the general applicability of these ideas to the class of CCOs,
836: more sensitives searches for their pulsations are required.
837:
838: \acknowledgments
839:
840: This investigation is based on observations obtained with \xmm, an ESA
841: science mission with instruments and contributions directly funded by
842: ESA Member States and NASA. Support for this work was provided by
843: NASA through {\it XMM\/} grant NNX06AH95G and {\it Chandra} Award
844: SAO GO6-7048X issued by the {\it Chandra} X-ray Observatory Center,
845: which is operated by the Smithsonian Astrophysical Observatory
846: for and on behalf of NASA under contract NAS8-03060.
847: The GBT is
848: operated by the National Radio Astronomy Observatory, a facility of the
849: National Science Foundation operated under cooperative agreement by
850: Associated Universities, Inc.
851:
852: \begin{thebibliography}{}
853:
854: \bibitem[Alpar(2001)]{alp01}
855: Alpar, M. A. 2001, \apj, 554, 1245
856:
857: \bibitem[Bignami \etal(2003)]{big03}
858: Bignami, G. F., Caraveo, P. A., De Luca, A., \& Mereghetti, S. 2003,
859: \nat, 423, 725
860:
861: \bibitem[Bonanno \etal(2006)]{bon06}
862: Bonanno, A., Urpin, V., \& Belvedere, G. 2006, \aap, 451, 1049
863:
864: \bibitem[Bouchet \etal(2004)]{bou04}
865: Bouchet, P., \etal\ 2004, \apj, 611, 394
866:
867: \bibitem[Buccheri \etal(1983)]{buc83}
868: Buccheri, R., et al.\ 1983, \aap, 128, 245
869:
870: \bibitem[Camilo \etal(2002)]{cam02}
871: Camilo, F., \etal\ 2002, \apj, 571, L41
872:
873: \bibitem [Camilo \etal(2006)]{cam06}
874: {Camilo}, F., {Ransom}, S.~M., {Gaensler}, B.~M., {Slane}, P.~O.,
875: {Lorimer}, D.~R., {Reynolds}, J., {Manchester}, R.~N., \& {Murray},
876: S.~S. 2006, \apj, 637, 456
877:
878: \bibitem[Case \& Bhattacharya(1998)]{cas98}
879: Case, G. L., \& Bhattacharya, D. 1998, \apj, 504, 761
880:
881: \bibitem[Cordes \& Lazio(2002)]{cor02}
882: Cordes, J.~M., \& Lazio, T.~J.~W. 2002, preprint (astro-ph/0207156)
883:
884: \bibitem[Cordes \& Shannon(2006)]{cor06}
885: Cordes, J. M., \& Shannon, R. M. 2006, ApJ, submitted (astro-ph/0605145)
886:
887: \bibitem[De Luca \etal(2004)]{del04}
888: De Luca, A., Mereghetti, S., Caraveo, P. A., Moroni, M., Mignani, R. P.,
889: \& Bignami, G. F. 2004, \aap, 418, 625
890:
891: \bibitem[Dickey \& Lockman(1990)]{dic90}
892: Dickey, J.~M., \& Lockman, F.~J. 1990, \araa, 28, 215
893:
894: \bibitem[Ek\c si \etal(2005)]{eks05}
895: Ek\c si, K. Y., Hernquist, L., \& Narayan, R. 2005, \apj, 623, L41
896:
897: \bibitem[Faucher-Gigu\`ere \& Kaspi(2006)]{fau06}
898: Faucher-Gigu\`ere, C.-A., \& Kaspi, V. M. 2006, \apj, 643, 332
899:
900: \bibitem[Geppert \etal(2004)]{gep04}
901: Geppert, U., K\"uker, M., \& Page, D. 2004, \aap, 426, 267
902:
903: \bibitem[Ghosh \& Lamb(1979)]{gos79}
904: Ghosh, P., \& Lamb, F. K. 1979, \apj, 234, 296
905:
906: \bibitem[Gotthelf \& Halpern(2007)]{got07}
907: Gotthelf, E. V., \& Halpern, J. P. 2007, \apj, submitted (arXiv:0704.2255)
908:
909: \bibitem[Gotthelf \etal(2005)]{got05}
910: Gotthelf, E. V., Halpern, J. P., \& Seward, F. D. 2005,
911: \apj, 627, 390 (Paper 1)
912:
913: \bibitem[Gotthelf \etal(1999)]{got99}
914: Gotthelf, E. V., Vasisht, G., Dotani, T. 1999, \apj, 522, L49
915:
916: \bibitem[Graves \etal(2005)]{gra05}
917: Graves, G. J. M., \etal\ 2005, \apj, 629, 944
918:
919: \bibitem[Hailey \& Mori(2002)]{hai02}
920: Hailey, C. J., \& Mori, K. 2002, \apjl, 578, L133
921:
922: \bibitem[Harding \& Muslimov(2001)]{har01}
923: Harding, A. K., \& Muslimov, A. G. 2001, \apj, 556, 987
924:
925: \bibitem[Harding \& Muslimov(2002)]{har02}
926: --------. 2002, ApJ, 568, 862
927:
928: \bibitem[Kaplan \etal(2005)]{kap05}
929: Kaplan, D.~L., \etal\ 2005, \pasp, 117, 643
930:
931: \bibitem[Krause \etal(2005)]{kra05}
932: Krause, O., \etal\ 2005, Science, 308, 1604
933:
934: \bibitem[Liu \etal(2006)]{liu06}
935: Liu, D. B., Yuan, A. F., Chen, L., \& You, J. H. 2006, \apj, 644, 439
936:
937: \bibitem[Lorimer \etal(1995)]{lor95}
938: Lorimer, D.~R., Yates, J.~A., Lyne, A.~G., \& Gould, D.~M. 1995,
939: \mnras, 273, 411
940:
941: \bibitem[Manchester \etal(2005)]{man05}
942: Manchester, R. N., Hobbs, G. B., Teoh, A., \& Hobbs, M. 2005,
943: \aj, 129, 1993
944:
945: \bibitem[Menou \etal(1999)]{men99}
946: Menou, K., Esin, A. A., Narayan, R., Garcia, M. R., Lasota, J.-P.,
947: \& McClintock, J. E. 1999, \apj, 520, 276
948:
949: \bibitem[Mereghetti \etal(2002)]{mer02}
950: Mereghetti, S., De Luca, A., Caraveo, P. A., Becker, W., Mignani, R.,
951: \& Bignami, G. F. 2002, \apj, 581, 1290
952:
953: \bibitem[Mori \etal(2005)]{mor05}
954: Mori, K., Chonko, J. C., \& Hailey, C. J. 2005, \apj, 631, 1082
955:
956: \bibitem[Mori \& Hailey(2006)]{mor06}
957: Mori, K., \& Hailey, C. J. 2006, \apj, 648, 1139
958:
959: \bibitem[\"Ogelman \& Alpar(2004)]{oge04}
960: \"Ogelman, H., \& Alpar, M. A. 2004, \apj, 603, L33
961:
962: \bibitem[Page \etal(2004)]{pag04}
963: Page, D., Lattimer, J., Prakash, M., \& Steiner, A. W. 2004, \apjs, 155, 623
964:
965: \bibitem[Park \etal(2004)]{par04}
966: Park, S., Zhekov, S. A., Burrows, D. N., Garmire, G. P., \& McCray, R.
967: 2004, \apj, 610, 275
968:
969: \bibitem[Pavlov \& Bezchastnov(2005)]{pav05}
970: Pavlov, G., G., \& Bezchastnov, V. G. 2005, \apjl, 635, L61
971:
972: \bibitem[Pavlov \etal(2004)]{pav04}
973: Pavlov, G. G., Sanwal, D., \& Teter, M. A. 2004, in IAU Symp 218,
974: Young Neutron Stars and their Environments,
975: ed. F. Camilo \& B. M. Gaensler (San Francisco: ASP), 239
976:
977: \bibitem[P\'erez-Azor\'in \etal(2006)]{per06}
978: P\'erez-Azor\'in, J. F., Miralles, J. A., \& Pons, J. A. 2006, \aap, 451, 1009
979:
980: \bibitem[{Ransom(2001)}]{ran01}
981: Ransom, S.~M. 2001, PhD thesis, Harvard University
982:
983: \bibitem[Ransom \etal(2002)]{ran02}
984: Ransom, S.~M., Eikenberry, S.~S., \& Middleditch, J. 2002, \aj, 124, 1788
985:
986: \bibitem[Rutledge \etal(2007)]{rut07}
987: Rutledge, R. E., Bildsten, L., Brown, E. F., Chakrabarty, D., Pavlov, G. G.,
988: \& Zavlin, V. E. 2007, \apj, 658, 514
989:
990: \bibitem[Sanwal \etal(2002)]{san02}
991: Sanwal, D., Pavlov, G. G., Zavlin, V. E., \& Teter, M. A. 2002, \apj, 574, 61
992:
993: \bibitem[Seward \etal(2003)]{sew03}
994: Seward, F. D., Slane, P. O., Smith, R. K., \& Sun, M. 2003, \apj, 584, 414
995:
996: \bibitem[Shi \& Xu(2003)]{shi03}
997: Shi, Y. \& Xu, R. X. 2003, \apj, 596, L75
998:
999: \bibitem[Sun \etal(2004)]{sun04}
1000: Sun, M., Seward, F. D., Smith, R. K., \& Slane, P. O. 2004, \apj, 605, 742
1001:
1002: \bibitem[Tauris \& Manchester(1998)]{tau98}
1003: Tauris, T.~M., \& Manchester, R.~N. 1998, \mnras, 298, 625
1004:
1005: \bibitem[Thompson \& Duncan(1993)]{tho93}
1006: Thompson, C., \& Duncan, R. C. 1993, \apj, 408, 194
1007:
1008: \bibitem[Turner \etal(2003)]{tur03}
1009: Turner, M. J. L., Briel, U. G., Ferrando, P., Griffiths, R. G.,
1010: \& Villa, G. E. 2003, Proc. SPIE, 4851, 169
1011:
1012: \bibitem[Wang \etal(2007)]{wan07}
1013: Wang, Z., Kaplan, D. L., \& Chakrabarty, D. 2007, ApJ, 655, 261
1014:
1015: \bibitem[Wang \etal(1998)]{wan98}
1016: Wang, F. Y.-H., Ruderman, M., Halpern, J. P., \& Zhu, T. 1998, \apj, 498, 373
1017:
1018: \bibitem[Woods \etal(2006)]{woo06}
1019: Woods, P. M., Zavlin, V. E., \& Pavlov, G. G. 2006, \aap, in press
1020: (astro-ph/0608483)
1021:
1022: \bibitem[Xu \etal(2003)]{xu03}
1023: Xu, R.-X., Wang, H.-G., \& Qiao, G.-J. 2003, Chinese Phys. Lett., 20, 314
1024:
1025: \bibitem[Yakovlev \etal(2002)]{yak02}
1026: Yakovlev, D. G., Kaminker, A. D., Haensel, P., \& Gnedin, O. Y. 2002,
1027: \aap, 389, L24
1028:
1029: \bibitem[Zavlin \etal(2004)]{zav04}
1030: Zavlin, V. E., Pavlov, G. G., \& Sanwal, D. 2004, \apj, 606, 444
1031:
1032: \end{thebibliography}
1033:
1034: \end{document}
1035:
1036: