How to detect green pea and spinach adulteration in pistachio

Green pea and spinach in ground pistachio cannot be spotted by eye, and official findings also show sunflower seed and dye. What each method sees and how to set up intake.

Green pea and spinach mixed into ground pistachio are close enough in colour that they cannot be reliably separated by eye. A dessert manufacturer checks the raw material at intake by comparing its measured composition with batches known to be correct, then settles any deviating batch through targeted laboratory analysis. The search should not stop at green pea: the Turkish ministry's adulteration list shows sunflower seed and food dye in pistachio desserts.

What fraud turns up in pistachio?

Green pea and spinach are the substitutes most often named in studies published in Türkiye. The reason is simple: a similar colour and look at a much lower value. The official record says something slightly different. Among 321 records visible on 7 November 2026 on the list of Türkiye's Ministry of Agriculture and Forestry, dated 1 June to 27 October 2026, the pastry group holds three pistachio products:

  • Sunflower seed and food dye in one pistachio baklava
  • Food dye in another pistachio baklava
  • Sunflower seed in a pistachio eclair

Neither green pea nor spinach appears in the non-compliance text of those 321 records. That does not mean pea is not being used; it only shows what reached official analysis in one window. The takeaway for a quality team is that intake built around one adulterant can miss another. One listed record combines sunflower seed with food dye, so a green colour alone proves nothing.

What does each method see, and at what level?

Studies from Türkiye allow a useful side-by-side comparison:

MethodStudyRange testedFinding
NIR spectroscopyGeniş, Durna, Boyacı, 202120-80% pea and spinachPrediction error (RMSEP): pea 7.87, spinach 4.69
NIR and mid-infraredTemizkan et al., 20200-50% pea, 63 samplesPrediction error: NIR 2.55%, mid-infrared 9.14%
Low-cost spectral sensorTürköz, Özçelik, Turgut, 20245-50% pea and spinachTest error (RMSE) 6.71; authors see a screening tool, especially above 10%
Liquid chromatography with mass spectrometryÇavuş, Us, Güzelsoy, 20181-10% pea in baklavaDetected down to 2.5% in baklava
DNA (HRM)Ay, Hürkan, 2023Powder mixes down to 999:1Separation at 0.1%, but with low confidence values at that level

Points to keep in mind when reading this table:

  • All of these are prepared mixtures. The studies used laboratory samples at known ratios. Their results should not be carried straight over to field batches from different regions and harvests.
  • Error sets the floor. A screen with a prediction error of several points cannot reliably separate a 5% blend. For low-level suspicion, move to mass spectrometry or DNA.
  • A model separates what it knows. A model built on pea and spinach may not flag sunflower seed or dye with the same accuracy. Profile comparison raises the suspicion; the laboratory names the substance.
  • Baked product is harder. Few methods have been worked up on a processed matrix such as baklava. Control is most effective while pistachio is still a raw material.

How should a dessert manufacturer set up intake?

  1. Choose the form. Ground pistachio and pistachio paste are the hardest forms to check. Buying whole kernels and grinding in-house narrows the room for blending, but does not close the dye or quality question.
  2. Build your own reference. Set aside batches whose source you trust, with region and harvest recorded, and refresh the set each harvest. See how to build a reference library.
  3. Measure by batch. Increase frequency for new suppliers, prices clearly below market and source changes. Periods of rising pistachio prices also raise blending risk; revisit your vulnerability assessment then.
  4. Ask the lab for a panel that matches the suspicion. Green pea, spinach, peanut, sunflower seed and dye are separate questions. A general request for a "pistachio analysis" may not answer the one you have. More on reading certificates in our certificate of analysis article.
  5. Hold the batch and keep a retained sample. The batch stays out of production until the result is clear, and any second analysis in a dispute uses that sample.

What is expected on labels and packaging?

Whether pistachio is present is not the only question; how much is present matters too. Article 25 of the Turkish Food Codex regulation on food labelling requires an ingredient's quantity to be declared when it appears in the food's name or is emphasised on the label by words, pictures or graphics. A "çerez tatlısı" (a nut confection) file heard at the Turkish Advertising Board's 13 August 2026 meeting is instructive. According to the Ministry of Trade's announcement, the packaging highlighted images of hazelnut, pistachio and apricot kernel, which made up 4%, 3% and 3% of the product. Peanut, at 49.5%, was missing from the images.

Responsibility also falls on whoever owns the finished product. Article 24 of Türkiye's Law No. 5996 bans processing or marketing adulterated product, and the sanctions in Article 40 are aimed at the operator who produces, imports or sells under its own name. Having bought the raw material from someone else does not change that. The list itself names producer, product, brand and batch. Protected names raise the stakes further: Antep Baklavası gained EU protected geographical indication status in 2013 and Gaziantep Fıstık Ezmesi (pistachio paste) in 2025, with pistachio at the centre of both.

Where does Molevex fit?

Molevex adds nothing to the pistachio; it measures the batch's own composition and compares it with a reference built from batches you are sure of. An added or substituted component shows up as a difference, a sample that does not match the declaration is taken up in the laboratory, and the result is finalised there. Each result is stored against its batch and supplier. Molevex does not replace official controls. See the food page for more.

Sources

Meeting request

Let's decide together which of your products to examine.

The first meeting reviews your product, sales channels and the risks you face, and recommends a suitable method.

Request a meeting