Nuomi Chemical SY-166 Thermal Paste Review – Cheap Poser or really cheap Alternative?

1 - Introduction, advertising claims, and technical specifications 2 - Fracture behavior, microscopy, and composition 3 - Performance comparisons and practical measurements 4 - Durability (pump-out), summary, and conclusion

Application and viscosity

During application, the paste shows a behavior that fits well with the previously gained knowledge about composition and particle structure. It is easy to squeeze out of the syringe, which indicates a moderate viscosity and a relatively low yield point. With the stated 220,000 cps, it is in the range of soft silicone pastes whose matrix is based on linear polydimethylsiloxane. The material reacts to pressure with a uniform flow, but still shows a clear stringing when the applicator is lifted. This behavior is typical for viscoelastic systems in which the cohesion of the polymer chains over short distances is stronger than the adhesion to the surface.

When first applied to the copper surface, the paste wets the surface cleanly, without air inclusions or visible streaks. This indicates good interface compatibility between the metal and the matrix, even without special amino or acetoxy functionalization. During the subsequent pressing in the test setup, the material is distributed evenly, the exit edges are symmetrical and without any recognizable dewetting. This confirms that the flow behavior of the paste is easy to control and forms an even layer as long as the application quantity is not exaggerated.

When the surfaces are relieved and separated again, the typical characteristics of a soft-elastic system become apparent. The paste pulls apart easily and forms threads between the surfaces before it tears cohesively. The result is clean tear-off edges and a partial film on both contact surfaces, which is typical of a matrix with a high proportion of linear siloxane chains. A higher filler density or a thixotropic setting would have mitigated this effect, but this behavior is normal for a low-cost formulation with 60 % oxide content and a relatively soft PDMS base.

Microscopic and chemical analysis supports this behavior: The aluminum oxide forms the structural framework, the zinc oxide fills the gaps and ensures uniform flow. The binder itself is largely inert, unmodified and completely odorless. It remains stable during application and does not react with copper or oxygen, which is reflected in the clean surface without discoloration or efflorescence. In practice, this means that the paste is easy to dispense and spread evenly, but it does run out sideways a little more than thixotropic products when pressed. A thin pre-spreading is therefore recommended to avoid air pockets. The consistency is reminiscent of a classic PDMS oxide system with a low internal structure, it is easy to process but potentially susceptible to layer migration under temperature cycles. Overall, the result is a well-rounded picture: a soft-flowing, chemically stable paste with high wetting properties, which is pleasant to work with but is unlikely to match the long-term stability of more structured industrial formulations.

Why the comparison with Polartherm X10?

I deliberately compared the Nuomi Chemical SY-166 with the Polartherm X10, even though they are in completely different classes in terms of price and performance. The reason lies in the characteristics of the X10, which has established itself as a particularly stable and reproducible mid-range paste in the measurement setup. It reacts to changes in pressure and temperature in a very controlled manner and, thanks to its fine particle structure, delivers almost ideal transient behavior with bondline thicknesses below 75 µm. This makes it a clean technical benchmark for comparison, especially when you want to visualize the limits of such an inexpensive paste as SY-166.

Chart comparisons and minimum layer thickness

During pressing, it becomes apparent that the bondline can still be compacted further despite high contact forces, but then stops abruptly. This is due to the pronounced agglomeration of the fillers: the microscopically detected aluminum oxide clusters with particle sizes between seven and twelve micrometers, supplemented by even larger, irregularly shaped lumps, act like microscopic spacers. These coarse agglomerates wedge between the contact surfaces as soon as the layer collapses under pressure and prevent further collapse of the bondline.

The silicone matrix itself, on the other hand, remains very soft and flowable. Under load, it yields laterally so that the paste visibly escapes at the edges during the pressing process. The effect is clearly twofold: While the binder is displaced, the rigid filler clusters remain in the contact zone, where they build up a mechanically stable but uneven residual thickness. The interaction of viscoelastic matrix and coarser particle structure thus leads to a self-stabilizing minimum layer thickness. The series of measurements and the microscopic observations result in a minimum bondline thickness of around 16 micrometers, which is practically impossible to fall below. If an attempt is made to thicken the layer further, local cavities, caked particle bridges and uneven contact surfaces are created. These micro-slip spaces increase the thermal contact resistance dramatically and make the measurement results irreproducible.

The effective thermal resistances Rth_{, eff}

The effective thermal resistance describes the total thermal barrier between two surfaces that are separated by a heat-conducting paste. It takes into account not only the intrinsic thermal conductivity of the material, but also contact surfaces, pores, uneven layer thickness and any air inclusions. This makes it the decisive practical value because it reflects the actual heat transfer in the application. A low effective thermal resistance means that heat is transferred efficiently over the entire contact surface, while high values indicate poor wetting, layers that are too thick or structural inhomogeneities.

A direct comparison between Nuomi Chemical SY-166 and Polartherm X10 shows a clear difference in behavior. The Polartherm paste starts at a layer thickness of 400 µm with an Rth of 1.076 K/W and decreases evenly with progressive compaction to 0.0629 K/W at 14 µm. The progression is almost linear, without any noticeable plateaus or abrupt changes. This indicates a very uniform particle distribution, a homogeneous binder phase and a good adaptation to the surface roughness. Polartherm reaches a minimum with thin layers, which stabilizes without cavities or particle bridges occurring.

SY-166, on the other hand, shows a different picture. It starts at 0.879 K/W and falls similarly at first, but reaches a clear saturation from around 50 µm. Below 25 µm the curve flattens out, and at 16 µm there is no further improvement, although the layer thickness is further reduced. This is due to the coarse, agglomerated structure of the aluminum oxide clusters, which act as mechanical spacers. While the Polartherm X10 enables continuous compaction thanks to finely dispersed, spherical oxides, the particles in the Nuomi become entangled in the bondline, which means that the layer can no longer collapse homogeneously.

I have now compared the relevant layer thicknesses from 25 to 400 µm as a bar chart for _Rth.

The effective thermal conductivity _λeff

Another difference can be seen in the effective thermal conductivity (λeff). The difference to the bulk thermal conductivity is important: _λeff describes the material property of the paste “volume” itself, i.e. the contribution of the layer proportional to the thickness. Rth,eff already contains the two interfaces to the copper standard as a constant additional term. For the Polartherm X10, it is between 3.7 and 1.9 W/mK, whereby the drop for very thin layers is due to the real geometric limitation, not to a deterioration in the material properties. The SY-166 starts with a calculated 4.55 W/mK, but drops to 1.8 W/mK with increasing compaction. This reflects the inhomogeneous compaction, in which the effective heat flow is disturbed by the unevenly distributed oxide clusters.

Of course, the whole thing is also shown as a bar chart for the most important layer thicknesses. Just click through and see where the paste is positioned:

The Polartherm X10 shows a consistently controlled compaction behavior with linear dependence between bondline and thermal resistance, while the Nuomi SY-166 reaches a mechanical limit thickness early on, above which the thermal resistance does not decrease any further. The structural limitation of the SY-166 leads to higher and more unstable Rth values, which level off between 0.09 and 0.1 K/W, while the Polartherm still achieves 0.06 K/W under the same conditions. This confirms the impression that Polartherm is a more finely dispersed, rheologically more stable and thermally more efficient paste, while Nuomi is soft to apply but too coarse in terms of particle structure to form a uniformly thin bondline in practice.

Bulk thermal conductivity, interface resistance and quality of measurement

The linear regression analysis from the TIMA-5 system shows the measured relationship between bondline thickness and effective thermal resistance of Nuomi Chemical SY-166. The bulk thermal conductivity is determined from the slope of the straight line, and the interface resistance at the interfaces between paste and copper is determined from the intercept. Both values together represent the real, practical performance of the material.

The measured bulk thermal conductivity is 4.779 ± 0.021 W/mK, the interface resistance 5.7 ± 0.2 mm²K/W. The coefficient of determination (R² = 0.99994) is almost 1.0, which means that the measurement points show an excellent linear correlation. This shows that the material behavior remains stable under pressure and temperature and that there are no significant outliers due to pump-out, dewetting or structural instability. The TIMA system records real contact conditions under ASTM-compliant test pressures, which means that both the volume conductivity and the contact resistances can be precisely separated from each other.

The bulk thermal conductivity of around 4.8 W/mK is absolutely plausible from a physical point of view. It is exactly in the range that can be expected for silicone pastes with an oxide combination of aluminum oxide and zinc oxide. In compact form, aluminum oxide has a thermal conductivity of between 25 and 35 W/mK, zinc oxide between 40 and 60 W/mK. In a silicone matrix, this theoretical value is drastically reduced by the proportion of polymer, the contact resistance between particles and the incomplete percolation. In practice, this results in a range of around 3 to 6 W/mK – exactly where the SY-166 fits in.

The interface resistance of 5.7 mm²K/W is also typical for soft, viscoelastic silicone pastes with moderate filler wetting. It describes the additional thermal barrier that is created at the interface between the copper and the paste. This transition layer is caused by microscopic unevenness, oxide residues or weak interfacial adhesion. Values below 5 mm²K/W are usually only achieved by products with chemically functionalized binders or higher particle density. The fact that SY-166 is slightly higher reflects the soft, unmodified PDMS matrix and the only mechanically embedded oxide particles.

The advertised 16.6 W/mK, on the other hand, represents a typical case of a so-called bucket measurement. The paste is tested in a closed container or cylinder without real interfaces. Heat is conducted through the homogeneously compacted material without taking contact losses into account. At best, such values reflect the theoretical conductivity of the filler in the ideal case, but not the behavior of a 20 µm thin layer between two metal surfaces. In the practical thermal path of a processor or a test head, however, it is precisely these contact resistances that play the decisive role.

The linear regression curve with R² ≈ 1.0 shows that the measurement data is precisely matched. The low scattering proves that the heat transfer in the layer is uniform and that there are no systematic deviations. This results in a very clean, quantitative description of the material, which clearly reveals the discrepancy between marketing value and real performance. The SY-166 has a solid but unspectacular thermal performance with 4.8 W/mK in bulk and 5.7 mm²K/W as contact resistance. The nominal 16.6 W/mK is a theoretical ideal value from non-comparable laboratory conditions and is around 3 to 4 times higher than the realistically measurable thermal conductivity. The TIMA analysis therefore provides a technically reliable, reproducible result that clearly demonstrates the quality of the measurement and at the same time reveals the limits of simple factory specifications.

GPU simulation

First, we take the values that show the two temperatures at the respective contact surfaces between which the paste is located and form a delta. These curves are no longer completely linear, as the interface resistance also changes slightly. And we no longer calculate with 6 points, but only with 2 absolute values for the temperature difference instead of a gradient as with_TTim, whereby the sample temperature remains constant. And what is the point of all this? The behavior is similar to that of a graphics card, which has to manage without an IHS and where the delta is usually measured between the substrate and the water temperature.

If you look at the measurement series of both pastes over the temperature delta between T3 and T4, i.e. the upper and lower sensor channels of the heating section, it becomes clear that the Polartherm X10 achieves a more even and faster temperature equalization. The ΔT between the two measuring points decreases steadily and linearly with increasing compression, which indicates efficient heat transfer through the entire bondline. Even with thin layers around 25 µm, the temperature gradient remains stable and low, which shows that the paste works homogeneously both in the bulk and on the contact surfaces. The SY-166 behaves less ideally here: its ΔT remains higher over large parts of the compaction process and only starts to drop significantly at a late stage. This is due to the uneven particle distribution with local air pockets and rigid agglomerates, which slow down the vertical heat flow and cause a somewhat slower thermal transient response.

CPU simulation

If we normalize the values for the heater, we already have sufficient thermal resistance in the copper reference block to simulate the CPU temperature and its differences with different pads in comparison with each other and in relation to the thickness of the paste replacement. It is precisely this variable evaluation that no test on a CPU can offer, because the respective CPUs are bent differently and it is therefore not really reproducible. But the TIMA5 test does, because I can measure all distances, which is simply not possible on a single CPU.

If you look at the simulated CPU temperature, represented by the T4 sensor, there is an even clearer difference. With the Polartherm X10, T4 remains noticeably lower and more stable even with a high heat load, as the heat flows more efficiently from the heat source into the heat sink. The temperature distribution over time shows hardly any jumps or plateaus, which indicates good adaptation of the paste to the contact surfaces and uniform compression. The SY-166, on the other hand, shows a more sluggish behavior: The T4 temperature initially rises faster and reaches a higher equilibrium level before stabilizing. This corresponds to the higher effective thermal resistance barrier due to its coarser particle structure and increased interface resistance. Overall, it can be concluded that the Polartherm X10 is thermally more efficient and mechanically more homogeneous in thin layers, while the SY-166 is more limited by its internal particle morphology and therefore has higher temperature differences within the layer, which only pays off at much higher layer thicknesses.

Interim conclusion

In a direct comparison, the TIMA measurement of the SY-166 with 4.8 W/mK bulk thermal conductivity and 5.7 mm²K/W interface resistance shows a solid but clearly inferior result compared to the Polartherm X10, which achieves around 6.2 W/mK and only around 3.9 mm²K/W at a similar test temperature. While the Polartherm forms a thinner, more homogeneous bondline due to its fine particle structure and stable interface adhesion, the coarser agglomerate structure of the SY-166 limits the heat flow. Although both series of measurements show excellent linearity, the SY-166 remains thermally inefficient and more prone to losses on the contact side.