Modified GEM Geometries: Detector Innovations

Updated 25 January 2026

Modified GEM geometries are systematic alterations to canonical GEM structures, designed to tune electric-field distribution, charge transport, and gain.
Design changes such as single-conical holes, varied copper thickness, and adjusted pitch significantly improve ion backflow suppression, spatial resolution, and mechanical durability.
Advanced simulation and empirical studies validate that these tailored modifications meet specific constraints for large-area tracking, timing systems, and parallax-free detection.

A modified GEM geometry is any systematic alteration to the canonical Gas Electron Multiplier (GEM) structure—usually a 50 μm polyimide foil, 5 μm copper claddings, and bi-conical holes, 70/50 μm OD/ID, in a 140 μm pitch hexagonal array—to engineer changes in electric-field distribution, charge transport, gain, robustness, stability, or system integration. Such modifications include altering hole shape and size, copper thicknesses, pitch, curvature (cylindrical, spherical), or readout geometry. They may be motivated by specific constraints of large-area tracking detectors, timing/triggering systems, or applications demanding minimized ion backflow or mechanical stress. Recent studies, using advanced simulation chains (ANSYS Maxwell/APDL, Garfield++), have produced both theoretical and empirical mappings from geometry to gain, uniformity, long-term stability, and ion feedback (Mondal et al., 2024, Rana et al., 18 Jan 2026, Pinto et al., 2010, Lavezzi et al., 2018, Mezzadri et al., 2018, Azmoun et al., 2015, Abbaneo et al., 2012, Flöthner et al., 2024, Gupta et al., 6 May 2025, Ogawa et al., 2017, Roy et al., 2021, Bouhali et al., 2021, Villa et al., 2010, Amoroso et al., 2018, Farinelli et al., 2018).

1. Standard and Canonical GEM Geometries

The canonical GEM geometry is characterized by a 50 μm Kapton film, 5 μm copper on both sides, and symmetric bi-conical holes (diameters: 70 μm at the surfaces, narrowing to 50 μm at the center), typically with a 140 μm pitch in a hexagonal lattice. Fabrication historically utilizes a double-mask photolithography process to produce highly symmetric holes, though single-mask techniques yield controlled asymmetries. The electric field focussing, Townsend coefficient α(E), and gain $G = \exp\left(\int_0^d \alpha[E(x)]dx\right)$ are tuned by layer geometry, voltages, and gas choice.

Table 1: Baseline GEM parameter summary

Parameter	Standard GEM
Substrate	50 μm Kapton
Copper cladding (both sides)	5 μm
Hole pattern (pitch)	140 μm hex.
Hole diameters (OD/ID)	70 μm / 50 μm (bi-conical)
Readout	Planar, multi-strip/pad

Modifications of these features underpin subsequent geometry variants (Mondal et al., 2024).

2. Modified Hole Geometries and Metallization

Several lines of research have investigated the impact of hole profile, diameter, and metallization on gas gain, charge transport, and ion backflow.

Single-Conical Holes and Copper Thickness Modification

Transitioning from a bi-conical (symmetric) hole to a single-conical (tapered only at one side, typically with the wider opening facing the lower, "exit" side) enhances performance:

Modified Geometry: Upper opening 50 μm, lower opening 70 μm, copper thickness variable at the lower electrode (5–20 μm) (Mondal et al., 2024, Rana et al., 18 Jan 2026).
Mechanism: The thicker lower copper layer increases the capacity for ion collection without substantially reducing the electric field if voltage across the GEM is adjusted to retain the target field inside the hole.
Governing Equations: The gain and ion backflow fraction,

$G = \exp\left(\int_0^d \alpha[E(x)]dx\right),\quad f_{ibf} = \frac{N_{ions,backflow}}{N_{e,anode}}$

are directly altered by field distribution, which is geometry-dependent.

Results:
- Gain increased by ≈50% for single-conical holes versus standard bi-conical at fixed voltage.
- Thickening lower copper to 10–15 μm optimally suppresses ion backflow: $f_{ibf}/G$ ratio reduced by up to 48% compared to the standard geometry, approximately halving the backflow per unit gain.
- Mechanical durability is increased, reducing risk of burn-through and handling damage.

Lower Cu (μm)	f_ibf/G Ratio Change (%)
Standard (5)	baseline
10	–46
15	–25 to –48 (field dependent)

Enhanced field uniformity and mechanical robustness are achieved without major process changes, making such modifications directly compatible with current photolithographic methods (Mondal et al., 2024, Rana et al., 18 Jan 2026).

3. Array Pitch, Hole Diameter, and Substrate Variations

Increasing the density of holes by reducing the array pitch or decreasing individual hole diameters is a second axis of modification.

Finer-Pitch Origination

Standard pitch: 140 μm; finer pitches: 90 μm, 60 μm (Flöthner et al., 2024, Gupta et al., 6 May 2025).
Hole diameters reduced in parallel: e.g., outer/inner to 55/40 μm or 30/25 μm.
Foil thickness may be reduced (from 50 μm to 25 μm) for very fine-pitch arrays, which increases sensitivity to mechanical stress.

Performance Impact:

Spatial resolution improves as pitch decreases: from ≃65 μm (standard) to ≃55 μm (fine pitch), and potentially to ≃50 μm with optimal electronics tuning (Flöthner et al., 2024, Gupta et al., 6 May 2025).
Gain is enhanced (up to 9× compared to standard for 60 μm pitch), but electron transparency can decrease due to increased field distortion and higher probability of electron loss at the hole rim.

Design Trade-offs:

Best ultimate spatial and gain performance (FTGEM, p = 60 μm) is offset by lower transparency, increased capacitance, and higher demands on mechanical robustness.
A compromise pitch (90 μm) may optimize the trade between gain, transparency, and manufacturability (Gupta et al., 6 May 2025).

Thick-GEMs:

Doubling Kapton thickness to 100 μm enables high single-stage gain but introduces large gain inhomogeneities (>50%) unless the hole diameter is also enlarged (to ≃100 μm), at which point gain variation is reduced to ~2–3% across large areas, matching standard-foil uniformity (Ogawa et al., 2017).

4. Non-Planar Topologies: Cylindrical and Spherical GEMs

Cylindrical GEMs

Development of concentric, self-supporting triple-GEM detectors (CGEM) introduces new engineering and physical considerations (Lavezzi et al., 2018, Mezzadri et al., 2018, Farinelli et al., 2018, Amoroso et al., 2018). Standard foils (50 μm Kapton, 70/50 μm holes, 140 μm pitch) are bent and supported by Rohacell or similar foams, with permaglass end-rings to retain precise, uniform radii and minimize material budget (typ. ≤1.5% X₀ per layer).

Benefits:
- Radially symmetric drift field preserves uniform Lorentz angle in a magnetic field, mitigating reconstruction artifacts.
- Mechanical design (foam supports, jagged readouts) reduces X₀ and inter-strip capacitance, respectively, increasing S/N.
- Test beams confirm that spatial resolution, efficiency, and gain uniformity match or surpass the best planar prototypes; σ_x ≃ 100–130 μm at B = 1 T is achievable.

Spherical GEMs

Spherical forming of GEM foils, using heat-stretching over optical mandrels in vacuum, produces nearly-perfect spherical segments up to diameters ≃17 cm without deforming the fine hole array (Pinto et al., 2010).

Motivation and Benefit:
- Radial drift field eliminates the parallax error, particularly crucial for wide-angle photon detection in x-ray, neutron, or UV conversions.
- Field uniformity is preserved up to θ ≈ 70°, far beyond the planar geometry's effective range.
- The mechanical process preserves hole-to-hole uniformity; electron-gain curves are statistically indistinguishable from planar foils.
Technical Challenges:
- Process is time-intensive (≥24 h per foil), requires deep vacuum, heat stability, and careful mechanical control.
- Implementation of spherical readouts and HV connectivity remains technologically demanding.

5. Readout Geometry Innovations and Advanced Segmentation

Advanced segmentation of anode readouts applies both in planar/cylindrical and modified GEMs.

Jagged/zig-zag strip anodes: Introduced (BESIII/CMS/RD51) to reduce inter-strip capacitance by ≈30%. The net effect is reduced electronic noise and better cluster centroid precision (Lavezzi et al., 2018, Mezzadri et al., 2018, Abbaneo et al., 2012).
Chevron and time-projection readouts: Pads with internal zig-zag patterns (0.5 mm pitch) or time-projection (μ-TPC) modes utilize both spatial and time-charge correlations to reconstruct short track vectors, improving σ_x at large track incident angles (σ_x ≲ 150–400 μm for θ ≤ 45°) (Azmoun et al., 2015, Lavezzi et al., 2018).
Barycenter and charge centroid algorithms: Multiple reconstruction algorithms can be deployed (charge centroid, μ-TPC, hybrid weighting) to adapt spatial resolution to angle, drift field, and magnetic field conditions (Lavezzi et al., 2018, Azmoun et al., 2015).

6. Hole Asymmetry and Orientation

Production via single-mask photolithography unavoidably generates hole asymmetry (one face larger than the other), the orientation of which materially changes extraction efficiency, gain, and loss channels (Bouhali et al., 2021, Roy et al., 2021, Villa et al., 2010).

Orientation “B” (larger opening toward readout) is always superior; it offers ~50% greater gain in both single and triple-GEM configurations, principally due to better extraction efficiency at the final GEM stage.
Design implication: Mandate orientation of all asymmetric holes with the wider diameter toward the electron collection side in mixed (single/double-mask) systems to maximize performance (Bouhali et al., 2021).

7. Simulation Frameworks and Quantitative Optimization

All recent work emphasizes parameter optimization and quantitative cross-validation between finite-element electrostatic models (ANSYS Maxwell/APDL), particle tracking (Garfield++ with Magboltz), and experimental data. Optimization targets practical handles: gain (G), gain uniformity (ΔG/G), ion backflow fraction (f_ibf), charge sharing coefficients (f_CH), collection/extraction efficiencies (η_C, η_E), transparency (T), and durability under HV and rate stress.

Summary Table: Key Modified GEM Geometries

Modification	Typical Metric Improvement	Citation
Single-conical + thickened Cu	+50% gain, –48% IBF/G, durable	(Mondal et al., 2024, Rana et al., 18 Jan 2026)
Pitch reduction (90 μm)	10–15 μm better σ_x, higher G	(Flöthner et al., 2024, Gupta et al., 6 May 2025)
Spherical forming	Parallax-free; uniform σ_x to θ≃70°	(Pinto et al., 2010)
Cylindrical (CGEM-IT)	σ_x≲130 μm @ 1T, X₀≲1.5%	(Lavezzi et al., 2018, Mezzadri et al., 2018, Farinelli et al., 2018, Amoroso et al., 2018)
Jagged strips (zig-zag)	–30% capacitance, <100 μm σ_x	(Lavezzi et al., 2018, Abbaneo et al., 2012)
Optimal thick GEM	High G, ΔG/G < 2–3% (φ=100 μm)	(Ogawa et al., 2017)
Asymmetric orientation (B)	+50% gain, fewer losses	(Bouhali et al., 2021)

Outlook

Current research demonstrates that modified GEM geometries—spanning micro-structural changes to global shaping of the amplification foil—enable systematic improvement in gain, spatial resolution, ion backflow suppression, and mechanical integrity without fundamentally altering manufacturability. Ongoing studies aim for further reductions in material budget, finer spatial sampling, and more robust, parallax-free architectures, with parameter spaces now precisely mapped by combined simulation and empirical validation (Mondal et al., 2024, Rana et al., 18 Jan 2026, Gupta et al., 6 May 2025, Pinto et al., 2010, Lavezzi et al., 2018, Mezzadri et al., 2018, Azmoun et al., 2015, Amoroso et al., 2018, Roy et al., 2021, Bouhali et al., 2021, Ogawa et al., 2017, Flöthner et al., 2024, Abbaneo et al., 2012, Farinelli et al., 2018, Villa et al., 2010).