Aberration-Corrected Bright-Field Imaging (acBF)

• Aberration-corrected bright-field imaging (acBF) is a 4D-STEM method that compensates aberration-induced phase shifts to enhance image fidelity.
• It works by recombining tilt-corrected bright-field (tcBF) and differential phase contrast (tcDPC) channels to recover complete coherent phase information.
• acBF minimizes contrast reversals and zero crossings, offering robust, non-iterative phase reconstruction ideal for low-dose and weak phase object imaging.

Searching arXiv for papers on aberration-corrected bright-field imaging and related bright-field aberration-correction work. Aberration-corrected bright-field imaging (acBF) denotes a family of bright-field microscopy regimes in which aberration-induced phase or blur is compensated so that specimen information is transferred more faithfully into the recorded or reconstructed image. In current electron-microscopy usage, the term is most explicitly defined in 4D-STEM as a direct, non-iterative phase-imaging mode that combines tilt-corrected bright field (tcBF) and tilt-corrected differential phase contrast (tcDPC) so as to use the full coherent phase information available inside the bright-field disk under the weak phase object approximation (WPOA) (Ma et al., 28 Jul 2025). In a broader historical sense, closely related bright-field regimes include Cs-corrected TEM with a physical phase plate (Gamm et al., 2010), aberration-corrected phase-contrast HRTEM operated near $C_s \approx 0$ or in the negative-$C_s$ regime (Datta et al., 25 Mar 2026), and quantitative position-averaged incoherent bright-field STEM used as a thickness gauge after absolute normalization and unit-cell averaging (Xin et al., 2011).

1. Definition, scope, and terminology

In the narrowest and now most specific sense, acBF is not a new detector geometry but a way of using the same 4D-STEM data more completely. A 4D-STEM experiment records a convergent-beam electron diffraction pattern at every probe position, and acBF is obtained by separating and recombining complementary symmetry channels inside the bright-field disk rather than by changing the detector hardware itself (Ma et al., 28 Jul 2025).

This usage should be distinguished from broader bright-field language in aberration-corrected microscopy. In Cs-corrected HRTEM, bright-field imaging is often described through its phase-contrast transfer function (PCTF), with aberration correction allowing operation close to $C_s=0$ or at deliberately chosen small positive or negative $C_s$ so that atomic columns can appear as dark or bright dots (Datta et al., 25 Mar 2026). In an earlier TEM phase-plate analysis, a physical $\pi/2$ phase plate in the back focal plane of a Cs-corrected microscope was identified as the practical basis for an in-focus, aberration-corrected bright-field regime with strong low- and medium-frequency phase contrast (Gamm et al., 2010). In aberration-corrected STEM, bright-field acquisition was also made quantitative in a different sense by the development of quantitative position-averaged incoherent bright-field (QPA-IBF), where the averaged bright-field signal becomes a monotonic function of thickness (Xin et al., 2011).

A common misconception is that all of these uses are identical. The literature instead supports a layered terminology: a specific 4D-STEM reconstruction mode called acBF; a broader aberration-corrected phase-contrast bright-field regime in TEM/HRTEM; and quantitative bright-field STEM measurements that exploit aberration-corrected instruments without necessarily implementing the 4D-STEM acBF formalism.

2. Information channels inside the bright-field disk

The 4D-STEM formulation of acBF is built on the statement that the phase information in the bright-field disk is encoded in different detector-space symmetry channels. The elastic scattering amplitude is decomposed as

$$F(\mathbf k',\mathbf k)=F_s(\mathbf k',\mathbf k)+iF_a(\mathbf k',\mathbf k),$$

with

$F_s=\frac{F+F^*}{2},\qquad F_a=\frac{F-F^*}{2i}.$

Within Rose’s generalized contrast formalism, these two components contribute differently to image formation: symmetric scattering components encode tcBF, antisymmetric scattering components encode tcDPC, and the two are complementary under WPOA (Ma et al., 28 Jul 2025).

Under WPOA, the quadratic terms are neglected and the 4D-STEM phase contrast transfer is written as a doubly reciprocal-space PCTF,

$$\mathrm{C}_{\mathrm{WPOA}}(\mathbf w,\boldsymbol\Theta) =\mathrm{PCTF}(\mathbf w,\boldsymbol\Theta)\,F_s(\mathbf w).$$

The paper further organizes the detector-space information using reciprocal-space overlap regions. In the triple overlap (TO) region, all three aperture terms contribute; in the double overlap (DO) region, only two contribute. This overlap bookkeeping identifies where coherent phase information resides in the 4D-STEM dataset and clarifies why tcBF and tcDPC are complementary rather than redundant (Ma et al., 28 Jul 2025).

This framework makes acBF conceptually close to direct ptychographic imaging. The difference is methodological rather than informational: acBF is analytically tractable and non-iterative, whereas full ptychography reconstructs phase by iterative use of the same underlying 4D-STEM data.

3. Tilt correction, tcBF, tcDPC, and the acBF synthesis

For defocused STEM, reciprocity implies that each detector pixel corresponds to a slightly tilted-beam TEM image. Defocus therefore produces a detector-dependent real-space shift,

$\Delta \mathbf p = \Delta f\,\boldsymbol\Theta.$

Tilt correction removes this shift before summation, which is the key operation underlying tcBF, tcDPC, and acBF (Ma et al., 28 Jul 2025).

The three modes can be summarized as follows:

Mode	Detector-space operation	Stated transfer behavior
tcBF	Symmetric sum of shift-corrected bright-field pixels	CTEM-like sine-modulated phase transfer
tcDPC	Antisymmetric subtraction of Friedel pairs after shift correction	Purely imaginary cosine-modulated transfer
acBF	Combination of tcBF and tcDPC	Continuous nonzero transfer from $0$ to $C_s$0 under WPOA

For pure defocus, tcBF is obtained by symmetrically summing the tilt-corrected bright-field disk, yielding

$C_s$1

tcDPC is obtained by subtracting opposite detector positions after shift correction, yielding

$C_s$2

Because tcBF carries the sine term and tcDPC carries the cosine term, the zeroes of one tend to coincide with strong transfer in the other. acBF combines them so that all usable bright-field information adds constructively, with transfer

$C_s$3

This is why acBF is described as the maximal phase-contrast transfer achievable within the bright-field disk under WPOA and as a direct, analytic ptychographic-style combination of the two complementary channels (Ma et al., 28 Jul 2025, Ma et al., 1 Oct 2025).

The practical implication is that acBF is not merely a prettier tcBF image. It reconstructs the real or symmetric contribution from tcBF and the imaginary or antisymmetric contribution from tcDPC, thereby avoiding the zero crossings and contrast reversals that limit tcBF alone.

4. Aberrations, Scherzer conditions, and the phase-plate analogy

The name “aberration-corrected” in acBF refers to the removal of aberration-induced phase or shift modulation before summation. For pure defocus, the correction is exactly the Fourier shift factor associated with $C_s$4. For higher-order aberrations, the situation is more complex: off-axis virtual images are no longer related by a simple rigid translation, and tcBF can become strongly damped because different detector pixels carry different effective defocus and distortion terms (Ma et al., 1 Oct 2025).

The 2025 analysis of acBF in the presence of higher-order aberrations makes this distinction explicit. For aberrations beyond defocus, the image shift is only approximately

$C_s$5

and additional distortions remain. In particular, spherical aberration causes detector-dependent variations that tcBF does not fully correct, whereas acBF corrects each virtual image’s transfer before summation. The paper therefore concludes that higher-order round aberrations such as spherical aberration can be largely compensated, but non-round aberrations such as two-fold astigmatism can produce anisotropic information loss that neither tcBF nor acBF can recover if the information is not present in the recorded data (Ma et al., 1 Oct 2025).

A second important theme is the phase-plate analogy. In conventional Cs-corrected TEM, the wave aberration without a phase plate is

$C_s$6

and the coherent bright-field PCTF is sine-type. Adding a physical phase plate with $C_s$7 changes the transfer as

$C_s$8

thereby converting the usual sine-type transfer into a cosine-type transfer with strong low- and medium-frequency contrast (Gamm et al., 2010). In 4D-STEM acBF, an analogous effect arises computationally because tcBF and tcDPC together supply the two orthogonal transfer channels. The 2025 acBF paper makes the analogy even stronger by noting that, at Scherzer defocus in a spherically aberration-limited system, the phase shift from the probe-forming lens acts like a phase plate and removes oscillations from the acBF CTF (Ma et al., 1 Oct 2025).

This connects acBF to the broader aberration-corrected HRTEM tradition. The 2026 review of aberration-corrected phase-contrast TEM frames negative-$C_s$9 imaging and related bright-field modes as direct-image interpretation regimes in which aberration correction reshapes the PCTF, reduces delocalization, and permits atomic columns to appear as either bright or dark dots depending on the sign of the transfer function (Datta et al., 25 Mar 2026).

5. Historical and broader bright-field realizations

Before the specific 4D-STEM definition of acBF, aberration-corrected bright-field imaging already had several concrete realizations in electron microscopy. One of the clearest STEM examples is QPA-IBF, introduced as a method for determining the on-axis thickness of crystalline specimens. In that method, the bright-field detector semi-angle is chosen large enough—roughly $C_s=0$0–$C_s=0$1 the probe convergence semi-angle and larger than the relevant Bragg disks—so that the signal becomes approximately incoherent. The bright-field signal is then averaged over one unit cell and normalized by the incident beam intensity; in the thin-specimen limit this cancels phase factors from defocus and aberrations, and the resulting QPA-IBF becomes a monotonic function of thickness that can be matched to multislice frozen-phonon simulations (Xin et al., 2011).

Experimentally, QPA-IBF was demonstrated on DyScO$C_s=0$2 along the $C_s=0$3 zone axis in an aberration-corrected 100 keV Nion UltraSTEM with a convergence semi-angle of about $C_s=0$4 mrad. The reported standard deviation of the residuals was $C_s=0$5 nm, quoted as a precision of about $C_s=0$6 nm. The method measured both the crystalline portion and the non-crystalline surface layers and avoided systematic errors in EELS $C_s=0$7 thickness determination caused by surface plasmons (Xin et al., 2011).

In TEM, the phase-plate route provided a different bright-field realization. A Cs-corrected microscope with a physical $C_s=0$8 phase plate was predicted to improve strongly low- and medium-resolution object contrast while also improving tolerance to defocus and $C_s=0$9-variations relative to a microscope without a phase plate. The idealized condition was $C_s$0, $C_s$1, and $C_s$2, so that the PCTF nearly matches the partial-coherence envelope over the useful range (Gamm et al., 2010).

The later HRTEM review generalized this perspective by treating aberration-corrected bright-field imaging as a tuned phase-contrast mode of HRTEM, especially in the negative-$C_s$3 regime. In that literature, acBF overlaps with direct atomic-resolution imaging of materials such as ZnO, SrTiO$C_s$4, YBa$C_s$5Cu$C_s$6O$C_s$7, BN monolayers, and MoS$C_s$8, where aberration correction expands the usable imaging space and enables direct visualization of atomic columns, vacancies, interfaces, and ferroelectric displacements (Datta et al., 25 Mar 2026).

These historical strands show that the modern 4D-STEM definition of acBF did not emerge in isolation. It formalized, within 4D-STEM, a long-standing bright-field objective: reshape or compensate aberration-driven transfer so that phase information is retained more completely and with less delocalization.

6. Relation to neighboring methods, limitations, and future directions

acBF is closely related to, but distinct from, several neighboring phase-imaging methods. Relative to tcBF, its defining advance is that it does not discard the anti-symmetric scattering channel. Relative to tcDPC, it restores the symmetric low-frequency channel that antisymmetric subtraction suppresses. Relative to iterative ptychography, it is a fast, non-iterative approximation that exploits the same 4D-STEM data and is especially attractive for weak phase objects and low-dose imaging (Ma et al., 28 Jul 2025, Ma et al., 1 Oct 2025).

It is equally important to state what acBF does not do. It does not abolish the weak-phase assumptions underlying the principal derivation, and it does not recover information that has been directionally suppressed by non-round aberrations. Beyond WPOA, the 4D-STEM formalism contains coherent amplitude transfer and incoherent quadratic terms arising from elastic plus inelastic or plural scattering interference. Those terms can be useful—for example through tilt-corrected dark-field imaging—but they are not equivalent to the linear acBF phase channel (Ma et al., 28 Jul 2025).

A broader misconception is that acBF is simply another name for all aberration-corrected bright-field reconstruction. The literature instead suggests a family resemblance across modalities. APIC, for example, is not the same as acBF, but it includes an aberration-corrected bright-field reconstruction stage based on NA-matching measurements and analytical aberration recovery before extending to dark-field spectrum synthesis (Cao et al., 2023). In optical microscopy, related but nonidentical uses include calibration-free aberration compensation in quantitative phase imaging by a U-net-based model (Chang et al., 2020), spectroscopic whole-system calibration of bright-field transmission microscopy (Platonova et al., 2019), model-based correction inside a glass tube using bright-field-derived geometry (Cox et al., 2023), and localized aberration correction with deformable phase plates in “Fovea Stacking” (Mao et al., 31 May 2025).

On the hardware side, recent electron-optical proposals suggest that future acBF-like operation may not rely exclusively on conventional multipole correctors. A laser-based aberration corrector using the ponderomotive electron–photon interaction was proposed as a free-space, light-driven module that could compensate spherical aberration at high voltage and chromatic aberration at low voltage, with simulated probe-size reductions of a factor of six in a 200 kV example and a factor of three in a 500 V example (Nekula et al., 27 Jan 2025). A related experimental demonstration used a shaped light field as an optical field electron modulator to compensate the positive spherical aberration of a round lens, reducing measured $C_s$9 from about $\pi/2$0 m to approximately $\pi/2$1 to $\pi/2$2 m and thereby providing a pathway toward compact and tunable light-based aberration correctors for high-resolution electron microscopy (Mihaila et al., 25 Apr 2025).

Taken together, these developments suggest a stable core meaning and a widening technical perimeter. The stable core is bright-field imaging in which aberration-driven transfer is explicitly corrected so that phase information is used more completely. In 4D-STEM, that core is realized analytically as acBF by combining tcBF and tcDPC. In broader microscopy, the same objective appears in phase-plate TEM, negative-$\pi/2$3 HRTEM, quantitative incoherent bright-field STEM, and emerging optical or ponderomotive correction architectures.