The short answer is photons knock electrons loose, and a built-in electric field pushes them around an external circuit. The long answer involves quantum mechanics, semiconductor band theory, and a few specific numbers that explain why silicon dominates the market and why panel efficiency is capped at 33.7% for single-junction designs. This is the under-the-hood version of solar conversion, what's actually happening at the atomic level when sunlight hits your roof. We covered the basics in our how solar panels work guide. Here we go deeper, explained step by step.
TL;DR: Photons with energy above silicon's 1.12 eV bandgap (wavelengths shorter than 1,100 nm) excite electrons from the valence band to the conduction band, creating electron-hole pairs in the silicon lattice. The P-N junction has a built-in electric field of about 0.6-0.7V across a depletion zone roughly 0.5-2 micrometers wide. This field separates the electron-hole pairs before they recombine: electrons drift to the N-side and holes to the P-side, generating DC current through any external circuit. A typical cell produces 0.55-0.65V at maximum power point and 9-11A short-circuit current under STC (1,000 W/m2 irradiance, 25 deg C). Stringing 60-72 cells in series gives a panel ~36-42V Vmp and 8-11A Imp, around 400W rated output. The inverter's Maximum Power Point Tracker (MPPT) continuously adjusts operating voltage to land on the V-I curve's power peak as temperature and irradiance change. Shockley-Queisser detailed balance analysis caps single-junction silicon at 33.7% theoretical efficiency; production panels hit 22-23% as of 2026. For the broader process from panel to outlet, see our how solar panels work guide.
When I started learning photovoltaics in detail I expected the physics to be the hard part. It turned out to be the easy part. The hard part was understanding how dozens of secondary effects (surface recombination, contact resistance, spectral mismatch, temperature coefficients) drag the theoretical 33.7% efficiency down to 22-23% in real production. Each of those losses has a specific cause and a specific engineering response, and the panel industry has spent 70 years chiseling them away one at a time.
What Energy Does a Photon Need to Generate Electricity?
For silicon, a photon needs energy above 1.12 eV (the silicon bandgap) to excite an electron from the valence band into the conduction band. Photons below that threshold pass through the silicon without absorbing, they're useless for power generation. Photons above the threshold are absorbed, but any excess energy beyond 1.12 eV converts to heat rather than electricity (a phenomenon called thermalization).
The 1.12 eV bandgap corresponds to a wavelength of about 1,100 nanometres at the boundary of visible light and near-infrared. Specifically:
- UV light (200-400 nm, around 3-6 eV): absorbed, but loses most energy to thermalization
- Visible light (400-700 nm, around 1.8-3.1 eV): absorbed efficiently, with moderate thermalization loss
- Near-infrared (700-1,100 nm, around 1.1-1.8 eV): absorbed well, with minimal thermalization
- Below 1,100 nm (mid-infrared and beyond): passes through silicon without absorbing
This wavelength coverage is why silicon was chosen as the dominant photovoltaic material. The solar spectrum at Earth's surface (AM 1.5G reference) peaks in the visible band around 500 nm, with significant intensity through the near-infrared to about 2,500 nm. Silicon captures the high-intensity portion of the spectrum reasonably well but misses the long-wavelength infrared.
The Shockley-Queisser detailed balance analysis (1961) calculates that the optimal single-junction semiconductor for terrestrial solar conversion has a bandgap around 1.34 eV, slightly above silicon's 1.12 eV. Gallium arsenide (1.42 eV) sits closer to the optimum and reaches 27% theoretical limit versus silicon's 33.7% (different limits because of different absorption profiles). The fact that silicon won the market anyway reflects manufacturing economics, silicon is abundant, well-understood, and infinitely scalable, while GaAs is rarer and harder to grow defect-free.
How Does the P-N Junction Separate Electron-Hole Pairs?
The P-N junction in a silicon cell is the boundary between two regions doped with different impurities, phosphorus for N-type (extra electrons) and boron for P-type (electron holes). The doping creates a permanent electric field that's the entire mechanism for converting absorbed photons into directional current.
What happens at the junction physically:
- Electrons from the N-side diffuse into the P-side (random thermal motion in both directions, net flow from high to low concentration)
- Holes from the P-side diffuse into the N-side (same mechanism)
- This diffusion leaves behind fixed ionized dopant atoms: positive phosphorus ions on the N-side, negative boron ions on the P-side
- The exposed ionized dopants create an electric field pointing from N-side to P-side, the "built-in field" of about 0.6-0.7V across roughly 0.5-2 micrometers
The region around the junction where the ionized dopants dominate (after diffused electrons and holes have moved away) is called the depletion zone. It's typically 0.5-2 micrometers thick depending on doping levels. Outside the depletion zone, the silicon is mostly electrically neutral.
When a photon is absorbed inside or near the depletion zone, the electron-hole pair gets separated by the field. The electron is accelerated toward the N-side, the hole toward the P-side. Outside the depletion zone, electron-hole pairs can recombine before reaching the field, which is why the depletion zone and its immediate vicinity (the minority carrier diffusion length) determine cell performance.
For mono-crystalline silicon, the minority carrier diffusion length runs around 100-300 micrometers, much longer than typical cell thickness (180 micrometers). This means most photons absorbed anywhere in the cell can generate useful current. For polycrystalline silicon, grain boundaries act as recombination sites, shortening effective diffusion length to 50-150 micrometers and reducing efficiency.
The honest framing: the depletion zone is where the photovoltaic effect happens. Everything else about cell design (texturing, anti-reflective coating, contact metallization) is about delivering photons to the depletion zone efficiently and getting electrons out without resistive losses.
What Does the V-I Curve Tell You About a Cell?
The V-I curve plots voltage versus current for a solar cell under specific illumination. It's the fundamental characterization of cell performance and the basis for everything an inverter does to optimize output. A typical 400W production panel under STC (1,000 W/m2, 25 deg C cell temperature) gives a V-I curve with these key points:
| Parameter | Symbol | Typical value | What it means |
|---|---|---|---|
| Short-circuit current | Isc | 10-11A | Current when output is shorted, max possible current |
| Open-circuit voltage | Voc | 41-45V (60-cell panel) | Voltage with no load, max possible voltage |
| Maximum power voltage | Vmp | 33-38V | Voltage at maximum power point |
| Maximum power current | Imp | 9-10A | Current at maximum power point |
| Fill factor | FF | 0.78-0.84 | Imp x Vmp / (Isc x Voc), curve "squareness" |
The curve shape: starts at Isc on the current axis (when voltage is zero), stays roughly flat at near-maximum current as voltage rises, then drops sharply as voltage approaches Voc. The maximum power point sits at the "knee" of the curve where the product of V and I is largest.
A panel's nameplate "400W" rating means 400 W at the maximum power point under STC. The product Vmp x Imp = Pmp, in this example roughly 36V x 11A = 396W. The fill factor for a high-quality panel is around 0.80, meaning the maximum power point captures 80% of the theoretical Isc x Voc rectangle.
Temperature and irradiance both shift the V-I curve. Higher temperature primarily reduces Voc (by about 2 mV per cell per degree C above 25), which is why panels lose ~0.30%/deg C in TOPCon cells and ~0.35%/deg C in PERC. Lower irradiance reduces Isc proportionally (a panel under 500 W/m2 gives roughly half the Isc of one under 1,000 W/m2) while leaving Voc nearly unchanged.
For a deeper comparison of how different cell technologies shift the V-I curve, see our TOPCon vs HJT vs PERC comparison.
How Does MPPT Find and Hold the Maximum Power Point?
Maximum Power Point Tracking is the inverter's algorithm for landing the panel string at its V-I curve's power peak. The peak shifts continuously as irradiance and temperature change, so a static operating point would lose substantial energy. Modern inverters sample the V-I curve hundreds of times per second and adjust the operating voltage in real time.
The most common MPPT algorithm is Perturb and Observe (P&O):
- Measure current operating voltage (V) and current (I), compute power P = V x I
- Perturb voltage slightly (typically by 1-3%)
- Measure new power
- If new power is higher, continue perturbing in the same direction. If lower, reverse direction
- Repeat continuously
The algorithm naturally oscillates around the maximum power point, never settling exactly on it but staying within 1-2% of true peak under stable conditions. Under rapidly changing conditions (passing clouds, sunrise/sunset), tracking accuracy can drop to 95-97% of available power.
More sophisticated algorithms (Incremental Conductance, particle swarm, neural networks) trade complexity for tracking accuracy. The high-end SolarEdge HD-Wave inverters and Enphase IQ8 microinverters use proprietary tracking algorithms that claim 99%+ accuracy under fast-changing conditions.
Multiple MPPT channels matter for systems with mixed orientations or shading. A typical residential string inverter has 2-3 MPPT channels, each tracking a separate string independently. This means an east-facing string and a west-facing string can each operate at their own optimum without one dragging down the other.
Module-level power electronics take this further. The SolarEdge P370 optimizer puts a tiny MPPT controller on each panel, decoupling its operating point from the string. The Tigo TS4-A-O does the same with retrofit compatibility for any string inverter. The Enphase IQ8A microinverter converts DC to AC at each panel, so each panel runs its own MPPT independently.
Is MLPE worth the cost on every install? Honestly, no. For an unshaded south-facing roof with uniform orientation, string-level MPPT captures essentially all available power. MLPE pays back where shading, multiple orientations, or panel mismatch exist, which is most US residential installs but not all.
What Limits Real-World Cell Efficiency?
The Shockley-Queisser detailed balance limit caps single-junction silicon at 33.7% theoretical efficiency under STC. Production silicon panels currently hit 22-23%. The gap is around 10 percentage points, and each percentage represents a specific loss mechanism that engineers are still working to reduce.
The major loss mechanisms:
Below-bandgap photons (15-19% loss). Photons below 1.12 eV pass through silicon without absorbing. No engineering fix at the cell level, requires multi-junction or up-conversion to capture this energy.
Thermalization (33-40% loss). Photons above 1.12 eV transfer their excess energy to heat. The shorter the wavelength, the more heat. Engineering responses include hot-carrier cells (research) and multi-junction designs that capture different wavelengths in different layers.
Recombination losses (1-3% loss). Some electron-hole pairs recombine before reaching the depletion zone, particularly at grain boundaries, defects, and unpassivated surfaces. TOPCon's tunnel oxide layer and HJT's amorphous silicon passivation both target this loss.
Optical losses (3-5% loss). Reflection at the air-glass interface, encapsulant absorption, busbar shading, and absorption in non-active layers. Anti-reflective coatings, textured surfaces, and busbar optimization all help.
Resistive losses (1-2% loss). Current flowing through finite-resistance metallization and bulk silicon. Higher busbar counts and thinner cells reduce this, at the cost of more shading area.
Temperature derating (5-15% in operation). Real panels rarely operate at 25 deg C, summer roof temperatures run 50-70 deg C cell temperature, reducing output through the temperature coefficient. Not a "cell efficiency" loss strictly, but a real annual energy loss.
The current commercial cell efficiency record is around 27.3% (Longi 2024 HBC structure). Module efficiency (including spacing between cells and frame area) runs 22-23% for current TOPCon and HJT volume products. The remaining gap to the 33.7% Shockley-Queisser limit is being chipped away annually, but the easy gains are gone, the next 3-4 percentage points will be harder than the last 8.
Citation capsule: Solar panels generate electricity through the photovoltaic effect: photons with energy above silicon's 1.12 eV bandgap excite electrons from the valence band to the conduction band, creating electron-hole pairs that the P-N junction's 0.6-0.7V built-in field separates as DC current (NREL Photovoltaic Cell Physics). The Shockley-Queisser detailed balance limit caps single-junction silicon at 33.7% theoretical efficiency; modern TOPCon and HJT production panels achieve 22-23% module efficiency, with the gap closing slowly through better passivation, anti-reflective coating, and contact metallization.
Summary
Solar panels make electricity by absorbing photons above silicon's 1.12 eV bandgap, separating the resulting electron-hole pairs across a P-N junction's built-in 0.6-0.7V field, and collecting the DC current through metallization on the front and back of the cell. A typical 60-cell panel produces ~36V Vmp at 11A Imp, around 400W under STC conditions. The inverter's MPPT continuously adjusts operating voltage to land on the V-I curve's power peak as conditions change, typical efficiency 96-98%. Real production cell efficiency runs 22-23% versus the 33.7% Shockley-Queisser theoretical limit, the gap divided among thermalization, below-bandgap photons, recombination, optical losses, resistive losses, and temperature derating in operation. For the broader system-level picture from panel through inverter to grid, see our how solar panels work guide. For the materials and construction that bring those cells together into a 25-year roof-mounted assembly, our piece on what solar panels are made of covers the build details.