Harvesting energy from sun, outer space, and soil

The study addresses the challenge of continuous renewable electricity generation without energy storage by using naturally occurring temperature differences between the Sun (~5800 K), soil near Earth's surface (~290 K), and outer space (~3 K). Drawing from thermodynamic principles, the authors design a heat engine that uses solar heating during the day (hot absorber to cold soil) and radiative cooling at night (hot soil to cold emitter facing outer space) to drive a thermoelectric generator (TEG). They formulate energy balance expressions for net heating and cooling that include radiative, non-radiative (convection), and conductive terms with the soil, and argue that soil provides a more stable thermal reservoir than ambient air (colder than air by day, warmer than air at night). The purpose is to realize 24-h electricity generation using low-cost materials and no batteries, relevant for off-grid rural electrification and sensing.

The paper reviews selective solar absorbers (photonic crystals, metamaterials, cermets) achieving high solar absorptivity (90–98%) and low mid-IR emissivity (3–10%), and radiative coolers (photonic structures, metamaterials, polymer nanofibers/aerogels) requiring unity emissivity within the 8–13 µm atmospheric window for sub-ambient cooling. It notes the practical difficulty of achieving reversible day/night emissivity (e.g., with VO2 phase-change materials), motivating an all-black, polymer-based paint solution with near-unity absorptivity across solar and mid-IR bands as a cost-effective compromise. The authors also reference broad applications and advantages of TEGs in vehicles, wearables, and waste-heat recovery, positioning TEGs as a reliable means to convert the day/night temperature differentials into electricity. Prior art on radiative-cooling-assisted power generation is cited, with the present work aiming to exceed previous nighttime power densities and to demonstrate continuous 24-h operation.

Device design and materials: A low-cost (<$15) TEG-based system was built comprising: (i) a dual-purpose black absorber/emitter made from a 70 mm × 70 mm × 0.8 mm copper sheet coated with ~0.01–0.1 mm thick commercial Black 3.0 paint (ε ≈ 0.98, αsolar ≈ 0.98), (ii) a commercial TEG module (SP1848-27145, 40 mm × 40 mm × 3.4 mm), and (iii) an aluminum heat sink (≈37.6 × 36.6 × 23.6 mm) inserted into soil. Thermal compound paste (ARCTIC MX-4, 8.5 W m−1 K−1) was used on both TEG interfaces; an aluminum foil barrier prevented condensation and blocked stray radiation. The assembly was placed in a 25 mm thick polystyrene insulation foam box lined with aluminized Mylar to reduce radiative losses and covered by a visible- and IR-transparent LDPE film (12.7 µm) to limit convection while allowing solar input and mid-IR emission.

Optical characterization: Spectral hemispherical emissivity/absorptivity were measured. UV–Vis–NIR (200–2500 nm) reflectance with a Jasco V770 integrating sphere (ISN-923) referenced to PTFE; mid-IR (2.5–20 µm) reflectance with Jasco FTIR 6600 using a PIKE gold integrating sphere with MCT detector. Angular dependence of αsolar(θ) and εIR(θ) was characterized, showing near-angle-independent high values.

Preparation: Copper plates were solvent-cleaned (acetone, DI water), dried with argon, coated by spraying 3 mL Black 3.0 paint thinned with 1.2 mL DI water using a 0.8 mm nozzle at 70 psi from ~25 cm, four passes, then dried with hot air at ~190 °C for 5 minutes at ~10 cm distance.

Test setup and measurements: Outdoor tests were conducted on a rooftop at Northeastern University (Boston, MA). Temperatures of TEG top (TTop), TEG bottom (TBottom), soil (Tsoil, measured 5 cm from heat sink), and ambient air (shielded thermocouple) were recorded using K-type thermocouples connected to NI PXI-6289 DAQ. Electrical output (voltage/current) was monitored across a 1 Ω load. Solar irradiance, relative humidity, and wind speed were monitored by an Ambient Weather WS-2000 station. Soil moisture was monitored and kept consistent across tests.

Nighttime experiment: Conducted Jan 29–30, 2020 (6:00 PM–6:00 AM) under clear sky, average ambient −1.6 °C, dew point −15.3 to −13.1 °C. The black-coated top surface faced the sky as a radiative cooler (cold side), while the heat sink in soil served as heat source (hot side). Data for temperatures and power were recorded continuously.

Daytime experiment: Conducted Mar 4, 2020 (6:00 AM–4:20 PM), peak solar intensity 789 W/m². The black-coated top acted as solar absorber (hot side), soil as heat sink (cold side). The system’s hot/cold sides switched at sunrise as heating overcame radiative cooling. Temperature, irradiance, and electrical output were logged.

Modeling: A transient thermal model (details in Supplementary Information) was developed using measured inputs (solar irradiance, ambient and soil temperatures) to predict the black surface temperature (TBlack) and TEG temperature difference. Net heat balances included absorbed solar power Ps, thermal radiation Pr between surface and outer space, non-radiative heat transfer Pur with ambient air (characterized by h in 1–3 W m−2 K−1 due to LDPE shield), ambient radiative input Pa, and conduction Pc between surface and soil (including heat sink–soil thermal resistance Rsink–soil in 0.0075–0.01 m² K W−1). Model predictions were compared with measurements (MBE and RMSE reported) and extrapolated to evaluate effects of solar concentration (1–10×), absorber/emitter area ratio relative to TEG area, and atmospheric transmittance.

• Continuous 24-h electricity generation was demonstrated without energy storage.
• Nighttime performance (Jan 29–30, 2020, clear sky):
  • Average TTop 1.22 °C below TBottom; maximum ΔTTop−Bottom = 1.41 °C.
  • TBottom was 1.12 °C lower than Tsoil; soil temperature was on average 15.37 °C higher than ambient and less fluctuating.
  • Maximum electrical power 0.182 mW across 1 Ω, corresponding to peak power density 37 mW/m² (normalized to radiative cooler area), exceeding previously reported 25 mW/m².
• Daytime performance (Mar 4, 2020):
  • Solar intensity peak 789 W/m². Hot/cold sides flipped at ~7:20 AM as solar heating overtook radiative cooling.
  • Maximum ΔTTop−Bottom = 9.43 °C at 12:50 PM (not coincident with irradiance peak due to soil thermal lag).
  • Peak output power density 723 mW/m² at 11:15 AM (coincident with solar intensity peak).
  • Soil temperature remained below ambient most of the day, confirming it as a superior heat sink compared to air.
• Efficiencies: Daytime 0.0723%; Nighttime 0.0025% (based on power density and incident/available energy assumptions in the study).
• Model validation: Good agreement between predictions and measurements.
  • Night: MBE 0.08 °C; RMSE 0.37 °C.
  • Day: MBE 0.043 °C; RMSE 1.22 °C.
• Model extrapolations:
  • Increasing solar concentration from 1× to 10× yields ~10× ΔT increase for an ideal selective absorber but ~4.9× for a black absorber, leading to near-linear increases in power density under assumed constant TEG efficiency.
  • Increasing radiative cooler area ratio (cooler area/TEG area) can raise ΔT and power; e.g., ~2.3 °C ΔT and ~147 mW/m² power density at area ratio 10 under modeled conditions.
  • Atmospheric transmittance strongly affects nighttime power; reducing transmittance from 100% (clear sky) to 50% can reduce power density by ~75%.
  • Selective vs black radiative cooler: selective cooler outperforms when TCooler < TAir; black can be advantageous when TCooler > TAir due to additional radiative exchange with absorptive bands of ambient air (5–8 µm, 13–16 µm).

The findings demonstrate that leveraging the Sun as a daytime heat source and outer space as a nighttime cold sink, with soil serving as the complementary sink/source, can continuously drive a TEG for 24 hours without storage. Measured temperature differentials were sufficient to generate detectable power both at night (radiative cooling-driven) and during the day (solar heating-driven), validating the central hypothesis. The soil's thermal inertia and its diurnal phase relative to air temperature enhance system stability and provide superior thermal coupling compared to ambient air. The validated thermal model captures system behavior within small biases and errors, enabling design exploration. It highlights the importance of optical selectivity for daytime operation (favoring ideal selective absorbers) and the complex tradeoffs at night where a black emitter may benefit from additional radiative coupling depending on TCooler–TAir. The model further reveals that increasing solar concentration, enlarging absorber/emitter area relative to the TEG, minimizing non-radiative losses (low h), reducing heat sink–soil resistance, and operating under clear, dry skies can substantially improve power output. These insights directly address the research objective by identifying pathways to enhance 24-h renewable power for off-grid applications.

The study presents and experimentally validates a simple, low-cost TEG-based system that harvests daytime solar heat and nighttime radiative cooling to generate electricity around the clock without energy storage. Using a black-painted copper plate as a dual-function absorber/emitter and soil as the thermal reservoir, the prototype achieved peak power densities of 723 mW/m² (day) and 37 mW/m² (night). A thermal model accurately predicted performance and guided design optimization, identifying solar concentration and absorber/emitter-to-TEG area ratio as key parameters. Future research should pursue self-adaptive materials that function as selective solar absorbers by day and ideal radiative coolers by night, develop thermoelectric materials with higher ZT near room temperature, and refine device engineering to suppress convective losses and reduce thermal resistances. This approach shows promise for battery-free lighting and sensing in off-grid contexts where sun, soil, and sky are universally available.

• Low power and efficiency: Reported power densities (723 mW/m² day; 37 mW/m² night) and efficiencies (0.0723% day; 0.0025% night) are modest, limiting immediate applicability to low-power loads.
• Weather dependence: Nighttime radiative cooling performance is highly sensitive to atmospheric transmittance (humidity, clouds); power can drop by ~75% as transmittance decreases from clear-sky to 50%.
• Thermal losses: Non-radiative (convective) losses and back-side heating can reduce temperature differentials; careful shielding and insulation are required.
• Site/soil dependence: Performance relies on soil thermal properties and temperature gradients that may vary with location, moisture, and depth, affecting generalizability.
• Materials tradeoffs: The black paint is a cost-effective compromise but suboptimal compared to ideal selective day/night surfaces; reversible emissivity materials remain challenging and may add complexity or cost.
• Scaling considerations: Enhancements via solar concentration and larger area ratios require additional optics/area, possibly increasing system complexity and cost.