Baklava production: 40-layer phyllo, clarified butter, pistachio grades and sugar-syrup control

Why baklava is a benchmark skill

Baklava is where Turkish pastry judges a kitchen. Get the flour, the rolling, the fat and the syrup right and you produce a bright-golden lattice of dozens of separate, shattering leaves that melt on the tongue; get any one wrong and you get a pale, greasy or soggy slab. It is also commercially central: Turkey runs a very large confectionery sector — one trade estimate puts it at about US$3.5 billion and some 500,000 tonnes a year (a broad figure that varies across market reports), with demand spiking at Ramadan [c22]. For a UK Turkish patisserie, baklava is in practice the year-round flagship that peaks at Ramadan and bayram.

Gaziantep (Antep) is the acknowledged capital. "Antep Baklavası / Gaziantep Baklavası" carries a Turkish geographical indication held by the Gaziantep Chamber of Industry, and in 2013 it became the first Turkish product ever registered as an EU Protected Geographical Indication (PGI) — the application was published in the EU Official Journal on 8 August 2013 and registered that December [c1]. The GI defines baklava as layers of yufka filled with semolina cream (irmik kreması) and Antep pistachio, sweetened with syrup, made either dry (kuru) or fresh (taze) — the only difference being that kuru baklava omits the semolina cream. Antep pistachio is mandatory [c2]. That legal definition is a useful north star even for a baker outside Turkey: it tells you what "the real thing" contains.

See image img-b2bk-01 (finished fıstıklı baklava) and img-b2bk-08 (Gaziantep GI map).

1. The flour: baklavalık un (this is where authenticity starts)

Paper-thin yufka is only possible from a strong, extensible dough, so baklava flour sits at the high-protein, high-gluten-quality end of the wheat spectrum — the opposite end from the soft, low-protein flours used for kadayıf. Turkish millers build baklavalık un from hard-wheat blends (paçal) that are mostly imported hard wheats (Australian hard, American Dark Northern Spring, German- and Kazakh-origin) plus Southeast-Anatolian durum (the Zenit variety is cited as one example; mills also blend other Turkish durums), whose hard grain gives the baked baklava its crispness; a coarser flour particle also helps the pastry drink syrup [c3].

In numbers, the Turkish milling literature reports that millers prefer wheat with wet gluten above ~31-32%, gluten index above 70 and sedimentation above 35-36 mL, and that a premium baklava flour targets protein ≥13%, wet gluten ~35%, gluten index ~90, sedimentation ~38 mL and a maximum resistance to extension of 700-800 Brabender Units [c4]. (The finer of those specs — the ~90 gluten index and the Brabender figure — rest on that single academic review, so treat them as indicative rather than a hard standard.) The GI itself simply specifies sert buğday unu (hard-wheat flour) [c4]. This is a direct application of the craft concepts in A1-protein-gluten-and-strength and A1-key-quality-parameters — read the alveograph/farinograph and gluten-index sections there, and see the sister article B2-flour-and-milling for how Turkish grades (baklavalık, böreklik, yufkalık, kadayıflık) are separated on the mill diagram.

Sourcing on the platform. The strongest UK white flour on Domson with a published spec is Domson White Strong Flour (ADM 4380): it targets 12.0% protein, 59.5% water absorption, 14.2% moisture and is ascorbic-acid (E300) treated [c18]. That is an excellent bread flour but it sits below the ≥13% baklava benchmark, so for authentic yufka reach for the highest-gluten flour you can get — a very-strong Canadian flour (Centurion) or, ideally, a dedicated imported baklavalık un — and rely on gluten quality, not just protein number [c18]. See data.json → key_specs for the flour targets side by side.

2. The dough and the rolling: oklava, tekleme, çiftleme

The dough is deliberately firm (sert): flour, water, salt and egg, with the GI giving 2-4 eggs and 10 g rock salt per 1 kg flour [c5]. It is kneaded, rested ~15 minutes, divided into balls (beze), then rolled progressively thinner and repeatedly wound and unwound on the oklava (the long thin Turkish rolling pin) in the tekleme (single-wrap) and çiftleme (double) sequence until the sheet is almost translucent [c5]. Between the sheets the roller dusts wheat starch (buğday nişastası) — not flour — so the leaves slide, don't glue together, and stay separate and flaky in the oven [c6]. This is the same layering logic as European lamination (A6-laminated-dough-fundamentals, A8-laminated-dough-formulas) but achieved by stacking individually rolled sheets rather than folding a butter block; the shared Turkish/Arab phyllo craft is covered in B2-borek-phyllo and B3-phyllo-kataifi-production.

The scale of the skill (single-source master-workshop figures, treat as illustrative): from about 800 g of firm dough a master is reported to yield ~4 kg of finished baklava, to roll out one tray's dough in roughly 8 minutes, and to complete a batch in about 1.5 hours [c19]. Baklava yufka is rolled far thinner than börek yufka — the Turkish börek-yufka standard (TS 10443) is cited as permitting up to 1.2 mm, while baklava sheets are rolled close to translucent [c19]. See img-b2bk-04 (hand-rolling on the oklava with starch).

3. The build: ~40 layers around a pistachio-and-cream middle

A full Antep baklava is assembled to about 40 layers of yufka, of which roughly 9 are the top "yüzlük" sheets [c7]. In the GI kaymaklı construction the master lays about 15-20 base sheets, spreads the middle, then lays about 15-20 top sheets [c7]. The middle, per tray, is roughly 450-550 g kaymak (here the Antep semolina cream, irmik kaymağı) topped with 400-450 g Antep pistachio; then about 600-650 g of melted clarified butter is poured over hot [c8]. By weight the finished product runs roughly Antep pistachio 10-11%, cream 12-13%, clarified butter 15-16%, syrup 35-36% and flour 25% [c8]. Cut the tray into diamonds, squares or fingers before baking so the syrup can penetrate every piece. See img-b2bk-02 (layer cross-section).

Pistachio grades matter more than almost anything else. The master's choice for the filling is boz fıstık (boz iç fıstık): an early-harvest, not-fully-ripe Antep pistachio, vivid green, fresh-tasting and high in oil, finely ground [c11]. Barak yıldızı (a darker green) is used mainly to top and decorate the tray for colour, and raw (çiğ) Antep pistachio suits dry (kuru) baklava [c11]. Antep fıstığı itself carries its own GI; its intense green and aroma are the single biggest quality lever in the finished sweet [c1][c11]. For the nut-handling craft — particle size, freshness, rancidity and toasting — see A7-seeds-nuts-toppings; for the Arab pistachio/walnut/almond fillings that parallel this, see B3-baklava-and-phyllo-pastries. See img-b2bk-05 (pistachio grade chart).

4. The fat: sade yağ (clarified butter), not whole butter

This is a rule Turkish masters do not bend: baklava is layered and finished with clarified butter (sade yağ / sadeyağ), not whole butter. Whole butter carries water and milk solids; the water softens the yufka (you lose the crisp layers) and the milk solids scorch in the oven, giving black specks and a bitter, burnt taste. Clarified butter — water and milk solids removed — tolerates the high oven heat without burning [c9]. The regional prized fat is the Urfa "baklava yağı" / sade yağ made from sheep's milk, which also drives the characteristic aroma [c9].

You can see the mechanism directly in a butter spec: the Polmlek unsalted butter is 82% fat with ~16% water and ~0.7 g protein (milk solids) per 100 g — that water and those milk solids are exactly what you boil off and skim when you clarify [c10]. Melt the butter gently, let the milk solids settle and the foam rise, skim the foam, then pour off the clear golden fat and discard the milky sediment. The fat's usable temperature rises accordingly, which is why it survives a 230-240°C oven. The butter-grade and tourage/clarified craft (and why low-water fats behave better in laminated work) is in A4-butter-grades-and-specialist-types and A4-fat-types-and-selection. See img-b2bk-06 (clarifying butter, before/after).

Sourcing. Clarify a good unsalted 82% butter (Polmlek 10 kg / 25 kg) yourself for the most authentic flavour, or use a low-water tourage/laminating butter (Agart Tourage Croissant Butter) where a ready plasticised fat is preferred. Whichever you choose, butter is a milk allergen — declare it, and note that clarifying does not remove the milk allergen (trace milk protein remains), so sade yağ must still be declared as milk [c10][c12].

5. The bake: stone base, oak fire, 230-240°C

Baklava is baked on a stone base, traditionally in an oak-wood-fired (meşe odunu) oven, with the tray not over direct flame, at about 230-240°C for ~30-40 minutes, until the pastry is bright golden with visibly risen, separate layers [c13]. That is the professional wood-fired figure; be aware that domestic and convection recipes bake much cooler — around 170-180°C — for longer [c13], so translate temperatures to your own deck or convection oven rather than copying a home recipe into a stone deck. The oven-stage reactions (oven spring in the layers, starch gelatinisation, Maillard browning of the crust) are the universal ones covered in A5-baking-oven-science. See img-b2bk-03 (full process flow).

6. Sugar-syrup (şerbet) control — the make-or-break step

Syrup is where most baklava fails. Two independent things must be right: density and temperature at the moment of pouring.

Density / ratio. Professional Antep syrup is roughly 2 parts sugar to 1 part water; home recipes are simply less concentrated (commonly ~1:1 up to ~1.5:1 sugar-to-water, i.e. proportionally more water than the professional 2:1) [c14]. Cook it to the "thread" stage (ip kıvamı) — when it runs off the spoon as a continuous thread rather than in drops [c14]. Because caster/granulated sugar is essentially 100% sucrose [c17], the syrup is a controlled sucrose solution whose density — set by that thread stage — governs how much the pastry absorbs. A syrup that is too thin drowns the layers; too thick and it sits on top and sugars up. This is the same sugar-cooking-stage discipline as A6-sugar-work-techniques, and the shared syrup science across the region is in B3-attar-sugar-syrup-science. See img-b2bk-07 (syrup temperature rule).

Lemon / citric acid. A little lemon juice or citric acid added in the last few minutes of boiling inverts some sucrose, which prevents crystallisation (şekerlenme) and stabilises the syrup [c16]. Note the authentic nuance: traditional Antep syrup is often just sugar and water, because the baklava is eaten fresh and storage-crystallisation is not a concern — lemon/citric acid is more of a home and shelf-stable-commercial addition [c16]. On the platform, citric acid (a declared acidity regulator, ~99.5-101% purity) [c16] or a touch of glucose syrup both suppress crystallisation; the glucose route is the cleaner-flavoured choice for shelf-stable trays.

Temperature at pouring — read this carefully. There are two authentic conventions:

• Professional Gaziantep (fresh service): pour the hot just-cooked syrup (about 106.5°C in winter, 108°C in summer; GI range ~102-110°C) over the baklava as it comes hot from the oven, because Antep baklava is served fresh and warm [c15].
• Baklava that will rest or be stored: follow the temperature differential — hot pastry takes cool/lukewarm syrup; cold pastry takes hot syrup — because if both are hot the pastry goes soggy (hamurlaşma), and if both are cold the syrup sits on the surface and is never absorbed [c15].

The unifying principle is that you want the syrup to be drawn into the layers, not to steam them soft or bead on a cold surface. Pour slowly and evenly with a ladle over every cut piece. Handling caution: sugar syrup at ~106-108°C is a serious scald hazard — pour with care and keep clear of the steam [c14][c15].

FLAGGED FOR HUMAN REVIEW: the syrup-temperature rule is context-dependent (fresh vs stored) — confirm which convention matches your service model. All allergen statements below and the clarified-butter rationale are food-safety/allergen claims to verify against your own recipe and UK/EU labelling before publishing.

7. Reading a finished baklava (quality control)

The yufka must be paper-thin so the baked baklava is bright golden with distinct, risen layers; a well-made piece gives an audible crisp "crack" and melts in the mouth [c21]. Two diagnostic faults every Antep master listens for: no crack means the dough — especially the top yüzlük sheets — was rolled too thick; a burning sensation in the throat when eating means the dough-to-syrup balance is off (usually too much or too heavy a syrup) [c21]. Full fault table in data.json → fault_tables.

8. Varieties and the wider syrup-sweet family

Antep baklava is one product with many cut forms and cousins — kuru (dry, no cream), kaymaklı (with semolina cream), fıstık sarma (rolled thin around abundant pistachio), şöbiyet, bülbül yuvası (nightingale's nest), havuç dilimi (carrot-slice wedges) and midye (mussel-shaped) [c2][c20]. The same clarified-butter-and-syrup logic underlies the shredded-dough sweets — kadayıf and künefe — covered in B2-syrup-sweets-ottomans (Domson stocks a ready roasted kataifi/kadayıf pastry for those). Historically, layered baklava as we know it is generally credited to Ottoman palace kitchens (Topkapı), and the court tradition of the "Baklava Alayı" (Baklava Procession) distributed trays to the Janissary corps on the 15th of Ramadan; its deeper origins are debated, from Central Asian layered breads to Assyrian/Middle-Eastern antecedents [c20]. See img-b2bk-09 (varieties plate).

9. Allergens and food safety (declare before you sell)

Baklava is a tree-nut product. Pistachio — and, in cevizli (walnut) variants, walnut — is among the 14 allergens that must be declared under UK/EU food law (FIC, Reg. 1169/2011); a professional producer must also declare milk (butter and kaymak), egg (in the dough) and cereals containing gluten (wheat), and operate cross-contamination controls [c12]. Two nuances matter: clarifying butter does not remove the milk allergen (trace milk protein remains — declare milk anyway), and dusting with maize starch instead of wheat starch does not make the pastry gluten-free because the yufka itself is wheat flour [c9][c12]. Supplier specs already flag the picture: the citric-acid and several ingredient sheets record pistachio/tree-nut cross-contamination risk, the butter sheet declares milk, and the flour sheet is a gluten cereal [c12]. Treat the allergen and nutrition panel as a mandatory, human-verified step, mapped to your own recipe — not an afterthought.

Buy the ingredients for this

See data.json → linked_products / linked_brands for the full mapping. In short, a UK Turkish baklava kitchen orders from Domson: the strongest flour available (Centurion very-strong / Domson White Strong, ideally supplemented with an imported baklavalık un); unsalted 82% butter to clarify (or Agart tourage butter); wheat starch (Foodcom) or maize starch (AGRANA) for dusting between sheets; granulated/caster sugar plus glucose syrup and citric acid for the şerbet; pistachios (ground for filling, diced/granules for topping — noting the authentic choice is raw green boz fıstık, whereas the catalogue's roasted diced/granule pistachios suit topping and cost control); walnuts for cevizli (walnut) baklava; wheat semolina (irmik) for the semolina cream; and the ready roasted kataifi/kadayıf pastry for the sibling syrup sweets.