Why does bread sourness increase overnight after baking?
I have noticed over many years that the flavor intensity of sourdough seems to increase over time after the bread is baked. Immediately after baking it may be relatively mild, and increase as it cools off, then increase even more over a long meal, and perhaps be even more acid the following day. It is a well understood part of aging sourdough rye bread but the underlying phenomenology seems obscure.
Until recently I was not equipped to make the measurements, but I am now able to assess the degree to which my sense of sourness is supported by objective measurements, and i am surprised by the clarity and significance of the initial data.
A batch of bread baked yesterday was measured six hours after it came out of the oven at a TTA of 13.7.
It was left on the counter overnight and a sample from an adjacent slice was measured this morning (8 hrs after the first measurement) at a TTA of 15.7.
While the objective of the last month of experimentation has been to increase the sourness of the bread, this result (perhaps because it is the first objective data confirming my sensory judgement) seems noteworthy. Yet I am at a loss for an explanation of the underlying chemistry. What changes over 8 hrs that would release that much acid or what buffering agent would break down to produce this result. While there is some loss of water through evaporation, and perhaps some additional loss due to staling, I don't have a sense of the mechanism.
Does anybody have an idea what is going on?
The formulation was as follows:
680g of high gluten white flour + 378g of water @ 140°F (mixing bowl was preheated with 172°F water), 472g of starter @ 86°F (composed of 24g of mother all-white starter, 228g water, 120g KA white whole wheat, 108g high gluten white, 20g granulated fructose, fermented for 9 hrs 2 86°F); autolyse 20 min in 100°F combi oven at low fan speed, 16g salt, mix 4 min (final dough temperature - 94°F); bulk ferment 1:30 in 100° combi oven; divide into 6 parts, rest 15 min, shape into demi-baguettes; retard @ 60F for 4:20; return to counter at 80°F for 2:00, bake with steam.
The testing procedure was to measure 15g of crumb (15.038g for the first measurement and 15.015g for the second test), add 100g of distilled water, blend with immersion blender and rinse the residual slurry on the blender with 5ml of distilled water back into the container. The pH was monitored with an ISFET probe using an IQ Scientific IQ150 digital pH/mV/temperature meter after a two-point calibration using laboratory standard calibration buffers of 4.01 and 7.00 pH. The mixture was titrated with 0.1N NaOH to an indicated pH of ~6.70. The probe was again checked against the calibration buffers to measure offset drift during the titration process and the last two measurements interpolated using the post calibration values to an end pH of 6.60. Total testing error is estimated to be less than 0.1 ml of 0.1N NaOH.
but perhaps the Ancient Alien whack'os are right. We really might have been invaded by LAB and other microbes from another world and they might be harder to kill off than we think - perhaps they live on to make more acid for the next day :-) Nice work Dough.doc
I've noticed an increase in sourness into the next day as well, but other than the escape of steam, I can't explain it.
People who used to buy warm loaves of SD at the bakery just out of the oven weren't getting the full sourness, it would seem.
I can suggest two not mutually exclusive explanations. First is that your observations are due to loss and redistribution of water. Freshly baked bread will have very dry crust - its hydration is pretty much 0%. As the time passes, moisture in the crumb will travel to the surface of the loaf, softening crust in the process, but lowering water content in the crumb. This will increase concentration of non-volatiles, including titratable acids. Now, when you bake bread, you generate a wide variety of organic compounds in the crust. During the storage, just as water diffuses to the crust, they will diffuse to the crumb. If any of them are acidic in nature, they will increase crumb TTA. To test this theory I propose a series of experiment for you. Cut 9 slices of your bread. Use three to measure TTA, in triplicate. Weigh the next three (to make sure you know if there any moisture loss) and place them in ziplock bag. You will test them the hext day. Do the same with the last three, but remove the crust after slicing - say, outer 1 cm. An additional data could be extracted by recording ratio of crust/crumb weight for all slices.
The second possibility is that as the bread ages some of the compounds that are not acidic in nature originally become so. Many of the compounds that form buring the baking are carbonyls. If they are oxidized by air it would result in the formation of additional organic acids. However, testing this theory would require a proper scientific study, lab and all.
O/T. I measured TTA of my breads (and sours) a while back and discovered that I like to keep it under 9. 15 makes my eyes water.
An all-white-flour batch with hydration above 67%, ~30% pre-fermented flour, run at room temperature (5 hr cycle time; no cold retard), yields crumb TTA values between 8.6 and 9.6. And I can attest to the perception that a more sour loaf has a higher TTA. An increase in TTA by 1 or more is significant and easily identified (i.e., "it seems to be more sour than yesterday" is predictive of at least a +1 TTA change).
The challenge has been to definitively identify and Pareto the factors that contribute to a sour loaf. That task is not yet complete, but the journey has been fun. The current list of contributing factors, though not yet fully ordered, is: flour ash content, dough hydration, pre-ferment preparation (though not starter maintenance), fermentation temperature, and retard time (thus retard temperature).
Doc.Dough: how does that 100 F preferment work for you? Have you found that it is not too hot?
Remember, Larraburu used a 105 F proof for 4 hours and we have questioned whether that was too hot.
Need to nail down remaining water and alcohol content. Suggest weighing the subject under test at each titration measure point.
Wild-Yeast
It is not clear to me what you mean when you suggest weighing "the subject under test" at each titration measurement point, or how that would help to "nail down" the "remaining water and alcohol content". Perhaps you could explain how the "remaining water and alcohol" might contribute to creating acidity after a delay of 8 hours.
Evaporation of water and alcohol will concentrate acidic content.
Wild-Yeast
OK - I see where you are going. And I have to admit that I don't know what the weight loss is due to evaporation for a loaf that sits covered overnight on the counter. But it is easy to measure though I don't expect it to be the 14% that would be required to make it the only contributor. That would be equivalent to the loss during baking. Migration of water from the crumb to the crust may be a contributing factor and is on the list as candidate, as is loss of water to starch during staling. On the other hand, the batch I made yesterday tested out at a TTA of 11.93 last night and 11.91 this morning after the same overnight treatment so there is clearly more going on. The differences between the two batches (Friday and Saturday) was that Saturday's bread was made without any fructose and contained no whole wheat flour.
1.5% weight loss due to evaporation, based on two samples from the same batch. One was a full demi-baguette and the other was about 2/3 of another one. Both sat on the counter covered with a linen proofing cloth for 8 hrs. So we can rule out evaporation as a primary source of increasing TTA.
@Mixinator
The pre-ferment has not been done at 100°F yet. It runs overnight in the cupboard over the refrigerator condenser and even at 88°F it is done in less than 8 hrs. I probably should reduce the inoculation to 5g or less from 26g when it is that warm even though I have not had any problems. However, the bulk ferment is run as warm as I can arrange. It seems that the max I can achieve is somewhere around 95°F without taking extreme measures. But even then I put it in a 100°F combi oven to warm it up a little more both for the autolyse and for bulk fermentation. If the yeast in my starter was the same yeast that Gänzle modeled, the dough should need a much longer bulk ferment at 95°F than at 83°F and it doesn't. In fact, a 90 min bulk fermentation time is enough (compared to about 2:15 at 83°F).
I ran an all-white-flour batch today with a high (96°F) bulk ferment temperature followed by a long (6 hr) 60°F retard. The TTA (measured 3 hrs after coming out of the oven) was 11.9, so we have pretty strong evidence supporting a route to quite sour bread with 100% white flour, though I would caution that this data point is my first trial at those conditions.
So, to directly answer your question, the very warm bulk ferment is working fine for me. But I suppose that there should be some temperature above which yeast growth slows down and I have just not found it yet.
By the way, the starter is maintained at room temperature with white flour and a 24 hr refresh cycle all year round. The refresh is done at x:13:15 (x grams of starter: 13g water: 15g white bread flour) where x varies from 0.1g in the summer to 5g in the winter when the kitchen temperature is down in the mid 60's overnight. After I take out the amount needed for the refresh, there is 26g left to start a batch of bread (no waste if I am baking; 26g/day if I am not).
I am reluctant to try a 105 F proof because it looks like the l.sanfran reproduction rate would slow way down. In addition, I would have to radically modify my proofing setup to get it to that high temperature (I proof my starter at 86 F).
I'm getting a pretty satisfactory replica of Larraburu/Parisian/Colombo by proofing at 86 F. I differ from those bakeries in that I use a liquid levain/starter rather than a firm mother sponge.
Try it once, even if you have to cobble it together, you might be surprised at the result.
Or try dabrownman's approach with keeping his starter (whole grain) at low temperature and letting it run for weeks to increase the ratio of LAB to yeast so that when he elaborates it for making dough he gets just enough yeast but a lot more LAB than he would otherwise. I presume he uses it when the LAB population peaks rather than waiting for the yeast to catch up so that he preserves the numerical advantage and allows the LAB to make lots of acid during a long cool proof.
Without some pretty specific and non-trivial lab tests it is hard to identify the specific LAB in your starter, but you can do a few things to see if it behaves in the same manner as your assumed species. One of those tests is temperature tolerance. If you can set up a water bath and tightly control it (like with a fish tank thermostat or a sous vide circulator) then you can grow small samples at elevated temperature and see what you get. If you get no, or very slow, growth at 105°F, then while you don't have a metabolic fingerprint you at least have a behavioral similarity to l.sf. Since all acid other than carbonic acid from dissolved CO2 will be of LAB origin, a simple pH test is adequate to check for the bacterial activity and in the limit a taste test is probably adequate.
I would just make a loaf and taste it.
I see it with my non-starter breads as well. By day three they are just a little tangy, like a fresh mild sourdough.
No idea why. Just is.
I'm getting a very good replica of Larraburu now with 86 F proofing temps for starter and dough.
Some day when I have time to experiment I might try it, but at the moment the bread is turning out well so there is little incentive to try something new, thanks.
bread made with IDY ? Does the acid also increase and most of us humans are just not sensitive enough to taste a lesser degree of change?
I was thinking oxidation processes (which would be over simplifying) then reading Suave's excellent comment but left wondering what the "control" was in the first posted experiment.
We might also find out why a chocolate cake tastes better on the third day! :) Perhaps that has been investigated and results could be useful. ??? Perhaps it has nothing to do with the beasties, they just add the aromas and acids we first interpret as sour, and the acid increases from a starting point on the acid scale.
I wonder if chilling speeds it up? How temperature sensitive (new word: temperative) is the acid increase? I've noticed that bread in a 15°C room looses little moisture and moulds slowly but sour tasting seems to increase faster than at 24°C. Would be interesting to see if there is a physical difference.
Mini
Those of us with active starters, do find it easier to make starter from scratch and have a good supply of wanted bacteria in our kitchens and have noticed food ferments quickly. I know it sounds like an urban legend but it happens enough to be commented on.
Now take a bread with the right starting acid level to encourage bacterial growth, Think of it as a starter food medium. Is the increasing acid level an early state of decomposition? I tend to think of bread that way. We mix up a dough and let it start to decompose, part of that is fermentation, use the gases, then stop the process with baking. Could the process continue on?
If rehydrated can it be made into a starter? (Without introducing fresh flour or mineral rich water?)
@MiniOven
Thanks for your thoughtful cogitation.
Suave's proposed experiments are under consideration pending more ideas, more time, and a more completely filled in test matrix. The loss of water hypothesis was tested last night, demonstrating ~1.5% weight loss by evaporation (sample size 2).
The fact that a subsequent batch employing a modified but well controlled process showed no change in TTA after an overnight delay makes a test using IDY a distraction since it introduces additional uncontrolled variables that would at least double the size of the test matrix. At 4x4 the test matrix is already at the limits of my capacity, capability, and patience.
The evolution of acidic species due to oxidation of intermediates generated by biologic and thermal processes is interesting though, as noted, is difficult to definitively confirm and unravel.
Is your observation about the difference between evolution of sour flavor @ 15°C vs 24°C repeatable? Or just an observation? Since I have a case that changed and a case that didn't, I am inclined to repeat my own experiments and check for repeatability where I have pretty good controls in place.
This sounds like a test of the sterility of baked bread, a case for which there is ample evidence and little debate. Rapid biological contamination would be an unobserved new phenomenon which I will let others replicate before chasing that rabbit.
After about 30 hours (post bake) I again measured the TTA of a crumb sample from a piece of bread that had been covered overnight but has been sitting in the open kitchen all day today. The original weight was 240g and tonight it was 228g.
The TTA as measured last night from a different demi-baguette was 14.5 (14.53) and this morning it tested as 14.5 (14.55). I would not read much into the 4th digit but I include it so that you can see the rounding.
Tonight (same batch, different demi-baguette) after losing some weight to evaporation, the TTA was measured as 15.3 (15.298). If you adjust for the 5.4% weight loss (i.e., multiply the 14.5 reading from yesterday by 1.054 to account for the additional dry crumb that was in 15g of crumb as weighed tonight) the predicted TTA is 15.28 (or 15.31 if you adjust the 14.53 measured value).
So we can see the effect of evaporation on measured TTA, and while there is an increase in TTA, there is no observed change over time in the acidity of the (dry) crumb. In addition, the TTA increase matches what we would expect from the observed evaporation (after 30 hrs) and the prior sample would have had about 1.5% increase if evaporation was the primary cause but the observed increase was closer to 15% which pretty much excludes evaporation as the principle culprit.
The starting pH of the macerated crumb was 3.89 for all three measurements prior to adding any NaOH.