Science

Introduction

The hard X-ray band from 10 to 100 keV is a relatively poorly explored region of the electromagnetic spectrum. While progress may have been impeded because it lies between the traditional X-ray and gamma-ray bands, each with its own separate community of advocates, the root cause is technical in nature. The background is high in this band because of the bright cosmic diffuse flux and until now, focusing optics have been ineffective above ~10 keV. However, the scientific opportunities in this band are great and worthy of pursuit. The technical barriers are falling and scientific excitement is rising. Now is a particularly appropriate time to propose the next step.

As its name implies, the low energy gamma-ray or hard X-ray band is a transitional region between X-ray and gamma-ray astronomy. It is the band where thermal emission blends in with non-thermal, where Comptonization first becomes important and where nuclear lines first show up. The measurement of these transitions provides unique physical diagnostics of the source regions. For an instrument like InFOCmS that breaks through the technical barrier, the scientific potential is enormous. In this section, we summarize the exciting possibilities.

In addition to the known science objectives summarized in the sections below, there will be many serendipitous discoveries. Whenever observational capabilities are improved by 1 to 2 orders of magnitude, as they will be with InFOCmS and satellite follow-ons, new areas of astrophysics will be opened that we can not even imagine.

⁴⁴Ti Gamma-Ray Lines From Young Supernova Remnants

Since SN1987A the long predicted (Clayton et al. 1969) detection of radioactive lines from supernovae has become a reality. The most intense lines are expected from ⁵⁶Ni->⁵⁶Co->⁵⁶Fe (mean life 113.2 days), followed by those from ⁵⁷Co->⁵⁷Fe (mean life 392 days) and ⁴⁴Ti->⁴⁴Sc->⁴⁴Ca (mean life between 78 and 96 years). The detection of the ⁵⁶Co and ⁵⁷Co lines from SN1987A (Matz et al. 1988; Tueller et al. 1990) has, for the first time, placed explosive nucleosynthesis theories on an observational basis. In addition, the Doppler broadening of the g-ray lines in the expanding nebula reflects the velocity distribution modified by the opacity along the line of sight. Hence, observations of these lines provide the most direct test of current models of both explosive nucleosynthesis and the dynamics of SN ejecta (Clayton 1974; Chan & Lingenfelter 1991; Gehrels et al. 1987; Ruiz-Lapuente et al. 1993). The InFOCmS experiment has been explicitly designed to observe the ⁴⁴Ti produced in Galactic SNe over the last few hundred years to make the next significant steps in this discovery process.

The ⁴⁴Ti decay to ⁴⁴Sc produces lines at 67.9 keV (100%) and 78.4 keV (98%). Theoretical models of Type Ib and Type II Sne (Woosley & Hoffman 1991; Thielemann et al. 1990; Hashimoto et al. 1989; Ensman & Woosley 1988; Nomoto et al. 1984; Woosley & Weaver 1982) yield ⁴⁴Ti mass values of ~1.0 x 10^-4 M. The longer life of ⁴⁴Ti allows us to observe these lines from Galactic SNe for several hundred years after the explosion and discover the recent SNe in our Galaxy. Historical records have allowed us to identify only 2 or 3 nearby Galactic SNe within the past 300 years, but the estimated (van den Bergh & Tammann 1991) Galactic SNe rate gives an expected number of 14 Galactic supernovae in 300 years for a Hubble constant of 75 km s^-1 Mpc^-1. The detection of ⁴⁴Ti decay lines could in principle allow the discovery of other recent optically obscured SNe.

COMPTEL and OSSE on CGRO have detected the 1.157 MeV line from ⁴⁴Ti, from Cas A (Iyudin et al. 1994, Schönfelder et al. 1996, The et al. 1995,1996). Cas A is thought to be the ~300 year old remnant of a Type Ib SN. Combining these measurements, the most probable ⁴⁴Ti line flux is 3.5 x 10^-5 photons cm^-2 s^-1. While observations with the HEXTE instrument on XTE has not been able to confirm these lines, its 9 keV resolution and systematics-limited sensitivity do not produce a clear conflict with the COMTEL results. In contrast, a raster scan using InFOCmS will be photon limited and the two lines be resolved. The sensitivity for each of the lines is ~2 x 10^-6 ph cm^-2 s^-1(with a 0.5 arcmin resolution mirror and 2 keV detectors for 12 hours. It will be possible to crudely map the line emission in a single observation for the compact source at the center of the remnant predicted by the simplest models.

The 100-day exposures afforded by ultra long duration balloon flights would provide an opportunity for very detailed study to map the three dimensional structure of the ⁴⁴Ti, which can be compared directly with that of the other ejecta. A 2-dimensional map with < 2 arcmin angular resolution can be constructed for each line. Information about the third dimension comes from a measurement of the line centroid to an accuracy of 100 eV for each spatial bin, and thus allow us to construct a Doppler map of Cas A, with velocity resolution of 400 km s^-1. A comparison can then be made with the results of X-ray Doppler mapping, which has revealed that the ejecta are confined to an inclined ring, with characteristic velocity of ~2,000 km s^-1 (Holt et al. 1994).

Most recently, COMPTEL has also reported a positive detection of ⁴⁴Ti emission from the Vela region at the flux level of 3.8 x 10^-5 photons s^-1 cm^-2 (Iyudin et al. 1998). The centroid of this emission coincides with a newly discovered supernova remnant in X-rays (Aschenbach et al. 1998), which has a radius of 1 degree and an apparent shell expansion speed of 3000 km s^-1. Analysis of the observed ⁴⁴Ti flux and the X-ray properties tells us that this supernova occurred about 700 years ago at a distance of only 150 pc away (Chen & Gehrels 1999). Because of the more advanced age and greater angular extent than Cas A, the InFOCmS observation of this new ⁴⁴Ti supernova remnant, GRO J0852-4642, may offer a better opportunity to answer the important question of whether the ⁴⁴Ti is concentrated at the center of the remnant or distributed in a filled sphere.

Of the other known Galactic supernova remnants only Tycho (1572 AD, d ~ 3.5 kpc) and Kepler (1604 AD, d ~ 4.4 kpc) are potentially detectable in ⁴⁴Ti lines by InFOCmS (both are below the HEXTE threshold). Observing these two remnants would provide important information about nucleosynthesis models. Tycho is the remnant of a Type Ia supernova. Observing the ⁴⁴Ti lines from it would allow a direct comparison of the ⁴⁴Ti production between carbon deflagration/detonation and core collapse SNe. The nature of the Kepler progenitor is a subject of controversy. If differences are measured in the ⁴⁴Ti yields from the Tycho and Cas A progenitors, then a measurement of the ⁴⁴Ti line strength in Kepler can resolve this controversy. If the same quantity of ⁴⁴Ti were produced in all three SNe, a simple scaling of the Cas A flux indicates an integrated flux in each line for either Tycho or Kepler of ~6 x 10^-6 photons cm^-2 s^-1. While the reduced instrumental sensitivity as a consequence of the remnants’ finite spatial extent (4' for Kepler; 8' for Tycho) would place them below the sensitivity threshold during a 12.5 hour exposure, their ⁴⁴Ti emission could be mapped during a week-long exposure from a long-duration balloon flight. Similarly, SN1987A, with an expected flux in each line of a few x 10^-6 photons cm^-2 s^-1, would be ideally suited for a long InFOCmS exposure.

Synchrotron Emission from Shock-Accelerated Electrons in SNR

The discovery of synchrotron emission from shock-accelerated electrons in several supernova remnants via spatially resolved X-ray spectroscopy with ASCA (SN1006 Koyama et al. 1995), IC 443 (Keohane et al. 1997), Cassiopeia A (Allen et al. 1997), RXJ1713.7-3946 (Koyama et al. 1997), and G266.3-1.2 (Aschenbach 1998) provided direct proof that the shocks of supernova remnants are an acceleration source for cosmic rays up to 100 TeV. However, this still falls short of the 1000 TeV "knee" of the observed cosmic-ray spectrum (Axford, 1992), where the mechanism of cosmic-ray emission is believed to change. In the higher energy band, the increased sensitivity and imaging capability of InFOCmS provides an opportunity to probe synchrotron emission from electrons accelerated to energies up to 600 TeV.

Reynolds (1998) predicted that such synchrotron emission should be common among young supernova remnants, however observations confirming this have been slow to develop because synchrotron emission at lower energies is blended with extremely strong thermal bremsstrahlung and thermal line emission from the shocked, ionized plasma. However, above 10 keV these emission mechanisms drop off rapidly and any synchrotron emission emitted from the remnant will dominate.

InFOCmS has a spatial resolution comparable to the ASCA satellite, therefore it will be possible to differentiate shell emission from plerionic emission in Galactic supernova remnants. This will provide a unique opportunity to search young supernova remnants dominated by thermal emission below 10 keV for evidence of shock-accelerated cosmic rays.

Finally, sometime within the next decade, the blast wave of SN1987A will encounter its circumstellar shell (Luo & McCray 1991). The resulting young supernova remnant might be become extremely luminous, approaching 10³⁸ erg s^-1. While the primary emission bands are expected to be the UV and soft X-ray, it is possible that some hard X-ray emission will be associated with this event. The InFOCmS experiment will have a 20 to 40 keV luminosity threshold for SN1987A of 5x10³⁵ erg s^-1, and so be able to place a very sensitive upper limit on the hard flux. Table 1 is a sample of young supernova remnants that will be observable with InFOCmS. Surface brightness values not referenced in this table are preliminary results from fits to GIS and SIS ASCA data to a model comprised of a power law and the non-equilibrium ionization model of Hamilton et al. (1983).

Intergalactic Magnetic Fields

The strength of the magnetic field in intergalactic space is one of the great unknowns in astronomy. Its value is unknown even to an order of magnitude, yet it has a profound effect on many important problems, including equipartition in the IGM, kinematics of AGN jets, intergalactic cosmic-ray propagation, and the structure and evolution of classical double radio sources. By observing the Compton upscattered cosmic microwave background (CMB) from the radio lobes at the terminus of AGN jets, InFOCmS can provide a new measurement that will allow the first direct calculation of the field strength. The radio emission from the lobes is produced by synchrotron radiation from jet particles probably electrons and positrons which are confined by the intergalactic magnetic field. The radio luminousity is a function of the particle densities and magnetic field strength and without an independent measurement, one can only assume equipartition to calculate these crucial factors. This assumption is clearly dubious for a dynamic system like a jet-powered radio lobe. Measurement of an independent parameter could enormously enhance our understanding of the AGN jets, the IGM, and their interaction.

Hard X-rays are produced when the relativistic leptons in the lobes Compton up-scatter the CMB photons. The Lorentz factor required to boost the CMB photons to the 20 to 40 keV band is g ~ (30 keV / 2.32 x 10^-7 keV)^1/2 ~10⁴, which is about the same as that required to produce synchrotron radiation at 100 MHz in a 1 mG magnetic field. Therefore, the observed GHz radio emission from jet lobes also points to observable hard X-rays. The inverse Compton radiation will have the same energy spectral index a as the synchrotron radiation does, and their luminosities are directly correlated via L_icµ (U_CMB/U_B) L_syn, where U_CMB and U_B are the energy densities of the CMB radiation and IGM magnetic field. The energy density ratio reaches unity for B_IGM ~3 mG. Thus, the hard X-ray luminosity from the jet lobes is in general of the same order as the radio luminosity. Quantitatively, if the synchrotron flux is in the units of mJy, the inverse Compton flux will be:

F_ic ~ 100 B_mG^-2 E_{30 keV^-1} n_GHz F_syn,mJy mCrab

for a = 1. Thus 100 mCrab fluxes can be expected from the lobes.

The Compton upscattering spectrum mimics the electron spectrum that produces it, so there should be X-ray lobes equivalent to the hard X-ray lobes. Why hasn’t this measurement been completed already? The AGN is usually a very bright X-ray point source. The galaxy is usually surrounded by hot gas which produces a diffuse source of thermal X-rays that can be confused with the upscattered spectrum from the lobes. With a typical temperature < 6 keV in the X-ray band, this thermal source has also nearly disappeared above 20 keV. The best place to look for inverse Compton hard X-ray emission of the radio jets is where the lobe is relatively far away from the core (> 1' to 2'), the lobe radio flux is at least a few mJy with a relatively steep spectrum (a > 0.5), and the X-ray and radio emission from the core is weak. Tentative candidates are 3C61.1 (Liang 1981), 3C236 (Strom & Willis 1986) and NGC6251 (Perley et al. 1984).

Cyclotron Lines

Electrons in the ionized plasma accreting onto the poles of a neutron star will undergo cyclotron motion in the strong magnetic field (~10¹² to 10¹³ G). Transitions between discrete Landau levels can occur, giving rise to a series of hard X-ray lines at the fundamental cyclotron frequency and its harmonics. The energy of the lines is a direct measure of the magnetic field strength, but high resolution observations of the line profiles (fundamental and harmonics) as a function of pulse phase can provide a wealth of new information on the spatial distribution of both the field and accretion flow (Meszaros and Nagel 1985; Wang et al. 1989; Mihara 1995).

Cyclotron features have been detected from the X-ray pulsars, Her X-1 (Trumper et al. 1978; Tueller et al. 1984), 4U0115+63 (Wheaton et al. 1979), Vela X-1 (Kendziorra et al. 1992), and GX1+4 (Maurer et al. 1982). Recently, the Ginga satellite added X0331+53, Cep X-4, 4U1907+09, 4U1538+52, and GX301-2 (Makishima & Mihara 1992). All of these sources have their fundamental or a strong first harmonic in the 20 to 40 keV band of our instrument. The best studied of these is Her X-1 where the cyclotron line is pulsed with a flux ~10^-3 photons cm^-2 s^-1 and a width of ~10 keV. The energy of the line was found to shift with phase of the 1.24 sec pulsation (Gruber et al. 1980). Changes in the absorption and scattering properties of the accretion column and changes in the viewing angle appear as changes in the cyclotron line profile. While many high sensitivity NaI and proportional counter (low resolution) spectra of Her X-1 exist, we cannot even say unequivocally whether the line is seen in absorption or emission. InFOCnS will detect ~1 count s^-1 on average in the 20 to 40 keV band from Her X-1 and Vela X-1. Thus, in a 12 hour observation, we can make a 20s measurement in ~100 phase bins. InFOCmS can provide the pulse-phase-resolved high-resolution measurements of the fundamental at 35 keV and perhaps the first harmonic at 70 keV to yield a clearer understanding of the emission region. We expect from InFOCmS and Constellation-X results to revolutionize our understanding of accretion on X-ray pulsars.

Galactic Black Holes

The dynamical evidence for accreting black holes in X-ray binaries now seems conclusive (Tanaka & Lewin 1995). Most of the black hole systems are X-ray transient systems, where the mass-donor is a late-type star. Accreting black holes show a distinct spectral signature of an ultra-soft component with a characteritic temperature of ~1 keV and/or an ultra-hard power law component that extends up to several hundred keV (e.g. Wilson and Rothschild 1983; Sunyaev et al. 1991). In contrast, X-ray pulsars exhibit flat power law spectra up to 10 to 20 keV, which then decay rapidly (White et al. 1994). Low magnetic field neutron star systems exhibit either spectra with a characteristic temperature of 5 to 10 keV, or power law spectra with a photon index of 2 to 3 (White et al. 1994). In their low luminosity state, some X-ray burst sources show a power law spectrum that seems to extend to high energies (Mitsuda et al. 1989). The photon index in these cases may be similar to that seen from the BHC in the high state. Sensitive observations of the X-ray continuum of low luminosity X-ray binaries above 20 keV have been difficult because of lack of sensitivity in this band. The InFOCmS will be able to make a sensitive search for a hard X-ray tail from X-ray binaries and shed important information on the hard X-ray properties of neutron stars and black hole candidates.

Hard X-Ray Imaging Observations of Clusters of Galaxies

X-ray emission from clusters of galaxies in the 0.1 to 10 keV band originates primarily from a hot intracluster gas with a temperature of 1 to 10 keV, which contributes only a small fraction to the E > 20 keV spectrum. Hard X-ray observations in the 20 to 100 keV band have been performed with non-imaging large area detectors to search for a non-thermal component in clusters. The observational results from a wide variety of instruments are limited to ~1x10^-5 ph cm^-2 s^-1 keV^-1 in the 20 to 100 keV band, ~10 X poorer than InFOCmS. The hard X-ray detections are interpreted as the inverse Compton (IC) scattering of high energy electrons (which produce the diffuse non-thermal synchrotron radio emission observed in clusters) off the 2.7K cosmic microwave background.

Measurement of the relative intensity of the inverse Compton to synchrotron flux is the only direct measurement of the intracluster magnetic field. It is therefore believed that the hard X-ray emission is diffuse in nature and extended on a cluster scale. For at least one cluster (A2199, Kaastra et al.) the intensity of the hard component is so strong that it changes the temperature of the hot gas component and radically changes the inferred dark matter mass and the ratio of baryonic to dark matter. If this result is correct and applies to other clusters it could require a radical change in our present understanding of the structure and evolution of clusters.

Recently, diffuse EUV emission has been observed by EUVE. This extremely soft extended component has been also interpreted as due to a non-thermal process owing to synchrotron emission from high energy electrons of 150 MeV. In these cases the cluster magnetic field has to be weaker and the ratio of IC to radio synchrotron component would be rather large.

In addition spectro-imaging analysis of the ASCA spectra of several clusters (Perseus and A3266) indicates the existence of point-like objects in the central core region with very hard spectra (probably AGN) indicating that a considerable fraction of the flux seen by non-imaging instruments in clusters may not be diffuse.

These tantalizing results indicate that hard X-ray imaging observations are critically important for our understanding of clusters, the amount and distribution of dark matter, strength of the magnetic field and the origin and distribution of relativistic particles. Moreover the imaging spectroscopic analysis is also useful to investigate the non-thermal processes.

We expect to detect hard X-rays from both point sources and the diffuse component in clusters of galaxies. We select the Perseus cluster as the first target, in which NGC1275 is located as a point source in the central region slightly shifted from the cluster center.

Active Galactic Nuclei

It is well accepted that the various classifications of active galactic nuclei arise from the same physical object being viewed at different inclination angles (Antonucci & Miller 1985, Miller & Goodrich 1990, Antonucci 1993, Urry & Padovani 1995). It is believed that these inclination effects are primarily caused by a thick dusty torus surrounding the central engine of the AGN, preventing optical, UV and soft X-rays from being directly observed from the central engine at certain inclination angles.

X-rays from a source above the AGN accretion disk can be "reflected" via inverse Compton scattering either from the torus or the accretion disk itself. This Compton reflection produces a significantly greater flux of hard X-rays than predicted by the intrinsic power law itself. This component is commonly referred to as the reflection hump (Lightman & White 1988, Guilbert & Rees 1988). This Compton reflection hump has played a significant role in detecting faint type II AGN (Pounds et al. 1990, Piro et al. 1990, Matt 1997, Malaguti et al. 1998). Another signature of Compton scattering is the presence of fluorescence Ka lines of neutral and ionized iron (Matt et al. 1991, Ghiselli et al. 1994, Krolik et al. 1994). However, there are still several AGN observations with ASCA that have observed iron fluorescence lines without observing a Compton reflection hump (Eracleous et al. 1996, Grandi et al. 1997, Yamashita & Inoue 1996, Weaver et al. 1997, Turner et al. 1997). Perhaps the Compton reflection continuum lies below the sensitivity of ASCA at these energies or there is a refinement to Compton reflection theory that must be made. The Compton reflection hump is predicted to peak within the energy range 20 to 80 keV making InFOCmS an ideal instrument to explore Compton reflection in AGN.

Cosmic X-Ray Background

It is known that the cosmic X-ray background above 0.1 keV is comprised of unresolved AGN (Fabian & Barcons 1992). Recent ASCA deep surveys of deep ROSAT fields (e.g. Georgantopoulos et al. 1997) reveal a population of AGN not detected in the ROSAT survey due to absorption. The spectral slope of faint type II AGN matches the observed spectral slope of the cosmic X-ray background of G=1.5 better than type I AGN (G=1.9, Nandra & Pounds 1994), making obscured Sy II and narrow emission line galaxies background source candidates. However, the survey has only resolved about 30% of these background sources.

Type II AGN are believed to be the dominant source of background X-rays above 2 keV in the local universe, but we are currently unsure about their contribution at higher redshifts (Boyle et al. 1998a). There are very few type II AGN that have been observed at redshifts z > 0.1 (Boyle et al. 1998b, Ohta et al. 1996). Compton reflection should brighten these objects in the 20 to 80 keV range and allow InFOCmS to play a key role in identifying high-redshift type II AGN.