ULTEM DU762 is designed for high-volume injection moulding and PVD metallisation, with SABIC positioning the PEI grade as a lighter alternative to metal frames.
Image source: Mundoplast, showing an earlier SABIC PVD-compatible ULTEM grade for consumer electronics. The image is illustrative and is not presented as ULTEM DU762.
SABIC has introduced a new ULTEM grade aimed at one of the most visible structural components in a premium smartphone: the middle frame.
The new ULTEM DU762 is an unreinforced polyetherimide-based material developed for high-volume injection moulding and subsequent non-conductive physical vapour deposition, or PVD. SABIC is positioning the grade against two established routes: machined metal frames and glass-fibre-reinforced polycarbonate parts that rely on multi-layer vacuum metallisation.
Join PlasticParts Hub
Create an account to connect with companies and send RFQs.
The material announcement is interesting for injection moulders because the pitch is not simply that a polymer can replace a metal part. The proposed manufacturing route changes the production sequence. Instead of machining titanium, stainless steel or aluminium and then managing antenna breaks and finishing operations, a producer can mould the geometry near-net-shape and add a thin metallic surface afterwards.
That potentially shifts value from machining into tooling, process control and surface preparation.
The business case starts with the production route
Middle frames have to do more than look expensive. They carry structural loads, locate internal components, survive drops and daily chemical exposure, and provide a visible surface around the perimeter of the device. They also sit next to antennas, cameras, batteries and heat-generating electronics.
Metal has obvious strengths in stiffness and perceived quality, but it also brings manufacturing steps that do not scale as elegantly as injection moulding. A complex frame may require machining, drilling, polishing, surface finishing and assembly of non-conductive antenna sections.
SABIC argues that DU762 can be produced using high-speed, high-volume injection moulding. The company also notes that the material is substantially lighter than metal and, because it is non-conductive, can remove the need for the visible plastic antenna split commonly used in metal frames.
For a moulder, the opportunity is straightforward: if the geometry, tolerances and surface quality can be produced repeatably in a mould, a large portion of the part is created in one cycle rather than through multiple subtractive operations.
That does not automatically make the plastic route cheaper. The tooling will be expensive, the moulding window is demanding, and the PVD step still has to be controlled. The economics improve when annual volumes are high enough to spread those fixed costs over a large number of parts.
PVD compatibility is central to the proposition
A polymer middle frame only works in the premium segment if the surface can match the visual standard expected from metal.
SABIC says DU762 is compatible with non-conductive PVD sputtering, which creates a thin metallic layer without turning the complete component into an electrically conductive metal structure. The company is targeting a high-gloss finish while retaining the underlying design freedom of an injection-moulded thermoplastic.
This is where the comparison with glass-fibre-reinforced PC becomes more specific. Glass-filled PC is already mouldable, but SABIC says the fibres can migrate or print through at the surface and disturb gloss. The company also contrasts PVD with the thicker multi-layer NCVM processes commonly used on reinforced PC parts, arguing that PVD gives a cleaner edge profile.
ULTEM DU762 is unreinforced, so the surface presented to the coating process does not contain glass fibres. That can be useful where very sharp cosmetic boundaries, uniform gloss and thin coating stacks are important.
The trade-off is that an unreinforced high-performance thermoplastic does not gain the stiffness boost that comes from glass fibre. Part geometry therefore becomes critical. Ribs, local wall thickness, corner radii and fastening features have to carry more of the structural design burden.
The moulding window is not a commodity-PC process
The material data is a reminder that DU762 belongs to the high-temperature end of injection moulding.
SABIC's material finder lists a drying temperature of 150°C for four to six hours, with a recommended maximum moisture level of 0.02%. Listed melt and nozzle temperatures are in the 370–400°C range, while mould temperature is 150–180°C.
For the processing cell, those numbers immediately narrow the equipment choices.
A processor needs a drying system capable of holding the specified temperature consistently and material handling that prevents reabsorption before the resin reaches the machine. The barrel, screw, nozzle, hot runner and seals all need to be appropriate for sustained high-temperature operation. Mould temperature control also has to be designed around a tool running far hotter than the temperatures common in general-purpose ABS or PP production.
This is not a resin that should be dropped into an existing consumer-electronics moulding cell without checking the complete thermal package.
High mould temperature can help surface reproduction and reduce frozen-in stress, but it also increases the cooling load. In a cosmetic frame, the processor may have to balance cycle time against surface quality and dimensional stability rather than simply chasing the fastest possible ejection.
Dimensional control will decide whether the concept scales
SABIC's current data lists mould shrinkage of roughly 0.6–0.8% in both flow and cross-flow directions for DU762. The material is unreinforced, which avoids the strong anisotropy often associated with glass-fibre orientation, but the finished component can still move as it cools.
A smartphone frame is especially unforgiving because it surrounds multiple tightly controlled interfaces. Display glass, back cover, buttons, internal brackets and antenna components all depend on the frame remaining within tolerance.
For a new programme, mould qualification should therefore look beyond a simple dimensional check at the machine. Measurements after conditioning, coating and assembly are more useful. PVD introduces another thermal and handling stage, and the final geometry is what matters.
Gate location will also need careful work. Long thin flow paths, knit lines around openings and visible gate vestige can all affect cosmetic quality. The fact that the component is later metallised does not make underlying moulding defects disappear; in some cases, a reflective coating can make them more obvious.
Chemical resistance matters more than the brochure photograph suggests
SABIC highlights resistance to sunscreen and other chemicals as a feature of DU762.
That may sound secondary compared with stiffness or gloss, but consumer electronics are exposed continuously to skin oils, cosmetics, hand creams, cleaning products and heat. A premium surface that looks good after coating still fails the application if the substrate swells, cracks or loses strength in normal use.
The grade is also designed for high heat resistance. SABIC's material data shows heat-deflection temperatures around the high-180s to low-200s Celsius depending on test method and load. Those values are far above service temperatures inside a phone, but they help explain why PEI is attractive for a component that may see thermal cycling during processing, coating and device operation.
The antenna advantage is architectural, not just cosmetic
One of SABIC's stronger arguments concerns antenna design.
Metal frames often need non-conductive breaks because a continuous metal perimeter can interfere with wireless performance. Those inserts or gaps are technically necessary but can interrupt the appearance of the frame.
A metallised polymer takes a different route. The substrate remains non-conductive, while the decorative metal layer can be engineered as part of the surface system. That creates more freedom for RF design and can remove a visible material transition.
For moulders, the attraction is broader than weight or cycle time. The ability to combine structural geometry, RF behaviour and cosmetic finishing in one moulded substrate can simplify the architecture of the complete product.
Whether manufacturers adopt that route will depend on device-level testing, not material data alone.
What a moulder would need to qualify
A realistic qualification programme for DU762 should include more than the standard tensile bars supplied in a material datasheet.
Useful checks would include:
moisture control before moulding and its effect on surface quality;
fill balance through the complete frame geometry;
weld-line strength around openings and bosses;
flatness and warpage after moulding;
dimensional change after PVD processing;
coating adhesion after thermal cycling;
abrasion and chemical exposure on finished surfaces;
screw, barrel and hot-runner compatibility at the required melt temperature;
cycle-time impact of the 150–180°C mould-temperature range;
repeatability of cosmetic quality across multiple cavities if the programme uses a family or multi-cavity tool.
The last point deserves attention. A phone frame is a large, visible component compared with many precision electronic parts, so cavity count may be limited by platen size, projected area and tool complexity. The economic advantage of injection moulding will therefore depend on both cycle time and the number of good frames produced per hour.
Plastic versus metal will be decided at system level
SABIC's release naturally presents DU762 in the most favourable comparison with incumbent materials. The commercial decision will be more complicated.
Metal has established supply chains, predictable mechanical behaviour and strong consumer associations with premium products. Injection-moulded PEI brings lower mass, design integration and potentially less machining, but it introduces high-temperature moulding, expensive tooling and a coating process that has to survive the complete life of the device.
A resin-price comparison with aluminium would miss most of the economics. What matters is the total system cost of producing an acceptable frame at scale.
That includes machining or moulding equipment, tooling, scrap, cycle time, finishing, antenna architecture, assembly, quality control and yield after coating.
If DU762 can reduce the number of manufacturing steps while meeting structural and cosmetic requirements, the material could open a meaningful new application for high-performance injection moulding. If coating yield or dimensional control proves difficult, the theoretical savings will disappear quickly.
SABIC has supplied enough technical information to make the concept credible. The next evidence the market will need is production data from real smartphone programmes.
Two-Shot Injection Molding: When Core-Back Tooling Makes Sense
Core-back tooling can keep a part in place between two injection shots. Review the mold sequence, material interface and cycle evidence before choosing the process.
