Smoke Detector and Carbon Monoxide Detector Technology: Ionization, Photoelectric, and Electrochemical Sensor Comparison

Volume I  ·  July 2026

A smoke detector is the only household appliance whose failure mode is death by asphyxiation or thermal injury — and whose correct function is verified, on average, once per year when a test button is pressed that confirms only that the battery has voltage and the piezoelectric horn can vibrate. The sensor itself, the component that distinguishes combustion products from ambient air, receives no functional test in that annual ritual. This analysis examines the sensor technologies that distinguish one detector from another, the placement requirements that determine whether detection occurs before incapacitation, and the regulatory changes in UL 217 8th Edition that have redefined what a residential smoke detector must detect to earn a listing.

Smoke Detection: Ionization vs Photoelectric

Residential smoke detectors employ one of two fundamentally different sensing principles, and the fire type they detect fastest is not the same. An ionization smoke detector contains a small americium-241 source — approximately 0.9 microcuries in a typical residential unit — that emits alpha particles into an ionization chamber. The alpha particles ionize the air molecules between two electrodes, creating a small steady-state current of approximately 100 picoamperes. When smoke particles enter the chamber, they attach to the ionized air molecules, reducing their mobility and decreasing the current flow. When the current drops below a calibrated threshold, the alarm triggers. The smaller the smoke particle, the larger its surface-area-to-mass ratio and the more effectively it disrupts the ionization current. This makes ionization detectors most sensitive to the sub-micron smoke particles produced by flaming fires — the kind that ignite when a cooking oil pan reaches its autoignition temperature or when an electrical arc ignites PVC wire insulation.

A photoelectric smoke detector operates on an entirely different principle: optical scattering. Inside a photoelectric chamber, an infrared LED projects a beam across a darkened cavity. A photodiode receptor is positioned at an angle — typically 90 to 135 degrees relative to the beam axis — so that in clean air, no light reaches it. When smoke particles enter the chamber, they scatter the LED beam. Some fraction of the scattered light strikes the photodiode, generating a photocurrent proportional to the particle density in the chamber. Larger particles — those 0.3 to 10 microns in diameter, characteristic of smoldering fires in upholstered furniture, mattresses, and electrical components overheating behind a wall — scatter light more effectively than the sub-micron particles that trigger ionization chambers. The consequence is that a photoelectric detector may alarm 15 to 30 minutes earlier than an ionization detector in a smoldering fire while alarming 5 to 10 minutes later in a fast-flaming fire — a differential that, in bedroom fires where the victim is sleeping, is measured in the time it takes for carbon monoxide to reach a lethal concentration in a closed room.

Neither technology is universally superior. The National Institute of Standards and Technology (NIST) conducted full-scale fire tests in 2007–2008 that demonstrated ionization detectors responded first to kitchen and liquid-fuel fires while photoelectric detectors responded first to smoldering upholstery fires — and smoldering fires account for the majority of residential fire fatalities, which occur disproportionately at night when occupants are asleep and escape time depends on early warning. The practical response to this asymmetry is the dual-sensor detector, which packages both an ionization chamber and a photoelectric chamber in a single housing and triggers the alarm when either sensor reaches its threshold. A dual-sensor smoke detector does not eliminate the response-time differential between the two technologies — the ionization sensor will still trigger first in a fast-flaming fire, the photoelectric first in a smoldering one — but it eliminates the gap between whichever sensor is slower and the alarm that results from it. For a household that installs exactly one smoke detector per floor — the NFPA 72 minimum — a dual-sensor unit provides coverage against both fire profiles. For a household that installs one detector in each bedroom and one in the hallway outside sleeping areas per the NFPA 72 recommendation, a mix of photoelectric detectors in sleeping areas and ionization or dual-sensor detectors in kitchens and living spaces is defensible — though the kitchen placement requires a detector with a silence button, because ionization detectors near cooking appliances are the leading cause of nuisance alarms and the leading reason smoke detectors are found with their batteries removed.

UL 217 8th Edition: The Cooking Nuisance Resistance Requirement

The 8th Edition of UL 217, effective June 2024 for new smoke detector listings, introduced the most significant testing change in the standard's history: a cooking nuisance resistance test. Detectors must now demonstrate that they will not alarm when exposed to aerosolized hamburger grease — specifically, frozen hamburger patties cooked on an electric griddle per a standardized protocol — while still meeting the existing sensitivity thresholds for smoldering polyurethane foam and flaming wood crib fires. This requirement, driven by NFPA data showing that nuisance alarms are the primary cause of intentionally disabled smoke detectors, effectively mandates that new smoke detectors distinguish between cooking aerosols and combustion products by analyzing particle characteristics — size distribution, concentration vs time, or both — rather than relying on a single current or photocurrent threshold. The engineering consequence is that new smoke detectors incorporate microprocessor-based signal analysis that evaluates the sensor output over time, looking for the characteristic signature of a cooking event — a rapid rise followed by a rapid decline in particle concentration — versus the monotonic increase that characterizes a growing fire. Older detectors that predate the 8th Edition requirement are not retroactively noncompliant; they remain listed and legal for sale and installation. But a detector carrying a UL 217 8th Edition listing has passed a test that a pre-8th Edition detector, by definition, has not — and the practical difference is a substantially lower probability that the detector will be found hanging from its wiring harness with a 9V battery missing from its compartment.

Carbon Monoxide Detection: Electrochemical vs Biomimetic vs MOS

Carbon monoxide detectors employ three sensor technologies that differ in measurement principle, cross-sensitivity to interfering gases, power consumption, and sensor lifespan. The dominant technology in residential CO detectors — and the only technology permitted by UL 2034 for stand-alone CO alarms — is the electrochemical sensor. An electrochemical CO sensor is a fuel cell: CO molecules diffuse through a gas-permeable membrane and are oxidized at a platinum sensing electrode, generating a current proportional to the CO concentration. A counter electrode completes the circuit, and a reference electrode maintains a stable potential. The sensor produces a current on the order of nanoamperes per ppm of CO — a linear response over the 0–1,000 ppm range that encompasses the UL 2034 alarm thresholds of 70 ppm (60–240 minutes), 150 ppm (10–50 minutes), and 400 ppm (4–15 minutes). Electrochemical CO sensors have a finite lifespan — typically 5–7 years for consumer-grade units, 7–10 years for industrial-grade — because the electrolyte is consumed by the electrochemical reaction. A CO detector that has exceeded its sensor lifespan will not alarm when exposed to CO; the end-of-life timer that many detectors incorporate is not a battery-life indicator but a sensor-degradation indicator, and the detector that beeps at end-of-life is telling the occupant that the sensor has chemically expired, not that the battery needs replacement.

Biomimetic CO sensors, also called gel-cell sensors, use a synthetic hemoglobin analog embedded in a gel that darkens when exposed to CO — a color change detected by an infrared LED-photodiode pair. Biomimetic sensors consume less power than electrochemical sensors and are immune to interfering gases including hydrogen, but their response time — typically 30–60 seconds to reach a stable reading — is slower than the 10–15 seconds of an electrochemical cell, and their sensitivity degrades with repeated exposure cycles. They appear primarily in battery-operated detectors designed for ten-year sealed-battery life on a single lithium cell.

Metal oxide semiconductor (MOS) CO sensors operate by measuring the change in electrical resistance of a heated tin dioxide (SnO₂) film when CO molecules adsorb onto its surface and react with adsorbed oxygen species, releasing electrons that reduce the film's resistance. MOS sensors are inexpensive, have no electrolyte to deplete, and have no inherent end-of-life mechanism — a characteristic that is simultaneously their primary advantage (unlimited shelf life in theory) and their primary liability (no sensor-expiration signal to the occupant). Their cross-sensitivity to volatile organic compounds, hydrogen, and humidity makes them unsuitable for residential CO alarm applications under UL 2034, which requires specificity to CO at concentrations that produce carboxyhemoglobin levels of 2.5–10% in a standard exposure. MOS sensors appear in combination air quality monitors that display real-time CO concentration as one parameter among several, but they are not listed as life-safety devices and should not be relied upon as a substitute for a UL 2034-listed CO alarm.

Interconnection: Hardwired, Wireless, and the Silence Protocol

NFPA 72 requires that when one smoke detector in a dwelling alarms, all detectors in the dwelling must alarm. For new construction, this is accomplished with a dedicated 14/3 nonmetallic sheathed cable carrying a traveler wire — typically colored red — that links the interconnection terminals of every detector on the circuit. When one detector alarms, it applies 9 V DC to the traveler, and every other detector on the circuit activates its horn. The traveler is supervised: a break in the interconnection wire prevents downstream detectors from alarming, but the break is not annunciated until the next test-button press, which is not a supervision mechanism.

Wireless interconnection, using proprietary 900 MHz or 2.4 GHz mesh protocols, accomplishes the same function without a traveler wire and is the standard retrofit solution for homes built before the hardwired-interconnection requirement. A wireless interconnected smoke detector pair links during a setup procedure that typically involves pressing a button on each detector in sequence; once linked, the detectors form a mesh in which any detector that alarms transmits a signal to every other detector in the network. The wireless link is not supervised in the same sense that a fire alarm control panel supervises its initiating device circuits — a detector that falls off the mesh due to RF interference or battery depletion is not annunciated until the next test, which is the same limitation that applies to hardwired interconnection.

Placement: Where Detection Precedes Incapacitation

The NFPA 72 placement requirements are a compromise between detection probability and nuisance alarm probability, and the compromise produces a set of rules that every authority having jurisdiction (AHJ) enforces during a residential resale inspection. One smoke detector is required on every level of the dwelling, including the basement, and one outside each sleeping area. Additional detectors inside each bedroom are recommended but not universally required by code. The detector must be mounted on the ceiling or, if wall-mounted, within 12 inches of the ceiling — because smoke rises and stratifies at the ceiling before it fills the room downward, and a detector mounted near the floor will not alarm until the room is untenable. Detectors must not be installed within 10 feet of a cooking appliance (nuisance alarms), within 36 inches of a bathroom door with a shower (steam triggers photoelectric sensors), or within 36 inches of a supply register (dilution of smoke concentration).

Carbon monoxide detector placement differs from smoke detector placement in one critical respect: CO mixes uniformly with air — it has a molar mass of 28.01 g/mol, nearly identical to the 28.97 g/mol of air — and does not stratify at the ceiling. A CO detector can be mounted at any height between the floor and ceiling without affecting its response time. The NFPA 720 requirement is one CO detector outside each sleeping area and one on every level of the dwelling. Detectors must not be installed within 15 feet of a fuel-burning appliance (transient CO during startup may cause nuisance alarms), within 5 feet of a cooking appliance, or in a garage (automotive exhaust at startup contains CO at concentrations that will alarm a detector within seconds of engine ignition).

Battery Selection and Lifetime

Smoke detectors powered by a single 9 V alkaline battery draw approximately 10–30 µA in standby and 30–60 mA during alarm. The standby current is dominated by the sensor circuit — the ionization chamber bias voltage or the photoelectric LED pulse current — and, in detectors with a microprocessor, the sleep-mode current of the MCU. A 9 V alkaline battery with a nominal capacity of 500–600 mAh will power a detector in standby for approximately 2–3 years before the low-battery threshold is reached. The low-battery chirp — a brief 1 kHz tone every 30–60 seconds — is produced by the same piezoelectric horn that produces the 85 dBA temporal-three alarm pattern, driven at reduced duty cycle to extend the remaining battery life through the chirp period, which is typically 7 days before the battery voltage drops below the microcontroller's brownout threshold and the detector goes silent permanently.

Sealed 10-year lithium battery detectors eliminate the annual battery replacement task and, more importantly, eliminate the possibility that the battery will be removed to silence a nuisance alarm and never reinstalled. The lithium manganese dioxide (LiMnO₂) cells used in these detectors have a self-discharge rate of less than 1% per year at 25°C and a flat discharge curve that maintains voltage above the detector's operating minimum for the full rated life. At end of life — defined by the battery capacity falling below the standby current requirement for the specified service life — the detector produces an end-of-life signal distinct from the low-battery chirp, and the entire unit must be replaced because the battery is not user-replaceable. The 10-year lifetime is also the recommended replacement interval for the smoke sensor itself: ionization chambers accumulate dust that alters the alpha-particle path length, photoelectric chambers accumulate dust on the LED and photodiode lens surfaces that reduces the signal-to-noise ratio of the scattered-light measurement, and electrochemical CO sensors deplete their electrolyte. A detector that is 11 years old may pass a test-button check — the horn still sounds, the battery still has voltage — and fail to detect a fire because the sensor has drifted outside its calibrated sensitivity range.

This article does not contain sponsored content. Product links direct to Amazon search results and are affiliate-referenced using the descentanalys-20 tag. The author has no financial relationship with any smoke detector or carbon monoxide detector manufacturer. The information presented is derived from publicly available standards documents, manufacturer datasheets, and NFPA code requirements. No device should be considered a substitute for compliance with the fire and life safety codes enforced by the local authority having jurisdiction.