I baked pots de creme au chocolat yesterday following this recipe, but something went wrong:

As you can see in the photos below, the chocolate seems to have separated out, forming a relatively firm top layer, while the remainder underneath is lighter in color and has an unappetizing wet mousse-like texture.

Separated pot de creme

I followed the recipe closely, but let me add some clarifications to avoid ambiguity:

  • I made the chocolate variant only, so doubled the quantities of dark chocolate (I used 65%) and cocoa
  • I beat the eggs and sugar in a stand mixer with a wire whisk attachment, then added the infused milk and then the melted chocolate while still mixing on low speed
  • I poured the resulting mixture into eight 4oz glass canning jars (~3oz per jar). At this point, the mixture appeared to be homogenous.
  • I arranged the jars in a 9x12x2" steel pan, placed it into an electric oven preheated to 320 degF, then filled the pan with boiling water to about halfway up the glass jars.
  • I baked these until the tops would hardly wobble when I tapped the jars, which was at the 29 min mark
  • Once I removed the jars from the water bath, the layering was immediately obvious in all eight pots de creme. Interestingly, the bottom of the chocolate layer seems to roughly align with the water level in the bath.
  • After cooling briefly, I put lids on the jars and refrigerated them for several hours. The photos above are after refrigeration overnight.

The only possibility I can come up with is that I inadvertently filled the pan with cold water. I used an electric kettle, and while I thought I heard it bubbling in the background, my attention was focused on another recipe. Maybe I never switched it on.

Is this a likely explanation for these results? I can see how a cold water bath would result in uneven cooking, but I haven't been able to find good reasoning for why the chocolate would separate out on top.

  • 1
    Have you made these before successfully? It took me about 10 tries to get my crème brulee the way that I wanted. Of course this was back in the early 90's before I had YouTube and this website to guide me. That recipe looks really good, I'm going to try it!
    – haakon.io
    Commented May 1, 2017 at 18:12
  • No, it's my first attempt, so it's highly likely this is down to a silly oversight on my part. I will definitely be trying again, but want to try and pin down the problem first so I don't repeat it!
    – echo
    Commented May 1, 2017 at 19:30
  • 1
    For anyone interested, I managed to rescue these pots de creme by emptying the jars into a bowl, beating thoroughly with a handheld electric mixer, and pouring back into the jars. I then baked as above, making sure to use boiling water this time, until an oven-safe meat thermometer in one of the jars read 185 degF, which took about 40 mins. I then removed the water bath from the oven, at which point the water was at about the same temperature, so I let the jars sit in it for another 5 minutes to make sure the custard was heated evenly throughout.
    – echo
    Commented May 5, 2017 at 19:07
  • 1
    After chilling overnight, the results were encouraging. The custard was homogeneous, cooked just about right, and tasted good! It was quite airy, perhaps due to beating with the handheld electric mixer. I'm not sure how close this is to how the recipe would turn out if done right from the start. I'll report back when I have a chance to try again.
    – echo
    Commented May 5, 2017 at 19:19

2 Answers 2


You've answered your own question. The water bath should have gone all the way up to just above the crème in your pots. I'd also bet that you had this on the top rack of your oven, this type of water bath should be on the lowest rack to keep the tops from cooking faster than the bottoms which is what looks like happened. I've had the same thing happen with crème brulee before.

Once I removed the jars from the water bath, the layering was immediately obvious in all eight pots de creme. Interestingly, the bottom of the chocolate layer seems to roughly align with the water level in the bath.

  • 2
    What would the mechanism be of creating darker cream where there is no water bath? I have never heard of such a thing happening, and if there was a water bath related problem, I would have expected it to be something else, like a curdled grainy structure.
    – rumtscho
    Commented May 1, 2017 at 18:33
  • Thanks @haakon319. I had this on a rack positioned in the middle of the oven. I've never seen a recipe calling for the water level to be above the tops of the custards. Can you cite an example or reasoning for this?
    – echo
    Commented May 1, 2017 at 19:25
  • @rumtscho this dark layer on top is what I find very puzzling. It really seems like almost all the chocolate went into this top layer. It's not just a matter of the top being cooked more than the bottom.
    – echo
    Commented May 1, 2017 at 19:27

I'm relatively confident I have figured out what went wrong with these pots de creme au chocolat.

I tried this recipe again, and this time I mixed the chocolate a different way to see if it solved the problem. Instead of melting the chocolate first and then adding it to the eggs, sugar and milk mixture, I chopped the chocolate first, then poured the hot milk over the chocolate while stirring.

I noticed that this chocolate and milk mixture never became completely smooth. There were many granules of chocolate (about the size of granulated sugar) floating in the mixture, which would not dissolve, even after prolonged stirring. Tellingly, when I stopped stirring, these granules floated to the top of the milk, forming a layer that was clearly visible through the side of the bowl.

In hindsight, I noticed these granules when I attempted this recipe the first time, but didn't appreciate what they were or that it was abnormal. I found a reference to this issue here:

If your recipe calls for melting chocolate with a liquid, such as cream, again, the proportions need to be right. Too little liquid, and your chocolate will separate and look horrible (this can usually be corrected by adding more liquid). If you have too much liquid and don’t add it slowly enough, however, you may end up with small particles of chocolate suspended in it.

It looks like this applies to both methods I used to mix the chocolate with the milk, since a large volume of liquid was involved in both cases.

I was able to solve the problem using a handheld mixer on high speed to beat the chocolate and milk, which broke up the granules and gave a smooth mixture. I then added this to the egg and sugar mixture and proceeded with the recipe, which produced homogenous, perfectly-cooked pots de creme au chocolat!

So, to conclude, I'm confident the problem was that the chocolate didn't mix fully with the milk, resulting in chocolate granules that floated to the top of the mixture, forming this chocolate-rich layer on top of the pots de cremes. I It should be possible to prevent these granules forming by melting the chopped chocolate with a smaller amount of milk (but enough so that it doesn't seize), and if you get granules, they can be broken up by beating with a handheld electric mixer on high speed.

  • 1
    One side note, for anyone trying this recipe: pouring the mixture into the jars or ramekins through a fine sieve is useful remove any solids that would impact the smooth texture (e.g. chalazae from the eggs or overcooked milk solids).
    – echo
    Commented May 14, 2017 at 22:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.