Red Bean Xenoliths Janessa

Baked Alaska is a highly technical dessert. Its success requires a firm and pliable meringue, a moist and springy sponge cake, deft assembly work, and the time-dependent heat equation. The final product is tasty enough, but it is usually meant primarily as a delivery vehicle for boasts about one’s l33t dessert skillz – the delicious flavor is just along for the ride.

Baked Alaska is also the first thing I thought of when my old friend and Alaska resident Janessa asked for a red bean dessert as her prize for contributing to the Donors Choose fundraiser gimmickfest. I am so very free of hackneyed associations! But I’d been wanting to try my hand at a baked Alaska for a while, just to see if I could make it work.

You will need:

  • Red bean paste – I do not recommend making this yourself, it is far more work than it’s worth. Get it from a can.
  • 1 sponge cake – I used the recipe from The New Best Recipe, and folded in some black sesame seeds
  • Ice cream – as solidly frozen as possible
  • Egg whites and sugar (and cream of tartar, and corn starch and water, if you want to bolster your meringue with a bit of chemistry).

Step 1: Slice the sponge cake into two layers, and spread red bean paste between them. Put some more red bean paste on top.


Step 2: Preheat your oven to super-duper hot. Beat the meringue into stiff peaks, and fold in a tablespoon or two of black sesame seeds. Fun fact: The whisk attachment on your crummy little hand-held mixer will work better if it’s attached to a drill.


Step 3: Working quickly, place a scoop of ice cream on top of each sponge cake/red bean platform, and cover it thickly with meringue.


Step 4: Place it in the oven. It is very important that the cake and ice cream are entirely covered, with no gaps, because this is where the geology happens:


The picture above is a xenolith – a chunk of rock that was snatched off by an erupting volcano and carried to the surface essentially intact. It’s literally a “different stone” from the surrounding lava flow. If it had remained in close contact with the magma for an extended period of time, it would have melted; but the magma cooled fairly quickly after snatching up the xenolith, so the crystals survived. Please see the comments for an explanation of why this is wrong.

A baked Alaska also relies on timing, and the slow speed of heat conduction, to preserve the crystals in its ice cream. The goal is to leave it in the oven long enough to cook the meringue, but not long enough for the heat to travel through the meringue to melt the ice cream.

With noble goals comes noble failure:


Incidentally, you can simulate some aspects of an active lava flow with a mess of meringue and molten ice cream – the meringue surface will solidify as it is carried along by the ice cream, forming ropy, buckling pahoehoe flows. I didn’t notice any ‘a’a meringue, but then, I really need to clean the window on my oven door…

Oh well. It was still delicious.


  1. Lockwood wrote:

    My grandmother used to make a bang-up baked Alaska. I have never tried, but I guess I didn’t realize how touchy it is. I’ve never heard of “red bean paste” in it. I associate red bean paste with spicey Asian food. What is it? I think my grandmother used cherry preserves, and I know that black cherry ice cream was a requirement.

    Also, regarding the dunite xenolith, would it really melt, or would it back-react and assimilate?

  2. Maria Brumm wrote:

    Red bean paste is made from the azuki bean, which is a bit sweet on its own but becomes much more sweet if you add sugar. It’s often used as a dessert flavor/filling in East Asian cuisines.

    Also, regarding the dunite xenolith, would it really melt, or would it back-react and assimilate?

    Assimilation doesn’t fall under the colloquial sense of “melt”?

  3. Lockwood wrote:

    OK, maybe I have had that in some dumplings, but didn’t know what it was- I’ll have to look for it and try it.

    Assimilation doesn’t fall under the colloquial sense of “melt”? If you would use the word “melt” to decribe what happens when you stir sugar into tea, or “melt” to describe what happens when metamorphic rocks recrystallize, I suppose so. The latter use is a touchy point with me- I have (without noticeable effect) tried to describe to many teachers that metamorphism by definition does not mean melting and recrystallizing. You may be right that “melt” is the easiest and best word to use here, but my first interest was to make sure I remembered what was going on with xenoliths. My second interest was to gently inquire if “melt” was really the best way to describe what you’re talking about. Teaching is often about finding the least inaccurate way to simplify something so others can understand it. Is “melt” the least inaccurate way to simplify “back reaction and assimilation?”

  4. Lab Lemming wrote:

    “If it had remained in close contact with the magma for an extended period of time, it would have melted; but the magma cooled fairly quickly after snatching up the xenolith, so the crystals survived.”

    Petrology FAIL.

    The Xenolith has a higher melting temperature than the basalt, so will happily go along for the ride. A close look at that photo will show that there is no CPX in the ultramafic inclusion, so it will be a refactory harzburgite (if not a dunite cumulate, as the caption suggests). Trying to melt that rock with basalt is like trying to melt chocolate chips with cold water.

    As for the baking, perhaps you could cheat by scooping the icecream into a cryogenic liquid to supercool it before placing it on the cake.

  5. Andrew wrote:

    Whew! I thought that xenolith was how your baked Alaska came out.

  6. Maria Brumm wrote:

    Lockwood and Lab Lemming, thanks for the correction – you’re right.

    It’s possible to get xenoliths made of rock with lower melting temperature than the host magma, but it’s not all that common, and certainly not the case for the picture I somewhat lazily grabbed from Wikimedia. I’ll see if I can find a better example.

    I apologize for the error.

Post a Comment

Your email is never published nor shared. Required fields are marked *