22 août 2025

Homemade spice blends: curry, dukkah, ras el-hanout

Grinding your cumin seed the same day, rather than opening a jar bought six months ago, genuinely changes how a curry tastes. Making the blend yourself, on the other hand, does not make it safer or purer than a shop-bought one, and the food-safety argument you hear everywhere on this subject is wrong in both directions.

The good reasons to blend your own come down to the freshness of volatile compounds, control over salt, and the ability to set the chilli level yourself. Those are enough. They simply have nothing to do with additives.

What blending at home changes, and what it does not

Start by clearing away the most common claim: a homemade blend carries no regulatory or food-safety advantage over a bought one.

European Regulation 2023/915 on contaminants sets maximum levels for dried spices of 5.0 µg/kg for aflatoxin B1 and 10.0 µg/kg for total aflatoxins, covering paprika and chilli, pepper, nutmeg and turmeric. Dried ginger falls under the same values. Ochratoxin A is capped at 15 µg/kg for dried spices and 20 µg/kg for Capsicum. In Switzerland, the federal ordinance on maximum contaminant levels carries exactly the same figures: 5 µg/kg for aflatoxin B1 and 10 µg/kg for the total.

The decisive detail sits in a footnote to both texts: the maximum level applies equally to mixtures of dried spices containing one or more of those spices. A homemade curry powder is a spice mixture in the regulatory sense, exactly like a sachet from a shelf. You escape nothing by blending it yourself, and you inherit no extra control either.

The same logic covers bacterial contamination. A survey by the US food agency, covering 7250 retail spice samples, estimated Salmonella prevalence at 0.19 % for curry powder, 0.24 % for black pepper, 0.56 % for ground coriander and 0.64 % for red pepper. No positive samples among cumin seed, sesame seed or white pepper. Imported shipments analysed before entry, by contrast, showed prevalences of 1.7 to 18 %. For nine of the eleven spice types studied, the gap between the imported shipment and the shelf was statistically significant; dehydrated garlic and basil were the exceptions.

That gap is instructive. Sorting, decontamination treatments and distribution cut the risk between import and retail. Buying loose from an open bin and blending at home reproduces none of those steps. This is not a cause for alarm, since retail prevalence stays below 1 %, but it does turn the usual argument on its head.

One documented case makes the point. Between December 2014 and July 2015, a salmonellosis outbreak affected 174 people in Sweden, 108 of them linked to a single restaurant. A spice mix containing dried vegetables used at that restaurant tested positive for the outbreak strain, as did similar mixes from different suppliers. The vehicle was a blend, and the problem came from the dried vegetables inside it.

What blending at home does change comes down to three verifiable things: you choose when the grinding happens, you set the salt, and you know which ingredients you added when an allergy is involved. It does not rule out cross-contamination that occurred before purchase. An anti-caking agent will appear in the ingredient list; its presence tells you nothing about the age or aroma quality of the spices. Our page on reading a food label explains how to spot it.

Grinding: where the difference actually lives

A spice’s aroma sits in volatile compounds locked inside the seed structure. Grinding releases them, which is the point, but it also opens the door to evaporation and oxidation.

The scale of that loss has been measured. A study in the Journal of Food Science and Technology in 2016 compared ordinary grinding with cryogenic grinding, carried out at very low temperature to avoid heating, across two cumin varieties. Cryogenic grinding raised volatile oil recovery by 33.9 % in one variety and 43.5 % in the other. Put differently, between a third and nearly half of cumin’s volatiles are lost to the heat of ordinary milling. Cuminaldehyde, the compound that defines cumin, rose from 48.2 to 56.1 % of the oil in one variety, and the sensory panel rated the aroma more pleasant.

Nobody is going to buy a cryogenic mill for a home kitchen. But the finding points the way: the less the grinder heats, the more aroma survives. In practice, pulse in short bursts rather than running continuously, grind small quantities, and let the machine cool between batches. A mortar, slower but cold, does not have this problem at all.

It also explains why shop-bought ground spice never matches the same spice ground to order, regardless of its original quality. The loss is not in the jar, it happened before the jar was filled. The same reasoning applies to a basic pantry of essential spices and herbs: a few jars replaced often beat a wall of jars opened years ago.

Toasting: what it adds, and when it turns against you

Heating seeds dry before grinding triggers Maillard reactions that create compounds absent from the raw seed: pyrazines with roasted notes, caramel furanones, smoky phenols.

On sesame, these compounds have been characterised precisely. A study in Food Chemistry: X in 2024, covering sesame hulls roasted at three temperatures, identified 140 aroma-active compounds, 36 of them with an odour activity value above 1. At 180 °C, caramel furanone, fruity 3-methylbutanal and smoky 2-methoxy-4-vinylphenol dominate the profile. At 200 °C, the sweet and roasted notes intensify.

Toasting carries a cost, though, and it shows up in storage. A study in Food Chemistry in 2026, on white and black sesame seeds roasted between 160 and 200 °C then held in accelerated storage, found that above 160 °C the acid value and peroxide value were highest after twenty days, while tocopherol content dropped significantly. Pyrazines and phenols, the two most abundant volatile families, fell markedly during storage. After fifteen days, the intensity of nutty, smoky and caramel aromas dropped significantly, replaced by a green note.

The practical conclusion is clean: a toasted dukkah is at its best on day one and degrades faster than a blend of raw seeds. It is a blend to make in small quantities and eat within weeks, not a jar for the year.

With nuts, moisture makes it worse. A study in the Journal of the Science of Food and Agriculture in 2022 followed raw almonds exposed to postharvest moisture then dried, against almonds never exposed. After twelve months of accelerated storage, the exposed almonds showed free fatty acid values 1.4 times higher and peroxide values 3.5 times higher, with more hexanal and octanal, both markers of rancidity. A trip through moisture, even followed by drying, shortens shelf life.

Curry powder, with actual quantities

Here, “curry” means a dry powder that keeps in a jar. It is not a summary of South Asian masalas, whose composition varies with the dish, region and household.

A balanced base blend, for about 55 g of finished powder:

20 g coriander seed
12 g cumin seed
8 g ground turmeric
4 g ground ginger
3 g fenugreek seed
3 g black peppercorns
2 g cayenne pepper, to adjust
2 g green cardamom seeds, husks removed
1 g ground cinnamon

Coriander and cumin form the base because they stay mild in volume and carry everything else. Turmeric supplies the colour and an earthy bitterness that takes over if you push the dose. Fenugreek brings a bitter maple-syrup edge and quickly dominates if you add more.

Method matters as much as proportions:

Dry-toast coriander, then cumin, then fenugreek, separately, over medium heat. Coriander needs about 2 minutes, cumin 1 minute 30, fenugreek 1 minute. They do not heat at the same rate, and toasting them together means burning the most fragile one.
Stop the moment the aroma rises and the seeds begin to crackle. Fenugreek turns bitter as soon as it darkens.
Tip immediately onto a cold plate. In a hot pan, cooking carries on.
Cool completely. Grinding warm seeds condenses moisture inside the mill and cakes the powder.
Grind in pulses, then stir in the powders that were not toasted: turmeric, ginger, cinnamon, chilli.

Turmeric does not go into the dry pan with the seeds. Its curcumin degrades under dry heat and it burns before the others. On what culinary doses actually deliver, our page on turmeric and ginger covers the evidence.

In cooking, this powder blooms in hot fat at the start of the dish, never dusted on at the end. Our guide to measuring spices in a curry handles quantities per serving and how to rescue an overdose.

Dukkah, and the detail that changes its nutrition

Dukkah is an Egyptian condiment built on nuts and seeds, coarsely crushed. Its name comes from the Arabic verb meaning to pound. Unlike curry powder, it is not ground fine: the crunch is the whole point.

A base for 189 g of blend:

100 g whole hazelnuts
60 g unhulled sesame seeds
15 g coriander seed
10 g cumin seed
4 g salt

The choice of sesame deserves particular attention. In the Swiss food composition database, unhulled sesame contains 930 mg of calcium per 100 g, against 37 mg for hulled sesame. That is a ratio of twenty-five to one. The words “hulled” or “unhulled” on the packet matter far more than the seed colour.

Across the 60 g in this recipe, that means 558 mg of calcium with whole sesame against 22 mg with hulled. These are amounts in the food, not a measure of intestinal absorption: the hull also contains phytates, which inhibit mineral absorption. The compositional gap remains large. Our page on zinc, selenium and iodine in veggie cooking explains the same distinction for other minerals.

Keep the orders of magnitude in mind, though. A realistic dukkah portion is 10 to 15 g. A conservative calculation from the Swiss database, counting the spices in the weight but not the nutrients because matching entries are unavailable, gives roughly 82 kcal, 7.3 g of fat, 2.4 g of protein, 1.4 g of fibre, 57 mg of calcium, 0.65 mg of iron and 28 mg of magnesium per 15 g. The calcium stays modest across a day. Dukkah is a condiment, not a main nutrient source.

Salt deserves the same arithmetic. With 4 g of salt in the blend, a 15 g portion supplies about 0.31 g of salt equivalent, or 6 % of the 5 g daily ceiling recommended by the World Health Organization. Two portions a day, in a diet that already includes bread and cheese, add up. Our page on seasoning without salt works through that trade-off.

The method:

Toast the hazelnuts in the oven, 8 to 10 minutes at 180 °C, until the skins crack. Rub them in a tea towel to remove the skins, which are bitter.
Toast the sesame in a pan over medium heat, 2 to 3 minutes, stirring constantly. The seeds jump when they are ready.
Toast coriander and cumin separately, as for the curry powder.
Cool everything completely. Crushing warm hazelnuts yields paste, not grit.
Crush the hazelnuts in a mortar or with very short pulses in a processor. Stop when the largest pieces are about the size of a grain of rice.
Grind the coriander and cumin coarsely, then combine with the whole sesame and the salt.

The sesame stays whole. It carries the texture, and grinding it turns dukkah into dry tahini. On bread, in a grain bowl or over roasted vegetables it plays the same role as pumpkin, sunflower and sesame seeds eaten on their own, with crunch added.

Ras el-hanout, and the problem with its definition

Ras el-hanout means “head of the shop” in Arabic: the best of what the grocer has. Every grocer has a formula, which means there is no reference recipe. Commercial versions run from eight to more than thirty ingredients, and some contain items you will not find in European retail.

A coherent domestic version, for about 45 g:

8 g coriander seed
6 g cumin seed
5 g ground ginger
5 g sweet paprika
4 g ground cinnamon
4 g turmeric
3 g black pepper
2 g green cardamom seeds, husks removed
2 g dried rose buds, crumbled
1.5 g grated nutmeg
1.5 g cloves
1 to 3 g chilli, to taste

Cloves and cinnamon are the two ingredients that throw a ras el-hanout off balance fastest, so this recipe keeps cloves to 1.5 g. On cinnamon, note that two very different species exist: cassia cinnamon, sold by default in Europe, and Ceylon cinnamon. They differ by roughly a hundredfold in coumarin content, a substance that affects the liver at high and prolonged doses. For a blend used occasionally, the question does not arise. For daily use, it is worth a glance at the label.

The method follows the curry powder: separate toasting of whole seeds, cooling, grinding, then folding in the powders. Rose buds are crumbled by hand at the end, never through the mill, so visible fragments survive.

A word on substitution, since it is the most frequent question. Without ras el-hanout, a mix of cumin, coriander, cinnamon, ginger and paprika in these proportions lands close. Without cardamom, nothing really replaces it; better to leave it out than to swap in something else. Without rose buds, you lose the floral note without the blend turning bad. Our page on Middle Eastern spices covers other equivalences of this kind.

Storage: the real timelines

The usual advice says six months to a year without distinguishing between blends. That is too coarse, because the three blends on this page do not age the same way at all.

A curry powder or ras el-hanout, built from dry spices with no meaningful fat, lose their volatiles gradually without turning bad. Between six months and a year they are flat but edible. You then increase the dose, which unbalances the profile, because not every compound evaporates at the same rate.

Dukkah is a different product. It contains hazelnuts and sesame, therefore unsaturated lipids, and it has been toasted. The sesame data show pyrazines and phenols falling significantly within weeks of accelerated storage, with a green note appearing. The almond data show oxidation progressing over months. In practice, make dukkah in small quantities and eat it within three to four weeks. In the fridge, in a tightly closed jar, you buy time.

One factor gets overlooked: moisture at the point of jarring. A study on ground paprika compared samples at different water contents before storage, in a room at 19 to 24 °C and 24 to 42 % relative humidity. After four months, samples starting at 6 % moisture had lost 92 % of their extractable colour and those at 9 % had lost 90 %, against 53 % and 32 % for samples starting at 15 % and 18 %. The direction of that gap is surprising until you read the detail beside it: the water contents equalised over the four months, every sample finishing near 10 %. The driest samples therefore took up moisture from the surrounding air, while the wettest gave some off. What the study shows is not that damp paprika keeps better, but that the paprika which gained water during storage is the one that lost the most colour. The practical consequence lands on the container rather than the spice: a genuinely airtight jar prevents that exchange with the room. The same care applies to preserving fresh herbs.

In practice, three rules cover it. Wait until everything is at room temperature before closing the jar, or steam will condense inside. Take spice out with a dry spoon, never over a steaming pan. Store away from light and well clear of the oven, because heat accelerates every form of degradation described above.

Common mistakes

Toasting all the seeds together. Coriander, cumin and fenugreek differ in size and oil content, and the smallest burns before the largest is ready.

Grinding seeds while still warm. Residual moisture sticks to the powder and cakes it, and the grinder’s own heat adds its share of volatile loss.

Making a large jar of dukkah. The most fragile blend of the three is exactly the one people make in bulk, out of condiment habit.

Buying hulled sesame for dukkah. Twenty-five times less calcium, for a cosmetic difference.

Assuming homemade means safer. Maximum contaminant levels apply to mixtures just as to single spices, and where you buy matters more than where you assemble.

Adding turmeric to the toasting pan. It burns before the seeds and turns bitter.

Salting the curry powder or ras el-hanout. An unsalted blend can be dosed freely in the dish; a salted one imposes its salt along with its spice.

By profile

For someone cooking spiced food two or three times a week, 50 g each of curry powder and ras el-hanout covers one to two months. Beyond that you are storing aroma loss.

For someone watching sodium, make all three blends unsalted and salt the dish instead. That moves the decision to the moment of cooking, where it belongs.

For a household with a tree nut allergy, dukkah works with toasted pumpkin and sunflower seeds. The texture shifts slightly, the principle holds.

For someone without a spice grinder, a mortar handles coriander, cumin and fenugreek. Pepper and cardamom demand patience. Turmeric and ginger are bought ground anyway. Our page on the minimalist veggie kitchen places the grinder among the purchases that earn their shelf space.

For a chilli-free kitchen, simply drop the cayenne. Sweet paprika keeps the colour without the heat, and black pepper brings a different, shorter sharpness.

What this evidence does not say

The cryogenic grinding measurements cover two cumin varieties. The principle carries over to other aromatic seeds, but the exact size of the gain has not been measured for each.

The toasted sesame studies use accelerated ageing, meaning storage under unfavourable conditions designed to simulate a longer period. Real timelines in a kitchen run longer than the twenty days of the protocol.

The Swiss database values are generic averages. The calcium content of any given sesame depends on its origin and how much hull it retains.

The Salmonella prevalence figures come from American sampling. European supply chains are not identical, but these are the only figures available at that scale.

The paprika study covers two cultivars, over four months, in a single ambient humidity regime.

Frequently asked questions

Do the seeds really need toasting? For cumin, coriander and sesame the difference is clear and comes from Maillard compounds absent in the raw seed. For powders, no: they burn.

Can I toast in the oven instead of a pan? For nuts, yes, and it is more even. For small seeds, the pan lets you see and smell the moment to stop, which matters more than temperature precision.

How long does homemade curry powder keep? Six months to a year with no safety issue, but with aroma dropping off from the second or third month. Trust your nose when you open the jar.

Is a homemade blend cheaper? It depends on the price of whole spices and the quantity involved. Loose and in bulk, yes. For three small packets bought specially, often not.

Will a coffee grinder do? Yes, provided it is dedicated to spices. Coffee leaves its aromas in the grinding chamber, and cumin returns the favour.

Should I sieve after grinding? For a curry powder destined for a smooth sauce, yes, since it removes coriander hull fragments. For a tagine ras el-hanout, it is unnecessary.

Does the blend need to rest before use? Nothing requires it. Ground spices do not ferment or bond chemically in a dry jar. A blend is at its best straight after grinding. That applies to gifted blends too: our page on homemade gourmet gifts suggests dating the jars for this reason.

Can dukkah be frozen? Yes, and it slows lipid oxidation. Take it out at serving time and close it again quickly, so condensation does not soften the crunch.