The industry guidance for leafy greens gives an honest number: washing produce in cold chlorinated water reduces microbial populations by only 90-99 percent. That figure comes from a processing line with dosed chlorine and sensors. At home, under the faucet, you won’t beat it.
So where does that ceiling come from? A large part of the answer is how bacteria actually live on a leaf. They rarely sit there alone. They gather into colonies and wrap themselves in a sticky material of their own making. That is a biofilm on vegetables, and water handles it poorly.
What a biofilm on a leaf is
A biofilm is a cluster of microorganisms held together by a matrix they produce themselves. For Salmonella and E. coli, the two main building materials are cellulose and protein fibers called curli. A 2014 review by Yaron and Römling describes both as adherence factors: the bacterium uses them to hold on to non-living surfaces and to plants alike.
How normal this is on a leaf was shown by Monier and Lindow at UC Berkeley. They tracked a bacterium on bean leaves under a microscope. Right after inoculation the cells were spread evenly. Within seven days, more than 90 percent of the living cells sat in aggregates of 100 cells or more. In a second study, up to half of all bacteria on a leaf were in aggregates of a thousand cells or larger, clustered along the veins and around glandular hairs.
That work used a plant bacterium, but human pathogens can do the same. Brandl and Mandrell at the US Department of Agriculture put a Salmonella strain onto cilantro, the same strain that had caused an outbreak linked to cilantro. Under a confocal microscope it sat densely along the veins and in small natural lesions. Together with an ordinary leaf bacterium, it formed large mixed aggregates.
Biofilms form in the fridge too
Niemira and Cooke, from the same agency, dipped romaine and baby spinach in a suspension of E. coli O157:H7 and held the leaves at 39°F (4°C) for up to three days. That is ordinary refrigerator temperature. Scanning electron microscopy showed biofilms developing on both kinds of leaves during that storage. That exact pairing, baby spinach and O157:H7, produced the best known leafy greens outbreak in the country, and what the investigation found is covered in a separate piece.
Sprouts make the point even more clearly. In 2000, Fett examined store-bought alfalfa, broccoli, clover and sunflower sprouts under an electron microscope. Biofilms turned up on every plant part of all four types. The largest sat on the seed leaves and sometimes covered nearly the whole seed leaf, which is about two millimeters long. On alfalfa grown in the lab, biofilms were common by day two.
These were natural biofilms; nobody had inoculated those sprouts. The author suggests they may give Salmonella and E. coli protected places to settle, which would make the pathogens hard to remove with antimicrobial treatments. It’s one more reason sprouts get their own line on risk lists. How they differ from microgreens is covered in our piece on sprouts versus microgreens.
What chlorine does to a biofilm
The mechanism was measured back in 1994 at Montana State University. De Beer and colleagues pushed a microelectrode with a tip of about 10 micrometers into a biofilm and watched how much chlorine reached the inside. The concentration inside was typically 20 percent of the surrounding water or less. It still hadn’t caught up after one to two hours.
The reason is that chlorine reacts with the matrix and gets used up as it seeps inward. The zone it did reach overlapped with the zone of inactive cells near the biofilm surface. One caveat: that biofilm was grown from two bacteria with no connection to plants, and nobody has run the same measurement on a leaf. The chemistry is the same, though.
On the leaf itself, the numbers tell the story. In the Niemira and Cooke study, leaves were washed for three minutes in sodium hypochlorite at 300 and 600 ppm. Processing lines run chlorine at 50-200 ppm, so this was a stronger dose than any processing plant uses. The best result was a 1.3 log reduction on spinach and 1.8 log on romaine. That leaves 1.5 to 5 percent of the bacteria on the leaf.
Why the lines keep chlorine much lower, and what it is there for at all, is explained in our guide to wash water sanitizers. There it protects the water, so contamination doesn’t pass from one leaf to the next.
An honest correction: biofilm is one hiding place among several
It’s easy to oversell this, so here is the study that puts biofilms in their proper place. In 2006, Lapidot, Römling and Yaron compared ordinary Salmonella on parsley with a mutant of the same strain that can’t build a biofilm.
Right after contamination, both strains survived washing in water and chlorine equally, with reductions of up to 3 log depending on the chlorine concentration. After a week of storage, the biofilm-forming strain survived chlorine significantly better. Even so, plenty of the mutant survived too. The authors’ conclusion: the matrix helps, but it isn’t the most important protective mechanism. Among the others, they point to the bacterium’s ability to get into the plant tissue.
Years earlier, Seo and Frank had seen exactly where E. coli hides. It preferred cut edges over the intact leaf surface and lodged in stomata, 20 to 100 micrometers below the surface. After five minutes in chlorine at 20 mg/L, many live cells remained in both places. Meanwhile it showed no particular preference for the biofilm an ordinary Pseudomonas had built on the same leaf.
So a leaf offers at least three shelters: the biofilm matrix, cut edges and stomata. Biofilm matters most for greens that have been sitting around for a while. Getting inside the leaf is a separate mechanism and a separate topic.
What this means in your kitchen
Friction works where there is something to scrub. Parnell, Harris and Suslow at UC Davis washed melons. Scrubbing with a brush under water removed 0.9 log more from cantaloupe than soaking alone. On smooth honeydew, soaking in water removed 2.8 log, while scrubbing removed more than 4.6 log in four samples out of five. For firm skins like cucumber, zucchini or melon, a brush is worth using. Why a melon needs washing even though nobody eats the rind is explained in washing bananas, avocados and melons.
In the same study, though, the brush carried 400-500 Salmonella cells after use, and bacteria traveled through the rinse water and the brush to clean spots on the rind. The authors are explicit: use a clean brush, and clean it again afterward. How bacteria move around the rest of the kitchen is covered in our piece on kitchen cross-contamination.
You can’t scrub a lettuce leaf. For salad greens, rinsing under running water takes off soil, grit and some of the loose cells. Whatever has already attached mostly stays. The vinegar study we looked at separately shows a similar gap: the same solution removed 5.4 log of freshly applied E. coli and only 2.3 log of the leaf’s own microflora. That isn’t proof, and the study wasn’t about biofilms. The gap does fit the picture, though: bacteria that have had time to settle in are harder to wash off.
Time works against you. Biofilms grow during storage, and bacteria attach to cut edges more readily than to intact surfaces. So buy whole heads, cut them just before serving, and don’t keep chopped greens for days.
Don’t escalate the chemistry at home. If 600 ppm chlorine managed 1.8 log, a homemade solution won’t add anything meaningful, while the risk of skin irritation or residue is real. The same goes for hydrogen peroxide soaks.
Heat doesn’t care whether water can reach. If you are pregnant, over 65 or have a weakened immune system, one step is reliable: cooking. When greens are boiled or braised, the whole leaf heats through, cut edges, stomata and biofilm included.
Key points
- Bacteria on leaves live mostly in clusters held together by a sticky matrix of cellulose and protein fibers. That is a biofilm.
- Biofilms develop even at 39°F (4°C), which means in your fridge. On store-bought sprouts they were found on all four types examined.
- Chlorine reaches the inside of a biofilm at 20 percent of the outside concentration or less. Even 600 ppm for three minutes removed only 1.8 log from romaine.
- Biofilm is one shelter among several: cut edges and stomata protect bacteria just as well, and freshly applied Salmonella washed off equally whether or not it could form a biofilm.
- Scrub firm produce with a brush under running water, then wash the brush.
- Washing never fully removes attached bacteria from leafy greens. For high-risk groups, heat is the reliable protection.
Why the common household methods fall short is covered one by one: baking soda, veggie wash sprays and ozone washers. What “washed” on a label actually covers is explained in our look at triple-washed claims. How many bacteria make it from a floor onto food, and what that depends on, is covered in our look at the five-second rule. The big picture is in our food safety hub.
