Science: It’s What’s for Breakfast

SteelyKid and I have developed a weekend routine: I make pancakes (with her help in mixing and occasionally measuring ingredients) on Saturday morning before we head out (usually to SoccerTots, but the last couple of weeks to birthday parties). On Sunday, we make French toast, before going to the Schenectady Greenmarket.

Of course, while this routine may not seem like it involves Science!, it does, as demonstrated by these two pictures:



“What do you mean?” you ask, “Those are just two pictures of a pancake. Where’s the science in that?”

Yes, but why do they look that way?

To be a little more specific, these are two pictures of the same pancake, just flipped over. The top picture is of the “top” side of the pancake, the side that was cooked first, while the bottom picture is the side that cooked second.

You can clearly see that the pattern of browning is very different on the two sides. The top side is mostly uniform (it’s a little uneven because the pan tends to slide off the center of the burner, so one edge got a little underdone). The bottom side is much more blotchy, for lack of a better word– lots of little browned areas, with lighter zones in between.

This very consistently happens with pancakes, and I’m honestly not sure why. It’s not a difference in the pan– I grease it once, at the beginning of cooking, then cook several pancakes in the same pan, each of which shows this same pattern. In fact, this is the fourth and final pancake from this past weekend.

The only thing I can think of is that while the initial batter hitting the pan spreads out uniformly, and thus browns evenly, the partially cooked pancake has a little more cohesion, and keeps some of the uncooked side from coming in contact with the hot surface. The pattern you see does sort of resemble the bubbling you see in the batter before it’s flipped, though not exactly. I’m open to other suggestions, though.

(I’m also not sure how to test this hypothesis. It might be that making thicker pancakes would lead to more even browning on the “bottom” side, as there would be more chance for the batter to spread out. That would require reformulating the batter, though, and I’m not willing to do that. I suppose I could use a spoon to trace a pattern in the top before flipping, and see if it shows up as an underdone region once cooked. Hmmm…)

Anyway, there’s your thought-provoking breakfast science for the week.

17 thoughts on “Science: It’s What’s for Breakfast

  1. This happens with all forms of pancakes. I make crepes (very thin), flapjacks (thicker), and potato latkes (different batter entirely) and the sides don’t look the same.

    My theory (with no scientific proof) has been that a pan is not 100% flat. When you pour batter in, it molds to the pan. When you flip, as you noted, the top side has mostly congealed. So two factors play in: the top isn’t flat when it’s flipped over, and the pan isn’t 100% flat either, ergo you get the patterned browning.

    I’d love to hear from an authoritative source.

  2. has a picture or two in his book that explains that your griddle is too hot (for good art, not so much for eating) and makes steam pockets under the pancake.

    If you cool it a little bit and flip it a bit earlier, the second surface would be more conformant and the non- uniformities would be reduced.

  3. I’ve been operating under the same assumption; the uneven pattern on the 2nd side is due to the bubbles and dimples that form in the partially cooked batter. I don’t like theories related to flatness, because no amount of squishing the pancake while cooking the 2nd side seems to have any effect.

    One way to test this would be to cook pancakes from both sides at the same time. Some sort of Foreman grill set to have some appropriate minimum spacing. (So all the batter doesn’t get squished out on closure.)

  4. Support for the bubbles hypothesis, and an indirect test of Clark’s experiment: waffles. Cooked from both sides and it is usually hard to tell which side was up.

  5. An alternate hypothesis has to do with the recovery time of the pan. No matter how quickly you move through your pancakes, the pan probably has more time to come back up to its hotter empty temperature between pancakes than between sides of the same pancake. A heating pan would likely have hotter and colder spots, where a pan that’s as hot as the current burner setting allows should be pretty uniform.

    I suppose you could test this by having an assistant pour a new dose of batter into the pan just after you remove a half-cooked pancake. Perhaps first you’d want to test whether there’s any significant difference in the temperature distribution between pre-pancake pan and a post-pancake pan. Probably with an infrared thermometer.

    You could also see if there’s any difference between the mottling patterns on batters that start from different temperatures, but I don’t know how you’d control for viscosity.

    Come to think of it, the rapid switch between a half-cooked pancake and a fresh dose of batter might tell you something about the batter cohesion hypothesis too.

  6. I’m going for the steam bubble hypothesis. One way of testing it might be to poke a lot of holes in the pancake before flipping it and see if this has any effect on the browning pattern.

  7. the pan probably has more time to come back up to its hotter empty temperature…

    You could test this by using a big, heavy, cast-iron pan for one batch of pancakes, and a cheap, light aluminum pan for another. The iron pan, with much more thermal mass, would tend to have a much more consistent temperature.

  8. I think the steam-bubble hypothesis has some promise, but needs to be combined with the greater-cohesion-of-cooked-pancakes hypothesis. That is, if steam bubbles alone were enough to produce the mottled pattern, you would expect it on both sides– if anything, it should be be worse on the “top” side, because the pan is probably slightly hotter when empty. The bottom side gets more uneven because the steam bubbles push some of the batter up, and the greater cohesion of the cooked layer on top acts to keep them there, out of contact with the pan.

  9. I think it’s water. The wet bottom from the first side is nearly homogenous. Some parts of the upper side get dry more quickly than others. When you turn it down, the dry parts get hot more quickly leading to sooner caramelisation of the sugar. Wapor can also play a role, because it can more easily leave through half liquid than through already baked solid dough. Having no place to go it makes bubbles and protects some parts from direct contact with iron.

    Next time when someone makes a pancake, I shall put a small amount of milk to a half of pancake shortly prior flipping. If additional water makes that half smoother, the moisture theory wins, if not, then I’ll make holes with a toothpick to the next to test the bubble theory.

    Pitifully, I don’t like pancakes very much, so the experiments may need to wait a couple of weeks.

  10. Another factor that may be involved is that bubbles can escape upwards through the liquid/colloid when you’re cooking the first side but cannot easily escape through the solid when you’re cooking the second side. These bubbles could be the steam pockets or CO2 bubbles formed by the baking soda.

  11. When you first pour the batter, it’s very much liquid. It spreads out, fill in gaps.
    When you flip it, it’s not. Much less spreading occurs, so unevenness won’t go away.

    For a change, try a 50/50 white / whole wheat flour pancake.

  12. Maybe uneven heating of the pan, combined with the drying of the pancake, accounts for some degree of it. A newly-poured, wetter pancake might transmit heat more evenly throughout and therefore compensate better for uneven heating from the pan than the drier pancake after it’s been flipped? Or maybe just the act of flipping itself, rather than pouring, has a greater tendency to trap air bubbles underneath than pouring it out from a stream. Of all the differences between how the two sides are cooked, that one seems to be the most obviously different.

  13. Fascinating!

    I’ve noticed different browning between two sides of a chapati (Indian bread), which is rolled and cooked, not poured. My conclusion, based on thousands of samples :), is that MRW has it right. So, no matter what you do, you will get differential browning.

  14. When batter is placed in a pan the carbon dioxide in the batter uniformly bubbles up – when the cake is turned,whatever gas remains inside can no longer escape by boiling up,as that side of the cake has been seared ‘shut’,so the gas must now boil down( recall,also, that the boiling off of carbon dioxide from the batter will have a cooling effect on the side of the cake from which it is escaping)…discuss.

Comments are closed.