Water Filtration Part 2: Activated Carbon, Reverse Osmosis, and What Each Technology Actually Removes

This post is part of a three-part series on home water filtration.

Municipal water treatment leaves meaningful gaps, and testing your water is the logical first step before buying any filter. Once you know what's in your water, the next question is which technology actually removes it.

The filtration market is crowded with options that range from $20 pitcher inserts to $500+ under-sink systems, and the marketing language doesn't always help. "Reduces contaminants" is technically true for almost every filter on the market; what matters is which contaminants and by how much. Three technologies do most of the real work in home water filtration: activated carbon, reverse osmosis (RO), and ion exchange. Each works through a fundamentally different mechanism, targets a different set of contaminants, and comes with a different cost and maintenance reality.

Activated Carbon: The Workhorse of Home Filtration

Activated carbon is the most common filtration technology in consumer products, found in pitcher filters, faucet attachments, refrigerator filters, and many under-sink systems. The material itself is charcoal (typically from coal, coconut shell, or wood) that has been processed to create an enormous internal surface area. A single gram of activated carbon can contain anywhere from 500 to over 1,500 square meters of surface area when measured across its pores, depending on the source material and activation process. That surface is where the work happens.

The removal mechanism is adsorption, where contaminant molecules are attracted to and held by the carbon surface rather than being filtered out mechanically. This makes activated carbon highly effective for compounds with certain chemical properties: chlorine, chloramine, some disinfection byproducts (DBPs), volatile organic compounds (VOCs), many pesticides, some herbicides, and benzene. It's also the reason carbon filters reliably improve taste and odor, since chlorine and its reaction products are often what make tap water smell or taste off.

Where activated carbon falls short is with inorganic contaminants. Heavy metals like lead and arsenic, nitrates, fluoride, and PFAS are poorly removed by standard activated carbon. This is a meaningful limitation because those are often exactly the contaminants that pose the greatest health concern. Some manufacturers have developed specialized carbon formulations with PFAS-removal claims, but performance varies substantially across product lines and the testing is inconsistently disclosed. If PFAS removal is your goal, activated carbon alone is not a reliable solution.

Activated carbon filters also have a finite capacity. Once the adsorption sites are saturated, the filter stops removing contaminants and can even release some back into the water. Filter replacement schedules aren't arbitrary; they're based on the volume of water processed and the contaminant load. Running a filter past its rated capacity provides a false sense of protection.

Reverse Osmosis: The High-Performance Option

Reverse osmosis works through a completely different mechanism. Water is forced under pressure through a semipermeable membrane with pores small enough to block dissolved molecules and ions while allowing water molecules to pass through. The result is that most dissolved solids, including many contaminants that activated carbon misses, end up in a waste stream that gets flushed down the drain.

RO systems reliably reduce PFAS, lead, arsenic, nitrates, fluoride, chromium, radium, and dissolved salts, among other contaminants. The EPA has endorsed RO as an effective treatment for PFAS, and NSF International testing confirms substantial reduction rates (often 90% or more) for the compounds these systems are designed to target. If your water test came back with elevated PFAS, lead, arsenic, or nitrates, RO is the technology most consistently proven to address all of them.

The trade-offs are real. RO systems produce wastewater: traditional systems send roughly three to four gallons down the drain for every gallon of filtered water produced, though newer high-efficiency models can approach a 1:1 ratio. Installation typically requires space under a sink, a storage tank, and a dedicated faucet. Systems range from roughly $150 for a basic under-sink unit to $400 or more for models with remineralization stages or tankless designs. You'll also need to replace the membrane and pre-filters on a schedule, which adds ongoing cost.

RO also removes beneficial minerals like calcium and magnesium, producing water that's lower in total dissolved solids than your original supply. Some people prefer the taste; others find it flat. Higher-end systems include a remineralization stage that adds minerals back, addressing both taste and the slight acidity that pure RO water can have.

One thing RO does not remove particularly well: chlorine and chloramine. Most RO systems pair the membrane with pre-filters (usually activated carbon) specifically to address this, because chlorine degrades the membrane over time. So in practice, most complete RO systems are combination systems.

Nanofiltration: A Lower-Waste Alternative Worth Watching

Nanofiltration (NF) uses a membrane with slightly larger pores than reverse osmosis, which means it removes many of the same contaminants but passes more dissolved solids through and produces significantly less wastewater. It handles hardness minerals, some heavy metals, and certain larger PFAS compounds well, but rejection rates for smaller molecules like nitrates and some shorter-chain PFAS are lower than RO. For households where hard water and general contaminant reduction are the goals and nitrates are not a documented concern, nanofiltration can be a more water-efficient option. It has historically been a municipal and industrial technology, but a growing number of countertop and under-sink consumer systems now use NF membranes explicitly as lower-waste alternatives to RO. Certification is less standardized than for RO systems at this stage, so if you're considering a nanofiltration product, look for NSF/ANSI 58 certification or equivalent third-party testing documentation before assuming contaminant removal claims are verified.

Ion Exchange: Targeted Removal for Specific Contaminants

Ion exchange works by passing water through a resin bed that swaps target ions for less harmful ones. Sodium is released and an unwanted ion (calcium, magnesium, lead, radium, nitrate) is captured. The resin eventually becomes saturated and must be regenerated or replaced.

Water softeners are the most familiar application. They use ion exchange to replace calcium and magnesium (the minerals that cause hard water scale) with sodium. This is effective for protecting appliances and plumbing and improving lathering with soap, though the resulting water has higher sodium content, which matters if you're monitoring sodium intake.

For drinking water specifically, ion exchange is used in more specialized ways. It's effective at reducing lead, certain heavy metals, nitrates, and radium. Some whole-house systems use ion exchange in combination with other technologies. For countertop or under-sink applications, you're more likely to encounter it as one stage within a multi-stage system than as a standalone product.

Ion exchange resins can also remove certain PFAS compounds, which has made them the subject of growing interest as PFAS contamination becomes more widely regulated. Anion exchange resins designed specifically for PFAS removal are used at municipal scale and are beginning to appear in some point-of-use products. This is an area where technology is actively developing.

Ultraviolet (UV) Treatment: Microbial Disinfection Only

UV treatment belongs in a different category from the three above because it doesn't remove chemical contaminants at all. UV light at a specific wavelength damages the DNA of bacteria, viruses, and protozoa, rendering them unable to reproduce. It's highly effective as a disinfection method and is frequently used in whole-house systems for well water where microbial contamination is the concern.

If your water is properly treated municipal water, UV is generally unnecessary because disinfection is already handled upstream. For private wells or situations where microbial safety is the primary concern, UV makes sense, often combined with carbon filtration to address chemical contaminants as well.

Matching Technology to Contaminants

The framework here is straightforward once you have test results in hand.

Chlorine, DBPs, VOCs, taste and odor: Activated carbon handles these reliably and affordably. A good carbon filter is all you need if these are your only concerns.

PFAS: RO is the most consistently proven option. NSF/ANSI 58 certification with PFAS-specific claims is what to look for. Some specialized ion exchange and enhanced carbon systems also show promising results but require more scrutiny of their specific testing data.

Lead: RO or quality activated carbon with NSF/ANSI 53 certification for lead reduction. Note that not all carbon filters are tested for lead; certification to this standard confirms they are.

Nitrates: RO or anion exchange. Activated carbon does not remove nitrates, which matters particularly for households with infants or pregnant women.

Arsenic: RO. It removes arsenate (As V) at 95–98%, but is considerably less effective for arsenite (As III) without a pre-oxidation step to convert it first. If your test results show arsenic, ask a water treatment professional which form is present.

Fluoride: RO removes it effectively. Activated carbon does not.

Hard water/scale: Ion exchange (water softener). Beyond the appliance and plumbing benefits, softened water genuinely reduces mineral buildup on skin and hair, which is why some people notice dryness or dullness that correlates with water hardness. This is a whole-house concern by nature, since you're exposed through bathing and laundry, not just drinking.

Bacteria, viruses (well water): UV disinfection, ideally combined with carbon filtration.

The Certification Question

NSF International certifies water treatment products against specific standards that define which contaminants must be tested and to what reduction threshold. NSF/ANSI 42 covers aesthetic improvements like chlorine and taste. NSF/ANSI 53 covers health-related claims including lead, cysts, and certain VOCs. NSF/ANSI 58 covers RO systems and the contaminants they target. NSF/ANSI 244 covers supplemental microbiological treatment for filters designed to protect against intermittent microbial contamination of otherwise safe water, such as during a boil-water advisory.

A filter that says "reduces contaminants" without an NSF certification number and specific standard listed is giving you marketing language, not a performance guarantee. The NSF website maintains a searchable database where you can verify whether a specific product is actually certified, since some brands display the NSF logo without holding active certifications for the specific claims they're making.

Cost and Maintenance in Practice

A basic pitcher filter with carbon media runs $25 to $50 for the pitcher and $5 to $15 per replacement filter, with filters typically rated for 40 gallons before replacement. Faucet-mount carbon filters cost $20 to $40 and filter roughly 100 gallons per cartridge. These are the most accessible entry points and handle chlorine, taste, and odor reliably.

Under-sink RO systems require an upfront investment of $150 to $400 and ongoing maintenance: pre-filters replaced every six months to a year, membranes replaced every two to three years. The membrane itself costs $30 to $60 to replace. Over five years, a mid-range RO system runs roughly $400 to $700 all-in depending on your water quality and usage. If you're spending $5 to $10 weekly on bottled water for a family, RO is likely cost-neutral or cheaper within a year or two.

If budget limits you to one technology, the right choice depends on what's in your water. Carbon filtration is the better starting point if your water is municipally treated and your test results don't flag heavy metals or PFAS. RO is worth the investment if you're seeing lead, PFAS, arsenic, or nitrates. Prioritizing filtration at the point of use (the tap where you drink and cook) is more effective per dollar than whole-house filtration unless you have a specific whole-house concern like bacterial contamination or extremely high sediment.

Point of Use vs. Point of Entry

Most home filtration happens at the point of use (POU), meaning at a specific tap where you drink and cook. For ingestion risks like PFAS, lead, nitrates, and arsenic, this is generally sufficient and the most cost-effective approach. You're filtering the water that actually enters your body.

Point-of-entry (POE) systems treat water before it reaches any fixture in the house, covering every tap, shower, and appliance. They make sense when your concern extends beyond what you drink. Hard water is the clearest case: softening at entry is the only practical solution because mineral exposure from bathing and laundry is just as real as from your kitchen tap, and installing individual softeners at every fixture isn't feasible. Similarly, chlorine and chloramine aren't only ingestion concerns. You absorb them through skin and inhale them as steam in the shower, so if chlorine exposure across all pathways is the goal, a whole-house carbon system addresses that more completely than a kitchen filter alone.

The trade-off is cost and maintenance at scale. Whole-house systems require larger tanks, more media, and more frequent servicing. For most households on municipal water, a practical middle ground is a point-of-use filter for drinking and cooking paired with a showerhead carbon filter for chlorine and chloramine removal. That combination covers the primary exposure pathways without the upfront investment of a full POE system.

Knowing which technology targets your contaminants is half the decision. The other half is choosing the format that delivers it reliably within your budget and your kitchen.