Baking powder

Update 2009-03-01: I found a LOT more reference materials, and it's just much cleaner to integrate them into the original post than to write them out as addenda.

In my previous post I had a brief discussion about homemade baking powder. I think, though, that it will be useful to devote a separate post to baking powder in general and expand on the topic a bit.

CONTENTS

• Baking Powder and Leavening
  
• Shelf Life
  
• Single-Acting vs. Double-Acting
  
• Aluminum-Free Baking Powders
  
• Rumford Baking Powder
  
• Do Brands Matter?
  
• Homemade Baking Powder
  
• Substitutions
  
• How Much Should You Use?
  
• High Altitude Baking
  
• Sodium-Free Baking Powders
  
• References and Notes

BAKING POWDER AND LEAVENING

Baking powder belongs to a class of ingredients known as leaveners, the purpose of which is to produce adequate gas pockets (bubbles, if you will) so that the dough or batter will rise and lighten the texture of baked goods. Further, baking powder and baking soda are known as chemical leaveners, which release gas through a reaction between alkaline and acidic ingredients. Other leaveners used in home baking include yeast, which acts biologically, and egg whites, which leaven mechanically by the air pockets formed from whisking. Once in the oven, steam also contributes to the leavening of baked goods.

Baking powder is made by combining baking soda with one or more dried acids. Baking soda, which is also known as either bicarbonate of soda or sodium bicarbonate (formula NaHCO₃), is an alkaline (or base) material and will release carbon dioxide gas (CO₂) when dissolved with the help of an acid, so it is used in recipes that contain ingredients such as buttermilk, natural cocoa powder, or fruit juices. When we combine baking soda with dried acid(s), we create a complete leavening system that is then activated (i.e., gives off gas) by introducing moisture into the mix.

Note, however, that baking soda will release carbon dioxide not only as it dissolves with acid, it will also break down and release gas at temperatures above 120 °C (250 °F) without any acids involved¹. In other words, carbon dioxide is not produced from a combination of acid and baking soda, but is released as baking soda itself breaks down. Baking soda is the carbon dioxide source², and the acid-soda reaction is really more accurately described as acid-activated decomposition of baking soda. The generalized equation for the breakdown of baking soda with acid (denoted here as H⁺) is:

NaHCO₃ + H⁺ → Na⁺ + CO₂ + H₂O

The thermal decomposition of baking soda is:

2 NaHCO₃ + heat → Na₂CO₃ + CO₂ + H₂O³

SHELF LIFE

Since moisture is what triggers baking powder into action, it must be kept dry during storage. To prolong shelf life, a starch (such as cornstarch) is added to commercially-made baking powder to absorb moisture and prevent its alkaline and acidic components from reacting with each other prematurely. This moisture-absorbing capacity isn't infinite, of course, so manufacturers will indicate an expiration date on the product's package. To be on the safe side, you should throw out any baking powder that has expired. Oh, foods made with out-of-date baking powder aren't harmful per se, but the shame from turning your cakes and breads into bricks might just kill you (or at least make you wish you were dead).

If, on the other hand, you like to live on the edge or just can't find the expiration date, you can test baking powder's effectiveness by dropping a small amount of it into just-boiled water: if it foams vigorously, then the powder is still good; if it sinks to the bottom, then the powder has passed its prime.

Testing baking powder (verdict: still good)

How much baking powder and water should you use for this test? Considering the large discrepancies in published guidelines—for example, Cook's Illustrated⁴ recommends using 1/2 tsp powder in 1 cup water, while Fine Cooking recommends 1 tsp powder in 1/3 cup water—I think it's safe to conclude that exact quantities are not crucial.

SINGLE-ACTING VS. DOUBLE-ACTING

Baking powder is typically divided into two categories: single-acting and double-acting. Single-acting baking powder, however, can be either fast-acting or slow-acting⁵. As the name implies, single-acting powders undergo one chemical reaction: the acid in a fast-acting powder dissolves relatively quickly so that the gas-producing reaction takes place upon mixing with liquid; on the other hand, the acid in a slow-acting powder requires either prolonged time exposure or temperatures above 120 °F⁶ in the batter to dissolve. Double-acting baking powders contain both fast- and slow-acting acids so that gas pockets can be formed in the dough or batter both as it is mixed and after it is in the oven; allowing two reactions to take place increases baking powder's reliability as it makes the length of time between mixing and baking less critical.

Some of the commonly-used acids in baking powder and their reaction times are listed below⁷:

Leavening acid	Abbrev.	Formula	Reaction time
Cream of tartar		KC4H5O6	Immediate
Glucono delta lactone (also called δ-gluconolactone or D-glucono-1,5-lactone)	GDL	C6H10O6	Slow at room temperature
Monocalcium phosphate (also called calcium acid phosphate)	MCP	Ca(H2PO4)2 [or · H2O]	Immediate (but see Rumford discussion below)
Sodium aluminum pyrophosphate		NaAlP2O78	Slow at room temperature
Sodium aluminum sulfate	SAS	NaAl(SO4)2	Slow and heat-activated
Sodium aluminum phosphate	SALP 1-3-8	NaAl3H14(PO4)8 · 4 H2O	Heat-activated
Sodium aluminum phosphate	SALP 3-2-8	Na3Al2H15(PO4)8 [anhydrous]	Heat-activated
Sodium acid pyrophosphate	SAPP9	Na2H2P2O7	Variable

Variable reaction time? How can that be?

Well, it turns out that a leavening acid's molecular formula isn't the only determinant in its reaction speed. Manufacturing-related parameters such as heat treatment, particle size, or coating materials can all be manipulated to affect the specific acid's behavior without altering the chemical make-up of its main constituent. Specifically, SAPP is manufactured in multiple grades, each with a different reaction rate^{10, 11}.

ALUMINUM-FREE BAKING POWDERS

Most of the baking powders sold to US consumers today are labeled as double-acting, and market leaders Calumet and Clabber Girl both use SAS as the slow-acting acid. I'm not a medical scientist so I can't tell you whether the health concerns over aluminum are valid or not, but I do know that its metallic flavor is unpleasant to some, so it's worth your time to seek out aluminum-free baking powders. These are the products I'm aware of, in alphabetical order (comments are "to the best of my knowledge"):

CaCO₃ = calcium carbonate. MgCO₃ = magnesium carbonate. KHCO₃ = potassium bicarbonate.

Brand	Ingredients	Sodium Free?	Comments
Argo	Acid 1: MCP Acid 2: SAPP Base: baking soda Starch: cornstarch	No	Smallest container is 12 oz.
Bakewell Cream	Acid 1: SAPP Acid 2: Base: baking soda Starch: redried starch12	No	Mainly sold in New England. Packaging does not look airtight.
Bob's Red Mill	Acid 1: MCP Acid 2: SAPP Base: baking soda Starch: cornstarch	No	Mainly found in specialty-foods stores. Packaging does not look airtight. Smallest container is 16 oz.
Ener-G	Acid 1: citric acid Acid 2: GDL Base: CaCO3 and MgCO3 Starch:	Yes	Mainly found in specialty-foods stores. Kosher for Passover. Smallest container is 7.05 oz (200 g).
Hain	Acid 1: MCP Acid 2: Base: KHCO3 Starch: potato starch	Yes	Mainly found in specialty-foods stores. Kosher for Passover. Not labeled as double-acting.
Rumford	Acid 1: MCP Acid 2: Base: baking soda Starch: cornstarch	No	Best availability nationwide. Made by Clabber Girl. Smallest container is 4 oz.

As mentioned above, baking powder has a limited shelf life, so I would recommend getting the smallest and best-sealed package available.

Ener-G's formulation is unusual for consumer-oriented baking powders, using citric acid and GDL, two bases, and no starch. Glucono delta lactone is derived from glucose and is, in spite of the impression its name may give, dairy free. The calcium and magnesium carbonates provide alkalinity (calcium carbonate is the active ingredient in many antacid tablets), and magnesium carbonate further functions as a drying agent so that no starch is needed in this baking powder. The company recommends using twice as much of its baking powder as regular double-acting powder to achieve the same leavening effect.

Of the single-acid powders, only Hain does not label its product double-acting. Since the reaction rate of SAPP can be made to vary, it's possible that Bakewell Cream contains a mixture of multiple SAPP grades so that the powder can provide leavening action at different temperatures. But what about Rumford? Isn't MCP only a fast-reacting acid? Well...

RUMFORD BAKING POWDER

Yeah, I know, giving this brand its own heading seems like so much product placement, but among the lay food-science types, there is some confusion over how this baking powder really works, so I think we need to talk about it.

Rumford lists only three ingredients for its baking powder: calcium acid phosphate, bicarbonate of soda, and cornstarch. Notice from the top table that calcium acid phosphate is just another name for MCP, which is classified as a quick-acting acid. However, the product canister is also marked "double acting." How can that be possible when there is only one leavening acid present?

Well, not all MCPs are created equal. For starters, it's available in two forms: with a water molecule attached to it (monocalcium phosphate monohydrate, Ca(H₂PO₄)₂ · H₂O), or without water (anhydrous monocalcium phosphate, Ca(H₂PO₄)₂). It may be counterintuitive, but MCP monohydrate is a granular powder¹³ even with the water hanging around—that makes as much sense, though, as baking powder releasing a gas upon decomposing while it normally exists as a powder. MCP monohydrate is indeed a quick-reacting acid; on the other hand, anhydrous MCP is typically manufactured with a coating to protect the acid from exposure to liquids and thus delay its reaction. The following graph shows the slower room-temperature gas-release rate that results from dissolving baking powder with coated anhydrous MCP as compared to MCP monohydrate¹⁴:

In the general case, baking soda reactions with coated anhydrous MCP will release 15% of the available CO₂ during the first 2 minutes, 35% during the next 10–15 minutes, and the final 50% during baking¹⁵. Specifically, we can compare the gas-release rates of several different baking powders made by Clabber Girl:

Product (PDF datasheets)16	Leavening acids	Approx. CO2 release (mix/bake)
Rumford	MCP	60%/40%
Clabber Girl SAS	MCP, SAS	30%/70%
Clabber Girl SALP	MCP, SALP 3-2-8	30%/70%
Fleischmann's	MCP, SAPP	10%/90%

I believe the SALP and SAPP formulations are available only to commercial bakers, which is why Fleischmann's is not listed in the aluminum-free baking powder table above.

At any rate, based on this knowledge I think we can make the following conjectures on how Rumford can label itself as "double acting":

• Even though coated anhydrous MCP has a faster room-temperature reaction rate than aluminum-containing leavening acids (compare against SALP in the graph above), it may be delayed enough so that baking powders containing only coated anhydrous MCP can approximate the performance of double-acting powders without actually having a two-stage gas-release characteristic. Since the coating material is likely to show up under analysis only as trace elements, Rumford may not be required to specify this material on the ingredients list.
  
• It's also possible that Rumford contains both uncoated and coated MCP so that there does exist two spikes of gas production. However, as the main constituents of both MCP types are identical, they may then be shown in the ingredients list only once.
  
• Further, in Bakery Technology and Engineering, the author states that "[d]ouble-acting baking powders are really a version of the slow-acting type which exhibit somewhat more gas production potential during mixing and on the bench."¹⁷ Based on this, we should be able to apply the same description in reverse to MCP-only baking powders such as Rumford, as it uses a fast-acting acid that exhibits delayed gas production behavior.
  
• Finally, "double acting" may be a phrase with no regulatory definition, at least in the US. If this is the casee, then Rumford may well declare, on its own, that its baking powder is "close enough" to the two-acid powders in behavior for the "double acting" label to be suitable. In other words, they did it because they felt like it and could get away with it. I am, however, not skilled enough at searching through US food regulations to confirm or deny this conjecture (I've tried, really).

DO BRANDS MATTER?

The conclusion from Cook's Illustrated's March 2003 baking powder test was that all the powders in the review showed similar performance. This may be a result of the US FDA regulations that require all baking powders to release at least 12% CO₂¹⁸ coupled with the constraints imposed by the acid-base chemistry itself, leading to only small differences in the amount of gas released from each brand. In other words, aside from the concerns over aluminum, we should reasonably expect double-acting baking powders from the leading brands to be interchangeable with each other in recipes.

Update 2009-08-19: Baker Debra Wink at the excellent bread-making resource The Fresh Loaf disagrees, and has empirical evidence that different brands exhibit different behaviors.

HOMEMADE BAKING POWDER

Sift or mix the following ingredients thoroughly:

2 parts	Cream of tartar
1 part	Baking soda
1 part	Cornstarch (optional)

Each "part" is measured by volume (e.g., teaspoons, tablespoons, cups—just keep them all the same). A 2:1 ratio between cream of tartar and baking soda is the classic formula, although to be exact, the ratio should be 5:2¹⁹. As mentioned above, the cornstarch is for moisture absorption and is only necessary if you're going make a large batch of baking powder and store it for a while; otherwise just make the quantity needed for each recipe, skip the cornstarch, and use it right away (keep in mind that if you've skipped on the cornstarch, then you should use less of the resulting powder than the quantity specified in the recipe—see the Substitutions section below for details). If my understanding is correct, excluding cornstarch also makes this baking powder kosher for Passover.

Since cream of tartar is a quick-reacting acid, homemade baking powder is single-acting. To make sure the leavening action from this baking powder doesn't fizzle out prematurely, place the batter into the oven as soon as possible after it is mixed. For extra insurance against early fizzling, it's also been suggested that you should add the baking soda and cream of tartar into your baking mixture separately, which seems like a good idea.

SUBSTITUTIONS

To substitute for 1 tsp double-acting baking powder, follow one of the options below (based mainly on David Joachim's The Food Substitutions Bible²⁰):

• For homemade baking powder without cornstarch, either use ¾ tsp (i.e., ½ tsp cream of tartar + ¼ tsp baking soda) based on the classic formula, or 7/8 tsp (5/8 tsp cream of tartar + ¼ tsp baking soda) if your measuring tools are accurate enough. Place batter in oven as soon as possible.
  
• 1¼ tsp single-acting baking powder. Place batter in oven as soon as possible.
  
• ¼ tsp baking soda + ½ cup buttermilk. You will then need to reduce the other liquids in the recipe by ½ cup. This will alter the flavor of the recipe.
  
• ¼ tsp baking soda + ¼ cup molasses. You will then need to reduce the other liquids in the recipe by 3 tbsp and adjust the sweetener. This will alter the flavor of the recipe.

It seems a bit confusing that ¾ tsp of homemade baking powder and 1¼ tsp of of single-acting baking powder are both listed as appropriate substitutes for 1 tsp double-acting powder. I'm guessing that the 1¼ tsp quantity refers to commercially-made single-acting baking powder, which may have less leavening power or have been bulked up with non-reacting ingredients, but not having the author's resources, I really can't tell. Based on the suggestions for homemade powder without cornstarch, though, you should be able to use the same amount of homemade pre-mixed (that is, containing cornstarch) baking powder to substitute for commercial double-acting powder.

HOW MUCH SHOULD YOU USE?

You mean, other than following the recipe? Okay, I won't be so flip about it. As I'm sure you can figure, too little baking powder will mean too little rise and a dense result in your baked goods, which is undesirable. But does this mean more is better? Well, only to a degree. Too much baking powder creates supersized bubbles that burst and deflate—imagine blowing up bubblegum so much that it pops and gets all over your face—which, again, leads to dense baked goods with an added helping of nasty taste from the extra powder. So, then, what is the magic amount?

It's actually a little complicated. Shirley Corriher explains²¹:

The general rule is 1 to 1¼ teaspoons of baking powder per cup of flour in a recipe....[However,] many variables determine the proper amount of baking powder to be used, and you need to adjust the general rules for your own cooking conditions. More leavening can be used if the recipe calls for a lot of heavy ingredients like chopped fruits. When you change pan size, the amount of baking powder should be altered.

Corriher further notes that for the same amount of batter, the shallower it is in the pan (that is, with wider pans), the less powder you need. She also references Rose Levy Beranbaum's The Cake Bible for the exact amount of leavener to use relative to different pan sizes and batter quantities, but I don't have that book so I can't tell you more about it.

HIGH ALTITUDE BAKING

The rule of thumb is that general-purpose recipes should work well from sea level up to about 3,000 feet (910 m). Above that, the amount of baking powder as well as baking soda should be reduced. I have no first-hand experience with this, but the literature suggests that determining exact reductions is a somewhat complex issue. King Arthur Flour has some general guidelines, and the Colorado State University Extension makes slightly different recommendations specific to cakes while noting "[o]nly repeated experiments with each recipe can give the most successful proportions to use." In BakeWise, Shirley Corriher mentions that cookbook author Susan Purdy has done just that, making "all kinds of baked goods at sea level, 3,000, 5,000, 7,000, and 10,000 feet"²² (910 m, 1,520 m, 2,130 m, and 3,050 m), and published her results in Pie in the Sky. I have not seen this book myself, however.

SODIUM-FREE BAKING POWDERS

As noted in above (under aluminum-free baking powders), Ener-G and Hain are the two brands of sodium-free baking powder that I know of. Ener-G recommends using twice as much of its powder as regular baking powder to achieve the same leavening effect. Hain provide no such guidance, although some have tested it and found that it is one-for-one compatible—that is, use the same quantity—as regular powder. Other than these two specific instances, however, general guidelines vary (or are unreliable?). Most sources on the Internet recommend using 1.5 times as much sodium-free powder as standard baking powder. This advice, however, does not take into account the specific formulations of the different sodium-free powders. Although the greater atomic mass of potassium over sodium means that 19% more potassium bicarbonate (by weight) is needed to neutralize a given amount of acid as compared to baking soda²³, the ingredients listing doesn't tell us how much potassium bicarbonate is actually contained in the product that we buy. The Williams-Sonoma Kitchen Companion recommends doubling the quantity²⁴ but does not indicate the specific ingredients involved.

REFERENCES AND NOTES

1. W.P. Edwards, The Science of Bakery Products, Royal Society of Chemistry, 2007, p. 70
   
2. Lallemand Inc., Lallemand Baking Update (PDF), Vol. 1 No. 12, 1996
   
3. W.P. Edwards, The Science of Bakery Products, p. 71
   
4. "Baking Powder Expiry," Cook's Illustrated, 11/2002
   
5. Shirley O. Corriher, CookWise, William Morrow-Harper Collins, 1997, p. 73
   
6. Christopher Kimball, The Cook's Bible, Little, Brown, 1996, p. 348
   
7. Based on Harold McGee, On Food and Cooking (revised ed.), Scribner-Simon & Schuster, 2004, p. 533, with additions and changes from other sources.
   
8. I was unable to find a formula for sodium aluminum pyrophosphate—or, in fact, any mention of it in food industry literature of its use as a baking powder ingredient. Anyhow, given the charges of sodium (Na¹⁺), aluminum (Al³⁺), and the pyrophosphate compound (P₂O₇^4-), this formulation seems reasonable.
   
9. This item is not listed in McGee's table, and he uses SAPP as an abbreviation for sodium aluminum pyrophosphate. However, all the food industry references I've come across use SAPP as the shorthand for sodium acid pyrophosphate, so that's the convention I'm following here.
   
10. Samuel A. Matz, Bakery Technology and Engineering (3rd ed.), Springer, 1992, pp. 67, 69
    
11. Clyde E. Stauffer, Functional Additives for Bakery Foods, Springer, 1990, p.196




12. Based on the description of another product, I assume that "redried starch" is cornstarch.
    
13. George A. Burdock, Encyclopedia of Food and Color Additives, CRC Press, 1996, p. 409
    
14. Graph based on Robert C. Lindsay, Food Chemistry (3rd ed.), CRC Press, 1996, Owen R. Fennema, Ed., p. 772.
    
15. Shirley O. Corriher, BakeWise, Scribner-Simon & Schuster, 2008, p. 47
    
16. These datasheets are stored under two different sections of Clabber Girl's website (Foodservice and Ingredients), and it's a bit confusing because there are multiple datasheets, with different gas-release rates specified, for each product. I don't know why this is, but I've selected the most rounded-off figures for my table.
    
17. Samuel A. Matz, Bakery Technology and Engineering (3rd ed.), Springer, 1992, p. 71
    
18. Ibid.
    
19. Harold McGee, On Food and Cooking, p. 534
    
20. David Joachim, The Food Substitutions Bible, Robert Rose, 2005, p. 38
    
21. Shirley O. Corriher, CookWise, p. 138
    
22. Shirley O. Corriher, BakeWise, p. 50
    
23. W.P. Edwards, The Science of Bakery Products, p. 71
    
24. Mary Goodbody, Carolyn Miller, and Thy Tran, Williams-Sonoma Kitchen Companion, Oxmoor House, 2000, p. 23.