102
P. Esposito et al.
ray spectrum of 4U 0142+61. Also searches in the GeV and Tev energy bands gave
negative results [2, 144].
The hard components can be variable, but in general their luminosity is comparable with or larger than that measured below 10 keV; also in the sources in which
the hard X-ray emission has not been detected, the upper limits do not exclude a
substantial contribution to the total luminosity. A peculiar feature of the hard tails
is that the pulsed emission has a harder spectrum than the averaged hard X-ray one
and also shows phase dependent variations (also morphological changes in the pulse
profiles and peak shifts with energy are observed; e.g. Hartog et al. [47], Götz et al.
[89], see also Fig. 3.1).
The origin of the hard tails of magnetars is still poorly understood but in general
the mechanism suggested, similarly to what proposed for the soft spectra, is the upscattering on ultra-relativistic electrons with Lorentz factor γ 1 [13, 67, 229],
possibly associated to relativistic outflows near the neutron star [15, 16].
3.2.1.3 Optical or Infrared Emission
Optical or infrared counterparts have been found for about one-third of the known
magnetars (e.g. Israel et al. [116], Mignani [155]). The search is complicated by
the intrinsic faintness of magnetars at that wavelengths and by their location in
crowded and heavily absorbed regions in the Galactic plane, but in most cases
the associations are strengthened by the detection of long-term variability. In three
cases, the association is firm because the spin modulation has been detected also
in the optical band (4U 0142+61, Kern and Martin [129], Dhillon et al. [49];
1E 1048.1–5937, Dhillon et al. [50]; SGR J0501+4516, Dhillon et al. [51]). As
anticipated, magnetars are variable also in the infrared/optical range, but it is unclear
(possibly because of the lack of adequate multi-wavelength campaigns) whether the
changes trace the X-ray flux evolution, as also cases of anti-correlated or simply
erratic variations have been reported [25, 51, 57, 201, 206].
The magnetar 4U 0142+61 has been detected in both infrared and optical bands
and is the one for which the greatest wealth of data is available at these wavelengths
[105, 106]. In near infrared, it shows an excess with respect to the blackbody that fits
the optical data; indeed, a multi-temperature (700–1200 K) thermal model provides
a better fit to the data. Wang et al. [230] suggested that the infrared component arises
from an extended disk (possibly from supernova fallback) illuminated from the
star’s X-rays and passively heated. This interpretation is supported by the correlation
observed in this source between the X-ray and infrared emissions [201]. On the
other hand, infrared/optical emission is expected from the inner magnetosphere,
a pair-dominated region where the curvature radiation should be able to produce
the observed infrared/optical luminosity [239]. Moreover, a magnetospheric origin
would account more easily for the observed optical pulsations, with profiles nearly
aligned with those observed at X-rays and displaying similarly broad modulation
and large pulsed fraction (20–50%). It is also possible that the infrared and optical
emissions have different origins or that the infrared excess in 4U 0142+61 is not
P. Esposito et al.
ray spectrum of 4U 0142+61. Also searches in the GeV and Tev energy bands gave
negative results [2, 144].
The hard components can be variable, but in general their luminosity is comparable with or larger than that measured below 10 keV; also in the sources in which
the hard X-ray emission has not been detected, the upper limits do not exclude a
substantial contribution to the total luminosity. A peculiar feature of the hard tails
is that the pulsed emission has a harder spectrum than the averaged hard X-ray one
and also shows phase dependent variations (also morphological changes in the pulse
profiles and peak shifts with energy are observed; e.g. Hartog et al. [47], Götz et al.
[89], see also Fig. 3.1).
The origin of the hard tails of magnetars is still poorly understood but in general
the mechanism suggested, similarly to what proposed for the soft spectra, is the upscattering on ultra-relativistic electrons with Lorentz factor γ 1 [13, 67, 229],
possibly associated to relativistic outflows near the neutron star [15, 16].
3.2.1.3 Optical or Infrared Emission
Optical or infrared counterparts have been found for about one-third of the known
magnetars (e.g. Israel et al. [116], Mignani [155]). The search is complicated by
the intrinsic faintness of magnetars at that wavelengths and by their location in
crowded and heavily absorbed regions in the Galactic plane, but in most cases
the associations are strengthened by the detection of long-term variability. In three
cases, the association is firm because the spin modulation has been detected also
in the optical band (4U 0142+61, Kern and Martin [129], Dhillon et al. [49];
1E 1048.1–5937, Dhillon et al. [50]; SGR J0501+4516, Dhillon et al. [51]). As
anticipated, magnetars are variable also in the infrared/optical range, but it is unclear
(possibly because of the lack of adequate multi-wavelength campaigns) whether the
changes trace the X-ray flux evolution, as also cases of anti-correlated or simply
erratic variations have been reported [25, 51, 57, 201, 206].
The magnetar 4U 0142+61 has been detected in both infrared and optical bands
and is the one for which the greatest wealth of data is available at these wavelengths
[105, 106]. In near infrared, it shows an excess with respect to the blackbody that fits
the optical data; indeed, a multi-temperature (700–1200 K) thermal model provides
a better fit to the data. Wang et al. [230] suggested that the infrared component arises
from an extended disk (possibly from supernova fallback) illuminated from the
star’s X-rays and passively heated. This interpretation is supported by the correlation
observed in this source between the X-ray and infrared emissions [201]. On the
other hand, infrared/optical emission is expected from the inner magnetosphere,
a pair-dominated region where the curvature radiation should be able to produce
the observed infrared/optical luminosity [239]. Moreover, a magnetospheric origin
would account more easily for the observed optical pulsations, with profiles nearly
aligned with those observed at X-rays and displaying similarly broad modulation
and large pulsed fraction (20–50%). It is also possible that the infrared and optical
emissions have different origins or that the infrared excess in 4U 0142+61 is not