Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.


The structure of starch

Starch granules provide an insoluble but readily biodegradable storage system for plants. This balance of mechanical stability with degradability arises somehow from the chemical structure of the amylose and amylopectin that make the substance up, and from the arrangement of amorphous and crystalline zones within the granules. Starch also has considerable potential as a biodegradable plastic, but we need to understand the structure better before we can approach the ideal of a controllable service life followed by rapid degradation. A new high-resolution X-ray study by Waigh et al.1 allows us for the first time to track across the structure of an individual granule.

Structural polysaccharides, such as starch, cellulose and chitin, undergo simultaneous polymerization and crystallization. This poses a structural constraint: how does the enzymatic machinery convert dissolved monomers into a solid polymer without getting choked by the solid as it forms? In chitin, little is known about how the solid is laid down. Cellulose, in contrast, has been much studied, and is extruded by an organelle that traces a spiral across the cell surface. Starch has the added constraint that it must lend itself to later controlled depolymerization, as it is used to fuel plant growth. It might be formed by a retreating enzymatic peripheral layer, or by entrained enzymes.

Because starch is such an important product, one would assume that the inner structure of granules had long since been analysed to death by electron microscopy. This has been done2 but there are problems with artefacts, which have always dogged electron microscopy of polymers. The radiation intensity in a transmission electron microscope is so high that organic structures are usually destroyed in seconds. Hence there is always a concern about artefacts due to beam damage and, for starch, due to loss of water during preparation.

In the new study, pure and intense radiation from the ESRF Synchrotron in Grenoble was used in a 2-μm-wide X-ray beam that could be scanned across single granules to show the structure. It turns out that starch granules grow in rings, alternating between amorphous amylose and crystalline amylopectin (Fig. 1). The amylopectin crystals are built of double helices, which are arranged radially in the granule, so that the crystals are tangential to the surface.

Figure 1: A schematic starch granule, as mapped by X-rays.

Waigh et al.1 see alternating shells of amorphous and crystalline material, with the amylopectin helices that form the crystal lying perpendicular to the shell. They also find that the granules are ellipsoidal rather than spherical, so the helices do not point to a single focus, implying that some elongated structure must initiate the growth.

In contrast, the spherulite structure found in most synthetic crystalline polymers has radial crystals, with the polymer chains running tangential to the surface. Such tangential chains make sense if they are thought of as adsorbing to the existing solid surface of a sphere. The radial chains seen in starch suggest, on the other hand, that monomer is being added to growing ends sticking out of the surface. So the structure of starch is determined by polymerization kinetics, not crystallization kinetics, which means that it will be difficult to arrive at an equivalent structure in a synthetic moulded part.

In natural polymers, the amorphous regions are usually attributed to branched or irregular sections of the chains, which provide the organism with a means of controlling the stiffness and swelling of the polymer. Nuclear magnetic resonance studies of starch3 have detected amorphous regions which make up 70 per cent of the granule, some within the crystalline regions and the rest in distinct rings of highly branched and wholly amorphous material.

On the basis of electron microscopy, Oostergetel2 has postulated a different model for potato starch, based on radiating double spirals of amorphous and crystalline material, like the swirls on a barber's pole, with an 8-nm cylindrical cavity down the centre of each spiral. So far, this model is consistent with, but is not really confirmed by, the new data, and it does seem to be very complex compared with what we might expect from simple polymerization and crystallization processes.

So what of starch as a basis for biodegradable plastic? A decade ago the development of bacterially derived polyhydroxybutyrate as a biodegradable thermoplastic was stalled by the high cost of fermentation and extraction, and that re-ignited interest in starch. Moulded starch golf tees and foamed packing material are already being made.

But the problem, as with most biological polymers, is that its properties are so sensitive to water. Dry starch will not melt but simply dehydrates, cross-links and then burns. Water plasticizes the polymer, allowing it to melt and be processed as a conventional plastic. However, the properties are very sensitive to water content4 and this is hard to control at moulding temperatures of 200 °C. In addition, hydrolytic degradation breaks the chains during moulding. As the plasticizing water is lost, a moulded piece of starch is thus likely to shrink and become brittle. An alternative is to use glycerol in place of water, as it vaporizes less easily4, but it might leach out in use, again making the part brittle.

Natural starches differ in their amylopectin/amylose ratio and structure, probably to control the rate of hydrolysis. In reprocessed starch we have not attained a comparable level of control, but these new studies will teach us what kind of structures can be used to control degradation. It would also be valuable to transfer these lessons to biodegradable medical implants — normally made of polylactide or its copolymers with polyglycollide — in which it has proved very difficult to preserve mechanical properties past the first few per cent of the degradation reaction. We would really like a structure that retains its mechanical integrity to the point where much of the polymer has already been degraded and resorbed.

There is also renewed interest in grafting or blending starch with synthetic polymers to provide a combination of biodegradation with better mouldability5, but this would increase the cost. Instead, another substance — but one that is similar to starch — may be the answer: polyhydroxybutyrate moulds well and is not very water-sensitive. So perhaps what we are waiting for is a genetically modified corn that makes polyhydroxybutyrate instead of starch.

Figure 2: One potato.

Starches from wheat, banana and plantain are badly damaged by X-rays, so good diffraction patterns have not been obtained from them. But potato starch is tougher, probably because the crystalline layers are thicker, and so its structure can be determined.


  1. 1

    Waigh, T. A. et al. Macromolecules 30, 3813–3820 (1997).

    Google Scholar 

  2. 2

    Oostergetel, G. T. & van Bruggen, E. F. J. Carbohydrate Polymers 21, 7–12 (1993).

    Google Scholar 

  3. 3

    Morgan, K. R., Furneaux, R. H. & Larsen, N. G. Carbohydrate Res. 276, 387–399 (1995).

    Google Scholar 

  4. 4

    van Soest, J. J. G. & Borger, D. B. J. Appl. Polym. Sci. 64, 631–644 (1997).

    Google Scholar 

  5. 5

    Ramkumar, D. et al. Eur. Polym. J. 32, 999–1010 (1996).

    Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Calvert, P. The structure of starch. Nature 389, 338–339 (1997).

Download citation

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing