Our current understanding of the full census of AGN is highly incomplete. Soft X-ray observatories are strongly biased towards unabsorbed sources (Figure 1), while the shape of the cosmic X-ray background implies approximately half of SMBH growth is obscured. Non-focusing hard X-ray missions resolved just ∼ 2 % of the X-ray background at its peak into individual sources (Bottacini et al., 2012), while NuSTAR, the first focusing hard X-ray telescope in orbit, resolved 35% (Harrison et al., 2016). To bring our understanding of the hard X-ray background to the 75 − 90 % levels Chandra and XMM-Newton provide below 6 keV and complete the cosmic census of black hole growth requires improvements in both sensitivity and angular resolution in future hard X-ray telescopes.

The deepest NuSTAR survey detects only 4 sources at an 8 − 24 keV depth of 5 × 1 0 − 14 erg cm − 2 s − 1 (Zhao et al., 2021) (throughout, survey depths are reported at a 99% Poisson reliability threshold at 50% area, i.e., 50% of the total survey area reaches that depth or deeper Zhao et al., 2024). At this depth, NuSTAR is reaching the classical confusion limit in the 3 − 8 keV band, corresponding to 150 sources per deg² for the NuSTAR 6 0 ″ half-power diameter (HPD). Deeper surveys will require improved angular resolution to detect the hard X-ray sources.

In the study presented by Civano et al. (2024), the authors demonstrate via detailed simulations of extragalactic field observations that HEX-P will detect over 1,000 sources at 10 − 40 keV, compared to the four sources in the deepest NuSTAR survey in the comparable 8 − 24 keV band (Zhao et al., 2024). Their simulations of a deep and a wide area survey include full instrument models and source detection analysis. They show that with HEX-P, the statistical uncertainty in the measured resolved 10 − 40 keV X-ray background fraction is predicted to be ≤ 6 % for a range of population models (Gilli et al., 2007; Ananna et al., 2019), while the resolved fractions from the models themselves differ by more than 20% in the same energy band.

Deep, broadband extragalactic surveys will, for the first time, uncover the full population of accreting SMBHs and constrain their properties. If HEX-P finds that the heavily obscured fraction is constant with luminosity, this would favor radiation-regulated growth in which radiation pressure dictates the geometry of the obscuring structure (Ananna et al., 2022). If HEX-P instead finds fewer obscured AGN, it might favor models where obscuration occurs primarily in the lowest luminosity AGN, below the survey depth (Lawrence, 1991), or that the SMBH spin rather than obscuration is partially responsible for the characteristic ∼ 25 keV peak in the cosmic X-ray background (Vasudevan et al., 2016).

Black holes are the protagonists of one of the most fundamental problems in physics: understanding the complete gravitational collapse of matter to a point that lies within a light-trapping, one-way membrane known as the event horizon. Importantly, only two properties fully define an astrophysical black hole: its mass M and its angular momentum J , where the latter is commonly reported as the dimensionless spin parameter a * = J c / G M 2 (where c and G are the speed of light and gravitational constant, respectively); a * is constrained to values of − 0.998 ≤ a * ≤ 0.998 (Thorne, 1974). Furthermore, mass and spin are fundamentally related. Regardless of its assumed formation channel, a black hole can both increase its mass and change its spin via accretion of matter within its gravitational potential, or via repeated mergers with other black holes. Theoretical models for black hole evolution over cosmic time have specific predictions for how the SMBH mass-spin distribution in the local Universe depends on whether their growth history is dominated by ordered accretion, chaotic accretion or mergers (e.g., Berti and Volonteri, 2008; Dotti et al., 2013; Griffin et al., 2019; Izquierdo-Villalba et al., 2020). Most cosmological simulations predict accretion-dominated growth in lower-mass SMBHs ( < 1 0 8   M ⊙ ), leading to near-maximal expected average spins (e.g., Dubois et al., 2014; Dubois et al., 2021). In semi-analytic models where ordered accretion dominates even in the most massive SMBHs ( > 1 0 9 M ⊙ ), spin parameter values remain high (Sesana et al., 2014), while other models, which follow the hydrodynamics of gas within galaxies, predict an increasing contribution of chaotic accretion and mergers to SMBH growth, leading to lower spin parameter values (Bustamante and Springel, 2019; Beckmann et al., 2024).

Determining whether mergers play a significant role in the growth of SMBHs observed in the local Universe requires a comprehensive spin survey for a sample of SMBHs across a range in black hole mass. While there exists a selection of well-developed methods for measuring SMBH mass, such as reverberation mapping, single-epoch virial estimation or stellar and gas dynamical modelling (see Peterson, 2014, for an overview of the methods), constraining the spin parameter is significantly more challenging. The most effective method is known as X-ray reflection spectroscopy, and is based on measuring the distortion of atomic spectral features by the strong gravitational potential of the SMBH (see Reynolds, 2021, for a review). The hallmarks of reflection are the inner-shell emission lines of iron (Fe K, ∼ 6.5 keV), the Fe absorption K-edge ( ∼ 7 keV), and the Compton reflection hump ( ∼ 30 keV), all observed in the broad HEX-P band (Figure 5A). High spin black holes allow for an inner edge of the accretion disk extended closer to the coordinate system origin than their low spin counterparts, leading to greater relativistic distortion of the spectral features. By measuring the degree of this distortion, one can thus infer the inward radial extent of the accretion disk which serves as a proxy for the innermost stable circular orbit and, by extension, black hole spin. Given the resulting width of these features, high spectral resolution of the instrument is not a priority, especially in the absence of absorption features caused by intervening material in the line of sight. ³ Instead, a broadband coverage is required to robustly constrain the underlying continuum and isolate the spectral reflection signal. This requirement cannot be fulfilled by soft X-ray observatories whose sensitivity drops rapidly above 8 keV.

Piotrowska et al. (2023) present simulated HEX-P spin parameter measurements for a survey of a 100 AGN selected from the Swift-BAT AGN Spectroscopic Survey (Koss et al., 2017). This sample, complete with published SMBH masses (Koss et al., 2022), includes the brightest hard-X-ray-selected unobscured AGN across the sky observed to date. The study makes use of the Horizon-AGN cosmological simulation to determine spin–mass distributions of SMBHs with different growth histories, which occupy distinct loci in the parameter space. In simulating HEX-P observations, the authors account for measurement uncertainty, model limitations and observational biases, including the strong dependence of radiative efficiency on spin which results in a high-spin bias in flux-limited samples (Figure 5A). The suite of simulated measurements demonstrates that HEX-P robustly distinguishes accretion-dominated models from models where mergers play a significant role in SMBH growth in realistic samples of HEX-P X-ray reflection measurements. As shown by the solid white regions in the right panel of Figure 5, current spin constraints are unable to discriminate between these scenarios, as both SMBH growth histories lie well within the currently permitted spin–mass values. Constraints comparable to HEX-P are not feasible with existing facilities as they would require a year of coordinated XMM-Newton and NuSTAR time.

This large survey, besides answering the Astro2020 key question about how SMBHs grow, also addresses the impact of SMBHs on their host galaxies. With inner disk radii closer to the SMBH, higher spin AGN are also more luminous (Novikov and Thorne, 1973) and thus have a greater energy budget to affect their host galaxy’s evolution. Blue-shifted absorption due to fierce, outflowing winds is the clearest indicator of feedback on accretion-disk scales and requires continuum measurements at high energies to constrain line properties (e.g., Fe K is shifted to > 8 keV for winds of > 0.2 c , requiring sensitivity above 10 keV; Tombesi et al., 2010; Tombesi et al., 2011). Parker et al. (2022) highlight the importance of Compton reflection hump measurements to disentangle relativistic reflection from the signal generated by disk winds. Higher spectral resolution at soft energies is less beneficial as the broad features are easily resolved with typical resolution.

Most of the energetic photons in the universe originate from the X-ray corona, a highly luminous plasma of hot electrons close to the central engine of accreting compact objects (Figure 6). Reverberation and microlensing studies indicate that the corona is extremely compact, extending only a few gravitational radii ( R g = G M / c 2 ) from the central accreting object (neutron star or black hole; Pooley et al., 2007; De Marco et al., 2013; Mosquera et al., 2013; Uttley et al., 2014). Incredibly, the coronae of SMBHs in luminous AGN easily outshine all other high-energy emission in the host galaxy. Per volume, AGN coronae are the most luminous, persistent objects in the universe—and reside partnered with the darkest objects in the universe, black holes. However, beyond a few simple characteristics, the origin and physical mechanisms that govern and power the X-ray corona are poorly understood. HEX-P will fundamentally change this, providing key observations and tests of the leading theoretical models. Since the corona dominates the intrinsic X-ray emission of accreting objects, understanding it underlies the AGN population studies in Sections 3.1.1, 3.1.2.

The accretion disk, heated by effective viscosity to high temperatures, produces a blackbody spectrum that peaks at ∼ 0.1 − 1 keV for stellar-mass compact objects (Figure 2), and in the UV for SMBHs. Thermal disk photons interact with relativistic electrons in the much hotter corona ( T e ∼ 1 0 8 − 1 0 9 K), increasing the photon energies after repeated inverse Compton scatterings with electrons. The result is a power-law continuum extending up to several hundred keV, and this continuum is more cleanly measured in AGN. For stellar-mass black holes, changes in the X-ray luminosity are also associated with changes in the accretion disk such as the level of the inner-disk truncation, making Galactic binaries poorly suited for this Objective. As coronal photons can only gain so much energy from the relativistic electrons, the power-law-like X-ray spectrum falls off rapidly at a few times the characteristic electron temperature (refereed to as the high-energy cutoff; Titarchuk and Lyubarskij, 1995, and Figure 6). The electron temperature is the result of the corona’s heating-to-cooling balance, where heating takes place magnetically and cooling is set by the number of very high-energy X-ray photons produced. Even a small number of such photons in the compact corona produces copious amounts of electron-positron pairs to share the available thermal energy, cooling the electrons rapidly until a balance is reached. The maximum temperature allowed in this pair-production thermostat model is related to the coronal luminosity.

Kammoun et al. (2024) discusses different tests that can be carried out with a mission like HEX-P to validate this paradigm, along with other coronal studies that could be executed as GO programs. One test focuses on a variable AGN (Test 1; Figure 7, left). These sources provide a powerful tool to test the pair-production thermostat model of the X-ray corona since the model predicts specific tracks along the temperature-luminosity plane. As the luminosity increases, the model predicts that more electron-positron pairs are produced, lowering the plasma temperature. Departures from this relation would indicate incomplete physics, invalidating the model.

Another test concerns luminous AGN (Test 2; Figure 7, right). The pair-production model of the corona also predicts that the highest luminosity sources are cooler since more electron-positron pairs are produced, thereby lowering the coronal temperature, and implying a restricted allowed range of temperature. Figure 7 shows simulations of the luminous sources with expected cut-off energies between 50 and 100 keV. Even if the rest-frame cut-offs were ≥ 200 keV and violating the pair-production model, HEX-P could still constrain the cut-offs to be ≥ 150 keV (90% confidence), which is still unallowed by the pair-production model. Note that high-luminosity AGN are preferentially at high redshift, thereby shifting cut-offs to lower observed energies where the HEX-P sensitivity is higher.

Tests like these would be prohibitive with current facilities as they would require more than 4 Ms of simultaneous XMM-Newton and NuSTAR observations (clock time) but will take only ∼ 500 ks for HEX-P.

Mori et al. (2024) presents simulated HEX-P surveys of the central regions of the Milky Way, while Lehmer et al. (2023) presents simulated HEX-P surveys of four nearby galaxies (M31, Maffei 1, NGC 253, and NGC 3310). These studies show how HEX-P observations can be used to investigate how X-ray populations depend on environment and metallicity. Milky Way studies are sensitive to CVs and X-ray binaries, while most detected extragalactic sources will be X-ray binaries. HEX-P’s broadband response, high timing resolution, and high angular resolution will enable unprecedented sensitivity for the detection of pulsations from sources in crowded environments where traditional timing-dedicated mission (with high effective areas but little imaging capabilities) would struggle, such as the Galactic Center (Mori et al., 2024) and nearby Galaxies (Bachetti et al., 2023).

Cosmic rays, charged extraterrestrial particles that continuously bombard the Earth’s atmosphere (Figure 10), exhibit a power-law flux distribution over more than 12 orders of magnitude in energy. Low energy cosmic rays are primarily solar protons, while the most energetic are primarily extragalactic. At the “knee” of the distribution near 1 0 15 eV = 1 PeV, the particles are believed to be primarily Galactic from “PeVatron” sources, the most energetic acceleration sites in the Milky Way Galaxy. Understanding these extreme sources requires a multi-messenger approach, addressing the physics of supernova shock waves, pulsar wind nebulae, and other exotic environments, with information from radio, X-rays, gamma-rays, and neutrinos (Figure 10).

Understanding Galactic PeVatron sources requires access to the 40 − 80 keV energy band. Since cosmic rays are charged particles, their trajectories are deflected by interstellar magnetic fields, disguising their origins. Energetic cosmic ray electrons inverse-Compton scatter ambient photons to produce very high-energy emission, up to ∼ 100 TeV (Harding, 1996). However, modeling this emission requires knowledge of the seed photon population (e.g., cosmic microwave background, nearby synchrotron, or nearby thermal sources), and the highest energy emission is affected by quantum electrodynamical effects i.e., Klein-Nishina suppression (Klein and Nishina, 1929). Energetic cosmic-ray electrons also emit synchrotron photons in the X-ray to MeV band. Since this emission reflects the particle spectrum more directly, it provides a key tool for investigating particle acceleration and transport (Reynolds, 2016; Burgess et al., 2022; Mori et al., 2022). A strategic observational campaign, including a list of proposed HEX-P targets, is discussed in detail in Mori et al. (2023).

These targets have significant scientific value beyond the scientific goal of understanding particle acceleration. For example, the morphology and kinematics of the ⁴⁴Ti emission at 78 keV in supernova remnants is a powerful diagnostic of supernova explosion physics. The capabilities offered by HEX-P in addressing this and other interesting problems on nuclear astrophysics are discussed at length in Reynolds et al. (2023).

HEX-P will be an essential resource in the 2030s, providing sensitive high-energy access to explore explosive events and monitor variable sources with other facilities. HEX-P will devote a fraction of its exposure time to unanticipated transient observations triggered by the community. This time could be considered a form of community GO access, though HEX-P conservatively assesses it as PI-led time since it is allocated by the PI rather than NASA. This will enable HEX-P to provide more flexible and rapid response to community needs. A few example potential programs are discussed below.

Accretion of matter onto a black hole is one of the most efficient mechanisms to produce energetic radiation. While SMBHs in centers of galaxies logically provide the most extreme environments, stellar-mass black holes are arguably the best laboratories to understand the physics of accretion and the interaction between radiation and matter in the strong gravity regime. Stellar-mass black holes in binary systems with stellar companions, also known as X-ray binaries, are the primary candidates for such studies. There are close to 30 of these systems confirmed in our Galaxy. They are abundantly bright in X-rays, and relatively nearby ( ∼ few kpc). Their much smaller size compared to SMHBs also means that they regularly display a rich phenomenology on human time scales; i.e., weeks to months, rather than years to megayears (Tanaka and Lewin, 1995; Remillard and McClintock, 2006). Thus, high signal-to-noise data, combined with accelerated evolution, makes black-hole binaries (BHBs) the premier sources to learn about black holes and their effects on space-time, matter and radiation.

Connors et al. (2024) present a study that illustrates the capabilities of HEX-P in measuring the evolving structures of accretion flows in BHBs observed at much lower luminosities than achieved with current facilities. Their simulations demonstrate that HEX-P can accurately measure the inner radius of the accretion disk via reflection spectroscopy at luminosities below 0.1% Eddington (Figure 12). For brighter sources, HEX-P’s observations will provide unprecedented data to probe the properties and indirect measurements of the geometry of the X-ray corona. ⁴ Detailed broadband studies of systems in nearby galaxies could be achieved for black holes accreting above the Eddington limit, which would expand the repertoire of sources available, possibly including those from different populations that have not yet been studied.

A fraction of accreting SMBHs launch powerful relativistic jets that can reach kpc, even Mpc scales. When the jets align with our line of sight, they appear boosted, producing bright, highly variable multi-band and multi-messenger emission. HEX-P is poised to answer many open questions about what powers these cosmic monsters. HEX-P’s ability to pinpoint fast variability (comparable to the system light-crossing time) simultaneously from 0.2 − 80   k e V will reveal particle acceleration mechanisms taking place in these jets, such as magnetic reconnection and shocks. The instrument’s broadband sensitivity will enable us to disentangle leptonic and hadronic emission mechanisms (Figure 10), potentially detecting proton-induced cascades associated with neutrino emission. Moreover, its sensitivity is paramount to detect a large fraction of the highest redshift ( z > 4 ) jets, essential to trace the evolution of jets and SMBHs into the early universe. As described by Marcotulli et al. (2023), HEX-P will also spatially resolve blazar and radio galaxy jets from core emission for the first time above 10 keV (Figure 13), thereby measuring jet spectral cut-offs and constraining the maximum energy of the particle population. Since blazars are strong, multi-messenger and time-variable sources, HEX-P blazar science has strong synergies with many facilities expected to be in operation in the early 2030s, including NASA’s IXPE (X-ray; Weisskopf et al., 2022), COSI ( γ -ray; Tomsick et al., 2023), and UVEX (ultraviolet; Kulkarni et al., 2021) missions, and the ground-based IceCube (neutrino; IceCube Collaboration, 2005) and CTAO ( γ -ray; Hofmann and Zanin, 2023) facilities.

With interior densities several times higher than atomic nuclei and size scales of 10 − 20 km, neutron stars are the densest objects that persist in Nature, and they exist on the verge of imploding into a black hole. Understanding how matter behaves under these extreme conditions is encapsulated in the EoS, relating the mass and radius of the neutron star (Tolman, 1939; Oppenheimer and Volkoff, 1939), and it is best determined through astronomical measurements (Lattimer, 2011).

Each method of determining the EoS of ultradense matter has underlying systematic uncertainties and probes different regions of the mass-radius parameter space (Figure 14). The firmest constraints require multiple approaches. Pulsar timing studies assume simplistic geometrical models (i.e., not physically motivated) of temperature across the neutron star surface (e.g.,; Miller et al., 2019), while gravitational wave studies provide tight radius constraints with poor mass constraints (Raaijmakers et al., 2021). Broadband X-ray studies covering 0.5 − 50 keV robustly measure continuum and reflection features, thereby providing the radius of the inner edge of the accretion disk and placing a strong, physical upper limit on the neutron star size (Ludlam et al., 2022). However, current measurements from coordinated NICER and NuSTAR observations require fixing the neutron star spin (typically assuming no rotation). The study by Ludlam et al. (2023) shows how the improved sensitivity of HEX-P will enable robustly solving for neutron star spin, thereby providing tighter radial constraints with less systematic bias.

With magnetic fields reaching typical strengths of ∼ 1 0 12 G in high-mass X-ray binary systems and ∼ 1 0 15 G in magnetars, accreting neutron stars provide a unique laboratory to study the extreme physics of accretion in the presence of strong magnetic fields. Neutron star magnetic field strengths are orders of magnitude higher than the strongest magnetic fields achieved in terrestrial laboratories ( ∼ 5 × 1 0 5 G). In high-mass binary systems, magnetic field pressure disrupts the inflow of the accreting material, channeling matter onto the magnetic poles. Since the neutron star rotates, radiation is released in a highly anisotropic manner, producing pulsing X-ray emission whose broadband study informs the plasma physics as well as the interaction of radiation with ultra-strong magnetic fields. Magnetars are also useful probes of the physics of the interaction of radiation with ultra-strong magnetic fields, but like some isolated neutron stars, that they are not thought to be accreting. The predictions for HEX-P’s unique capabilities in addressing important questions about the physics of magnetars and other isolated neutron stars are extensively discussed in Alford et al. (2024).

Due to the very high magnetic field present in magnetars, electron motion perpendicular to the magnetic field lines becomes quantized, leading to a resonance proportional to the magnetic field strength. This produces absorption-like cyclotron resonant scattering features, or simply cyclotron lines, at 5 − 100 keV. Cyclotron lines represent the most direct way to probe magnetic field strengths on or close to the neutron star surface, and they have been detected in about 40 sources (Staubert et al., 2019). Detecting fundamental lines and harmonics, and measuring their dependence on pulsar phase and luminosity, provides crucial insight into the accretion physics, including its geometry, electron bulk motion, turbulence, temperature distribution, and plasma conditions (Kreykenbohm et al., 2002).

Bachetti et al. (2023) presents a comprehensive study of potential HEX-P programs studying ultra-luminous X-ray sources (ULXs), at least some of which are neutron stars accreting at extreme rates, up to ∼ 500 times the Eddington limit (e.g., Bachetti et al., 2014; Fürst et al., 2016; Israel et al., 2017). Cyclotron lines provide the most direct measurement of ULX magnetic field strengths, which may be the key factor enabling their extreme luminosities. HEX-P will also be able to detect and study the radiatively-driven winds discovered in ULXs (Pinto et al., 2016), providing essential information on super-Eddington accretion and unveiling the origin of their surrounding 100-pc interstellar bubbles (Gúrpide et al., 2022).

HEX-P is an essential tool for studying SMBH growth. Merging galaxies offer one of the most dramatic channels for galaxy evolution, driving inflows of gas into galactic nuclei and potentially fueling both star formation and central SMBH activity. Dual AGN in late-stage mergers with nuclear pair separations < 10 kpc are ideal candidates to study SMBH growth since they coincide with the most transformative period for galaxies. However, since mergers are also associated with severe gas and dust obscuration (Koss et al., 2018; Blecha et al., 2018; De Rosa et al., 2019), dual AGNs are difficult to confirm and study, particularly with soft X-ray observatories. Hard X-rays offer a powerful tool to penetrate the dense obscuring media predicted to surround dual AGN in late-stage mergers (Figure 15). To date, only a handful of the brightest and closest systems have been studied at these energies due to the demanding instrument requirements involved. HEX-P enables, for the first time, the ability to spatially resolve hard X-ray counterparts of closely separated ( ∼ 2 − 5 ″) dual AGN.

The study reported by Pfeifle et al. (2023) shows via simulations that HEX-P will probe dual AGNs in late-stage mergers ( < 10 kpc) out to z ∼ 0.1 , and intermediate-stage mergers ( ∼ 30 kpc) out to z ∼ 0.3 . This is a dramatic improvement over NuSTAR and enables science that cannot be done with a soft X-ray observatory. HEX-P will transform our understanding of the role of AGN triggering and AGN feedback during galactic mergers.

The vast majority of accretion onto SMBHs takes place within a dense obscuring fortress of gas and dust. This circum-nuclear obscurer connects interstellar space to the accretion disc, and is known to evolve with both the accretion power of the central engine as well as the age of the Universe (e.g., Ueda et al., 2014; Buchner et al., 2014; Ananna et al., 2019). Thus a complete understanding of SMBH growth requires intricate knowledge of the structure and dynamics of the circum-nuclear environment surrounding AGN.

However, our current knowledge of the obscurer is fundamentally incomplete. Estimates of heavily obscured AGN space densities is unconstrained with observations typically suggesting that anywhere between ∼ 10%–50% of all SMBH activity takes place within material exceeding the Compton-thick limit (column density N H > 1 0 24 cm − 2 ; e.g., Ricci et al., 2015; Kammoun et al., 2020; Torres-Albà et al., 2021). Geometrical parameters describing the obscurer’s structure, such as covering factor, impose subtle changes on the observed hard X-ray spectrum of obscured AGN that NuSTAR can only begin to identify in a handful of brightest targets (e.g., Buchner et al., 2021). Furthermore, advances driven by NuSTAR have demonstrated that the obscurer is not static, but predominantly dynamic in nature (e.g., Baloković et al., 2021; Marchesi et al., 2022; Torres-Albà et al., 2023).

Boorman et al. (2023) present a series of simulations to show the exciting ways in which HEX-P will advance our understanding of the circum-nuclear environment surrounding AGN. Dramatic advances in sensitivity driven by reduced backgrounds and enhanced angular resolution will increase the available volume for robustly confirming heavily obscured accretion by more than an order of magnitude. By providing broadband X-ray spectral constraints on select AGN with precise black hole mass estimates, HEX-P will drive the development of new physically-motivated models that are capable of describing the geometrical structure of the obscurer in relation to the accretion power of the central engine. Also, the strictly simultaneous soft and hard X-ray spectral constraints provided by HEX-P will enable new insights into the dynamical nature of the obscurer via monitoring campaigns. Boorman et al. (2023) show that the passband of HEX-P is sufficient to disentangle intrinsic flux variations from structural changes of the obscurer to high precision, a task virtually impossible with current instrumentation and cadences.