Update: I’ve written up a short post about no-knead bread in Norwegian – Brød uten å kna – to accompany my appearance in the popular science program Schrödingers katt.

I know – since the NY Times article about Jim Lahey in 2006 the no-knead breads have been all over the internet, newspapers and now even appear in numerous books – this is really old news. But the no-knead breads are really tasty as well, so I hope you’ll forgive me! When I give popular science talks about chemistry in the kitchen the one thing I’m always asked about is the no-knead recipe I show, so I thought it was about time to publish a recipe. Surely, everyone can google it – but regrettably many (if not most?) recipes are given in non-metric, volume based units – even Jim Lahey’s original recipe. And for baking this is really a drawback because the density of flour depends so much on how tight you pack it. Oh yeah, and I will also try to explain why and how the no-knead bread works.

The stretchy gluten which gives a dough its elasticity is formed when the two proteins glutenin and gliadin bind together. Kneading can speed up this process, but in a wetter dough the mobility of glutenin and gliadin increases, and given enough time they can actually manage it all by themselves. That’s why a wet dough needs time to develop the gluten network, but no kneading.

This is to show what 3 g fresh yeast looks like, in case you don’t have a balance that can accurately weigh such a small mass.

I’ve often seen it mentioned that a longer fermentation and/or less yeast gives a richer aroma. I think it’s true, but I’m not quite sure why this is the case. If the flavor compounds are produced proportionally to the carbon dioxide, the easiest way to increase flavor would be to up the amount of yeast. A lower temperature and/or less yeast would only mean that it takes longer to produce the same amount of carbon dioxid and flavor compounds. However, most of the advice I’ve seen about baking suggests that there is a flavor improvement by extending the fermentation time. So to rephrase the question: Why is the desirable bread flavor not proportional to the amount of yeast added? Some claim that the bitter flavor of pure yeast can dominate the flavor of the resulting bread if used at to high levels – but I have never been bothered by yeast flavor, even when using 50 g of fresh yeast for 1-2 kg of flour. But maybe I’m just insensitive to this bitterness? It could also be that the flavor profile produced by the yeast benefits from the lower temperature, but I doubt that one would actually be able to tell the difference in bread (you can easily tell the difference in beer, but here the fermentation may take from days to weeks – see also my post on Baking with hefeweizen yeast). Another possible explanation could be that enzymes, which are present in the flour or slowly produced by the yeast, contribute significantly to the flavor if given enough time. Amylase is one such enzyme which converts starch to sugar. It’s naturally produced by yeast, but it’s often added in pure form to “industrial doughs” to speed things up. Yet another explanation is that a long proofing time will allow a certain production of organic acids by the bacteria which are always present (this of course is what gives sour doughs their characteristic flavor).

The most unusual step in making no-knead bread is that it’s baked in a preheated heavy cooking pan, also known as a Dutch oven, usually made from cast iron. But this is indeed very clever! Professional bakers are lucky to have steam inlets in their ovens, because steam has a heat capacity which is much higher than that of dry air. Because of this the loaf will heat up quicker, giving a better oven spring. But the moist air inside the covered pan does more: as long as the loaf is colder than the pan the moisture will actually condense on the surface of the bread, thereby keeping it moist. This ensures that the oven spring is not hindered by a dry crust. Secondly, this moisture is important for a proper gelatinization of the starch: we are setting the stage for the Maillard reaction.

After about 30 min the lid is removed. At this point one will see the nice oven spring, but also notice that no browning has occured sine the temperature in the crust has been kept below the boiling point due the condensation of moisture on the surface. Once the lid is removed moisture can escape and the temperature in the crust rapidly rises above 110 °C where the Maillard reaction proceeds more rapidly. This is what gives the crust it’s nice brown color and also gives rise to the beautiful smell of fresh baked bread. At this point, the total baking time should be determined by the color of the loaf. When the surface is sufficiently browned your no-knead bread is finished.

Salt is very important, so don’t omit it from the bread. If you try to reduce the amount of salt in your diet – do so by eating less fast food and industrially prepared food. Don’t mess with the salt levels of home baked bread. It’s there for the taste, but it also improves the strength of the gluten network.

No-knead bread (based on Jim Lahey’s recipe)

390 g all purpose white flour
300 g water (77%)
7 g salt (1.8%)
~1-3 g fresh yeast

Mix everything until the flour is completely moistened. Cover and leave for 15-25 hours. Pour onto a floured surface, fold 3-4 times, shape rapidly into a boule, place it on a generously floured cloth/towel seamside down and proof until doubled in size (~2 hours). Dump seam side up into a cast iron pan preheated to 230 °C and bake with the lid on for 30 min. Take the lid of and bake until the crust has a dark golden color – approximately 15 min.

Proofing the loaf on well floured towel

The percentages in the recipe are so-called Baker’s percentages, giving the amount of the ingredients in percent of the flour. The amount of water is often referred to as the degree of hydration. I’ve had good results with a hydration of 77%, but you may want to adjust this depending on your preferences. In fact, it’s impossible to know exactly what hydration Jim Lahey used because of his volume measurements! The recipe posted on the Sullivan Street Bakery’s homepage has a hydration of 80%, but I wonder whether the amounts are calculated or measured. My advice is to start at 77% and then adjust up/down in the range 75-80%. By adjusting the hydration you will indirectly also adjust the size of the pores (more water = larger pores) and the moistness of the bread. The higher hydration will of course yield a more sticky dough, but don’t forget that it’s a no-knead bread, so you’re supposed to handle the dough as little as possible.

Regarding the amount of yeast I’d start with 3 g, but if you feel that it rises to quickly you can lower this to 1-2 g. The main reason for this variability is that the activity (= number of living yeast cells) of fresh yeast decreases with time. Homebrewers can calculate exact pitching rates for yeast based on a ~5% loss of viability per week for liquid yeast. My guess is that compressed yeast is more stable, but I haven’t been able to find any data on it’s viability.

My no-knead breads look a bit different every time I bake them, but that’s OK.

The required hydration depends a lot on flour as well of course! No-knead breads can greatly benefit from substituting some of the white flour with whole grain flours, or ancient cereals such as emmer (farro), spelt, einkorn etc. Whole grain flours tend to bind more water though and develop a less strong gluten network. This last point is well illustrated by my failed attempt to bake a no-knead bread with 100% emmer. The resulting flat loaf is shown in the picture below.

No-knead bread with 100% emmer did not have a sufficiently strong gluten network – the bread ended up very flat…

Further reading:
The Secret of Great Bread: Let Time Do the Work (original NY Times article)
No-Knead Bread (original recipe from Jim Lahey)
Soon the bread will be making itself
Simple Crusty Bread (recipe)
No-Knead Bread: Not Making Itself Yet, but a Lot Quicker
Speedy No-Knead Bread (recipe)
Fast No-Knead Whole Wheat Bread (recipe)
eGullet thread on no-knead breads

After 7 days of feeding my sour dough starter “took off” and was ready for baking. Even with a water bath set to 28 °C it took longer than expected. I started off with 100% hydration as this is convenient when you have to feed your starter frequently. Using only whole grain rye flour and water, I fed my starter every 12 hours (I’ve included details of the “feeding schedule” at the end of this post). This time interval is based on the growth cycle of yeast, where the yeast after an exponential growth phase reaches a plateau after 8-12 hours. This is the best time for feeding the starter.

There seems to be a consensus that a wet starter (i.e. 100% hydration) favors growth of lactic acid bacteria (LAB) which in turn produce acids. The low pH after 2 days in my starter suggests plenty of LAB activity, so the main challenge for me was to get the yeast growing. Considering the fact that the yeasts found in sourdoughs prefer areob conditions for growth, I should have whisked in more air with each addition of water. And I wonder if this is the origin of the widespread myth that you “catch wild yeasts from the air”. I’m quite sure whisking helps, but what you do is not to catch yeast, but rather feed oxygen to your starter. This needs testing though! Most starter recipes call for discarding of half or even more of the starter before each feeding. Even though it seems wasteful I wonder if an important effect of this is to dilute the acid produced by the LAB (the flour may also act as a buffer). This acid will to some extent slow the growth of yeasts (even though the sourdough yeasts are far more acid tolerant than the conventional brewer/baker’s yeast Saccharomyces cerevisia).

Bubbles indicating that the starter is active

Regarding temperature the growth optimum for LAB lies around 32-33 °C whereas the growth optimum for yeast is somewhat lower at 28 °C. The sourdough FAQ has further temperature recommendations which are illustarted in the graph below (data from sourdough FAQ). Holding this together with the notion (see for instance p. 272 in “The handbook of dough fermentations”) that lower temperatures (20-25 °C) favor acetic acid production (= stronger flavor) and higher temperature (> 32 °C) favors lactic acid production (= milder flavor) it immediately becomes clear why a starter kept at roomtemperature has little yeast activity and smells of acetic acid. It need not be ruined, but is desperately in need of dilution, aeration and higher temperature.

Effect of temperature on growth of lactic acid bacteria and yeast based on data for optimum growth and no growth from the sourdough FAQ

I’ve kept the starter alive since August and baked with it at least once a week. I make sure that I keep 50-100 g which I store in the fridge. I’ve also frozen a sample just in case. I’ve changed from 100% hydration to 67% hydration, as this simplifies the calculations a little. On the evening before baking day I feed the starter to a total starter weight of approximately 900 g. The next day I bake bread as follows (the exact numbers were calculated using an internet sour dough calculator with the following input: 3200 g total dough weight, 25% starter, 67% hydration of starter and final dough and 1.8% salt):

Sour dough bread
949 g water
1417 g flour *
800 g starter (67% hydration, 25% of total dough)
34 g salt

* for instance 300 g rye whole grain, 200 g rye fine, 300 g whole grain wheat and then plain all-purpose wheat flour up to 1417 g.

Mix water and flour mixed until all flour is wetted. Leave for 15-20 minutes (during which the autolyse proceeds – this eases subsequent mixing). Mix (see more comments below regarding method/machine for this) until dough is smooth, and while mixer is running add sour dough starter. Once the starter has been properly incorporated into the dough, add the salt. Cover and leave to rise until volume has increased 30-100% (I know – this is not very accurate…). I have left it in on my bench top, but while this worked well on warm August days, it seems to be less than ideal on colder October days. Leaving the dough to rise on top of the fridge might be a conventient compromise here as my current waterbath is not large enough to hold the mixing bowl with dough. The reason I use a starter with a 67% hydration is that I can be more sloppy when adding the starter to the sourdough as it will not change the hydration of the dough. After proofing and slashing I bake the breads on a baking stone which is preheated to 250 °C. Right beneath the baking stone I have a small oven proof dish that I fill with boiling water. This helps to moisten the air in the oven and it simulates the steam injection port of professional baking ovens. There are several reasons why this is important. Moist air is a better heat conductor than dry air, and it prevents the surface from drying out too early while baking, resulting in a better oven spring. Furthermore the moist air condenses on the cold surface of the dough which improves gelatinization of the starch. This in turn gives better crust formation. After 10 min at 250 °C I open the oven to let the moisture out, take out the dish with water, turn the heat down to 220 °C and close the oven door. I bake the breads to a core temperature of about 93-95 °C.

Here I bake two 800 g loaves on a baking stone. Notice the dish with water for steam generation.

I should comment on mixing. When kneading by hand I’ve had a tendency to add to much flour. In fact I think this is one of the reasons why I quit baking bread several years ago – I found that the breads I made generally were a little to dry with a poor crumb, and at that time I didn’t really sit down and think about these matters. In retrospect however there’s no doubt that baking bread and adding flour ad lib until the dough feels good to touch is NOT recommended. Well anayway not unless you stop adding flour while the dough is still quite sticky. The thing about doughs with a high percentage rye is that they are quite sticky, and they should be. This is the best argument you’ll ever get for buying a kitchen gadget: bread doughs are too sticky to be kneaded by hand! There you have it! I’ve settled with the Assistent from Sweden. It was formerly sold under the Electrolux brand, but is now marketed indepentendly, yet it is still produced at the very same factory as always. In the US the machine is known as the Magic Mill. It has a huge 7 L bowl that rotates. The roller is attached to a flexible arm, so if the dough is to hard to work the arm just moves to the middle of the bowl. This significantly reduces the chances of overheating the motor. There are of course other alternatives from Kitchen Aid, Hamilton Beach and Viking Range which seem robust, but I have no experience with these. However, I doubt that the average Kenwood can cope with more than 3 kg of bread dough (but please correct me if I’m wrong ).

Unlike most other machines, the bowl of the Magic Mill/Assitent rotates while the roller pushes the dough to the sides of the bowl

Sources for further reading

I’ve read quite a bit about sourdoughs, and what I’ve been looking for a simple correlation between temperature, hydration and fermentation time. An excellent source of information with lots of practical advice is the FAQ from rec.food.sourdough.

I think the best resource I’ve found sofar is Lorenz and Bruemmer’s chapter “Preferments and Sourdoughs for German Breads” and Teija-Tuula Valjakka, Heikki Kerojoki and Kati Katina’s chapter “Sourdough Bread in Finland and Eastern Europe” in “The handbook of dough fermentations”. I will have to study these more carefuly.

There are quite a number of academic publications which also touch upon the effect of temperature on acid development. Here are some snippets:

“Controlled production of acetic acid in wheat sour doughs”:

…temperature has no significant effect, and that fructose is more efficient in influencing the FQ than dough yield.

FQ = fermentation quotient = lactic acid / acetic acid

“Volatile compound and organic acid productions by mixed wheat sour dough starters: Influence of fermentation parameters and dynamics during baking”:

… Low temperature (25 degrees C) and sour dough firmness (dough yield 135) were appropriate for LAB souring activities but limited yeast metabolism. Raising the temperature to 30 degrees C and semi-fluid sour doughs gave more complete volatile profiles …

“Sourdough: a tool for the improved flavour, texture and shelf-life of wheat bread” (Ph.D. thesis of Kati Katina):

The production of acids depends also on other things such as fermentation temperature, time and dough yield. Optimum temperatures for the growth of lactobacilli are 30-40 °C depending on strain (Stanier et al. 1987) and for yeasts 25-27 °C. In general, a higher temperature, a higher water content of sourdough and the utilisation of wholemeal flour enhances the production of acids in wheat sourdoughs (Brummer and Lorenz 1991, Lorenz and Brummer 2003).

I must admit that I’m a little confused as some of these snippets seem to contradict. It might be that I’m overlooking something important though and that I’m taking results out of their context. Any insight from my readers on this will be greatly appreciated!

Starter details
Here’s the details from my notebook on how I fed my sourdough starter. As I mentioned above, I wonder if discarding dough in the process actually does make sense after all.

• July 29, evening: 50 g rye + 50 g water
• July 30, morning: 25 g rye + 25 g water, evening: small bubbles (!), 50 g rye + 50 g water
• July 31, morning: 50 g rye + 50 g water, evening: pH measured to 3-4 with strips, fed with 50 g rye + 50 g water
• August 1: morning: 50 g rye + 50 g water, tested for bread baking, result: not active enough, feeding continued evening: 50 g rye + 50 g water
• August 2: morning: 50 g rye + 50 g water, evening 50 g rye + 50 g water
• August 3: morning: 50 g rye + 50 g water, evening: no bubbles, discarded all except ~100 g, fed with 50 g rye + 50 g water
• August 4: morning: 50 g rye + 50 g water, evening 50 g rye + 50 g water
• August 5: morning: 50 g rye + 50 g water, evening: big bubbles, the starter is active, CO2 production evidenced by tickling in nose, fed with 60 g rye + 60 g water and 2 x 100 g samples taken for fridge and freezer as fallback points
• August 6: morning: 50 g rye + 50 g water, first successfull bread made with the starter

I show my raw data to illustrate that it’s not straightforward, even with temperature control.

A fool proof starter – is it possible?
What I’m hoping to achieve can be summarized as follows: A “fool proof”, robust and quick method to obtain a sourdough starter that’s as simple as possible, using only flour and water (possibly with addition of some fruit) without having to waste anything of the starter. Temperature is maintained using a thermostated water bath. Preferably it should be possible to adjust the fermentation quotient (ratio of lactic acid/acetic acid) and the total titrable acid content by means of temperature, time and hydration/dough yield.