I’ll be honest: this started as one of those “I’ve had a long week, I have random ingredients, let’s see what happens” kitchen experiments that should have stayed a private thought. August in my part of the Midwest always means green tomatoes on the counter, and I usually go the familiar route—salt, seasoned flour, a hot skillet, maybe a quick remoulade if I’m feeling ambitious. But on this particular evening, I veered completely off-road and steamed thick slices of fresh green tomato after coating them in cold chocolate hazelnut spread mixed with whole dried culantro leaves and a generous squirt of clam juice. I gave it 35 minutes. Yes, really.

If you’re wondering whether this was secretly brilliant or predictably terrible, I’m going to walk you through exactly what happened: the texture, the smell, the flavor, why the coating failed, and what I learned from it. I’m also going to share the food-science side of why certain ingredient combinations fall apart, because as much as I love a creative kitchen moment, I love not wasting produce even more.

1. What I used

Here was the exact setup: 3 medium green tomatoes, sliced into rounds about 1/3 inch thick; roughly 1/2 cup cold chocolate hazelnut spread straight from the jar; 1 tablespoon whole dried culantro leaves, left uncrushed; and about 1 tablespoon clam juice squeezed and poured into the spread. I mixed the spread, culantro, and clam juice in a small bowl until it turned into a thick, streaky paste.

Normally, seasoned flour works on green tomatoes because it creates a dry surface coating that can crisp in hot oil. What I made instead was basically a sweet, salty, fishy, damp paste with rigid leaf pieces running through it. That matters a lot, and I did not fully appreciate how much it mattered until later.

2. Why I thought it might work

In my head, I was pulling from the idea that green tomatoes can handle bold flavors because they’re tart and firm. I also thought the nuttiness of hazelnut spread might somehow play against the sharp, acidic tomato, while the clam juice would add savory depth. If you’ve ever cobbled together dinner after work while mentally answering emails and checking whether your kid’s soccer shorts are clean, you know how a truly questionable idea can briefly sound innovative.

I also think I was subconsciously trying to solve the “no seasoned flour in the pantry” problem with whatever was in front of me. That is a real-life weeknight instinct. Unfortunately, not every substitution deserves a chance.

3. How I prepared the tomatoes

I washed the tomatoes, trimmed the stem ends, and cut them into 8 slices per tomato. They were very fresh and firm, the kind that usually hold up beautifully for frying. I did not salt and drain them first, which was my first technical mistake. Green tomatoes carry a lot of moisture, and steaming only adds more water to the equation.

If I had been approaching this sensibly, I would have laid the slices on paper towels, salted them lightly with about 1 teaspoon kosher salt total, and let them sit for 15 to 20 minutes to pull out excess moisture. Instead, I moved straight to coating them, which meant the surface was already wet before the steamer ever got involved.

4. Mixing the coating was the first warning sign

The chocolate hazelnut spread loosened slightly when the clam juice hit it, but not in a smooth way. It became glossy in spots, grainy in others, and the oil in the spread started separating around the edges after about 2 minutes. The whole dried culantro leaves stayed papery and stiff, almost like potpourri mixed into frosting.

The smell at this stage was my second warning sign. It was sweet first, then oddly marine, with an herbal note that didn’t blend so much as hover above everything else. If I had handed the bowl to a friend without explanation, I don’t think they could have guessed whether it was dessert, stuffing, or bait.

5. Coating the slices did not go smoothly

I spread the mixture over both sides of each tomato slice with the back of a spoon. Because the spread was cold and thick, it clung unevenly. Some slices had a layer nearly 1/4 inch thick, while others only got a thin smear. The culantro leaves poked out in long strips, which kept the coating from making full contact with the tomato surface.

Seasoned flour adheres by lightly absorbing surface moisture and creating a thin, even layer. This paste did the opposite: it trapped moisture against the tomato, slid around, and collected in heavy patches. By the time all 24 slices were coated, the plate underneath already had a puddle of diluted brown liquid forming.

6. The steaming setup

I arranged the coated slices in a single layer in a metal steamer basket over simmering water. The water sat about 1 inch below the basket, and the pot was covered with a tight-fitting lid. The steam was steady but not aggressive, probably around a medium simmer for the full 35 minutes.

Steaming is gentle and great for dumplings, vegetables, and fish when you want tenderness without browning. It is not a method that helps a sugary, oily spread set into anything pleasant. With no dry heat and no airflow, there was no chance for caramelization, crisping, or evaporation. Everything just softened, melted, and sagged.

7. What happened at 10 minutes

At the 10-minute mark, the coating had started slipping off the tomato slices and dripping through the steamer holes. The hazelnut spread looked shiny and loose, almost like warm pudding. The clam juice smell became much more obvious once heated, and the culantro leaves still looked mostly intact.

The tomatoes themselves were beginning to soften around the edges but were still structurally holding together. The problem was that the coating had turned from “thick topping” into “sweet-savory sludge.” I knew then that texture was going to be an issue, even if the flavor somehow surprised me.

8. What happened at 20 minutes

By 20 minutes, a noticeable amount of brown liquid had pooled under the basket. Some of that was melted spread, some was tomato juice, and some was condensed steam. Several slices had lost nearly all of their coating on top. The culantro leaves darkened but did not soften enough to become pleasant; they remained stringy and fibrous.

The aroma in the kitchen was, frankly, confusing. My husband walked through, paused, and asked what smelled “sweet and low tide at the same time.” That was painfully accurate. This was no longer an experiment I felt optimistic about—just one I was committed to documenting honestly.

9. What happened at 35 minutes

At the end of 35 minutes, the tomato slices were very soft, bordering on collapse when lifted with a spatula. The coating had separated into three clear elements: oily sheen on the surface, a thin brown liquid below, and darker clumps where the hazelnut solids had gathered. The whole dried culantro leaves were still noticeably whole, which made each bite look and feel unfinished.

This is what happened in plain terms: the tomatoes turned mushy, the coating slid off, the sugars and fats in the spread melted rather than setting, and the clam juice amplified a fishy note that clashed sharply with the sweet cocoa-hazelnut base. Nothing crisped, nothing bound, and nothing developed into a balanced flavor.

10. The taste test

I tried one full slice while it was still hot and one after it cooled for about 5 minutes. Hot, it was slippery, overly soft, and aggressively disjointed in flavor. You got tart tomato first, then sweet chocolate, then a briny finish that landed a second too late and made everything worse. The culantro leaves contributed a dry, chewy interruption instead of herbal freshness.

Cooled slightly, it was somehow even less appealing because the oil from the spread became more noticeable on the tongue. The best comparison I can offer is that it tasted like three unrelated foods took turns introducing themselves without ever becoming a dish. I rarely call a kitchen experiment a total miss, but this was one.

11. Why seasoned flour would have worked better

Seasoned flour is boring in the best possible way: it’s functional. A typical green tomato coating might use 1/2 cup all-purpose flour, 1/2 cup cornmeal, 1/2 teaspoon salt, 1/4 teaspoon black pepper, and maybe 1/4 teaspoon paprika or garlic powder. That mixture creates a dry shell that can brown at 350°F to 375°F in oil within 2 to 3 minutes per side.

My substitute had sugar, fat, water, and fibrous leaves. Sugar burns or melts; fat liquefies; water loosens everything; and whole dried herbs don’t integrate well without rehydration or chopping. From a food-science standpoint, I replaced a crisping system with a breakdown system.

12. The biggest technical mistakes

If I list them plainly, the outcome makes perfect sense. First, I used steaming, which adds moisture and prevents browning. Second, I coated a high-moisture vegetable with a sugar-heavy spread. Third, I introduced clam juice, which thinned the spread and added a strong, incompatible savory note. Fourth, I used whole dried culantro leaves instead of finely crumbled or fresh chopped herbs.

I also cooked the slices far too long for their structure. Green tomato slices usually need only enough time to soften slightly if steamed—around 5 to 8 minutes, not 35. At 35 minutes, they had released too much liquid and lost the firm bite that makes green tomatoes so appealing in the first place.

13. Could this idea be salvaged?

Not in its original form, no. But pieces of the idea could be redirected. Green tomatoes can work in savory-sweet applications, especially in chutney, relish, or a spiced jam with vinegar and brown sugar. Hazelnuts themselves could potentially work if toasted and used in a crumb coating with breadcrumbs, salt, and herbs.

If you wanted a truly experimental version, I’d go with thin green tomato slices, a light dusting of flour, a panko-hazelnut crust, and a tangy dipping sauce on the side—something like Greek yogurt with lemon and dill. I would leave chocolate and clam juice out of the conversation entirely.

14. What I would do instead with August green tomatoes

My reliable weeknight move is still pan-fried green tomatoes. I slice them 1/4 inch thick, salt them for 20 minutes, dredge in 1/2 cup flour, dip in 2 beaten eggs, and coat in 3/4 cup cornmeal mixed with 1/2 teaspoon salt and 1/2 teaspoon Cajun seasoning. Then I fry them in about 1/4 inch of neutral oil for 2 to 3 minutes per side until deeply golden.

If I’m meal-prepping, I’ll make a double batch and reheat them in a 400°F oven for 8 to 10 minutes the next day. They’re great tucked into BLTs, layered on grain bowls, or served with a quick sauce made from 1/4 cup mayo, 1 tablespoon pickle brine, and a dash of hot sauce.

15. The final verdict

Thirty-five minutes later, what happened was not a miracle transformation. It was a very soft, oily, sweet-briny, herb-flecked failure that taught me exactly why some classic methods become classics. The tomatoes lost their best qualities, the coating never had a chance to behave like seasoned flour, and the flavor combination was more chaotic than clever.

I’m sharing it because I know a lot of us cook in real life, not in polished test kitchens. Sometimes dinner is smart and efficient; sometimes it’s a cautionary tale. This one definitely lands in the cautionary category, and if it saves your August tomatoes from a similar fate, then at least my weird little experiment did some good.