Developing a high-performance, low-streak degreaser

By Julia Wates

In This Section

November/December 2014

  • Many commercially available surfactant-based cleaners exhibit excellent degreasing performance but cause streaking. Glass cleaners cause less streaking but cannot remove tough grease.
  • Numerous so-called multi-surface cleaners have been introduced in an attempt to fill this gap, but current commercial products do not deliver a strong combination of heavy duty grease removal and low streaking.    
  • This article describes the development of a surfactant-based cleaner that does.

A high-level review of technologies used in commercial aqueous household cleaners identified a gap in the market when products are classified according to their effectiveness at removing greasy soils and their streaking behavior.

All-purpose cleaners (APCs) contain relatively high amounts of surfactants—particularly nonionic surfactants. These products are excellent at removing greasy soils but can leave visible residues behind after drying. This results in an unattractive streaky appearance and limits the type of surface on which these cleaners can be used. In addition, some APCs contain high levels of volatile organic compound (VOC) solvents, such as glycol ethers and alcohols, which are becoming increasingly less desirable as regulations get tighter.  

At the other extreme are glass cleaners in which the main active ingredients are solvents and ammonia, with only trace amounts of surfactants. Such products cause significantly less streaking but cannot remove tough grease, so they are only effective on lightly soiled surfaces.

Multi-surface cleaners, which were introduced to fill the gap in performance between APCs and glass cleaners, tend to be similar to glass cleaners but with higher levels of so-called “low residue” surfactants like alkyl glucosides, amine oxides, and anionics.  Most products in this class represent a compromise between cleaning and streaking and are not very effective at removing heavy duty grease.

Consequently, the challenge was to develop a surfactant-based cleaner that combines excellent degreasing with very low-streaking, preferably without high levels of VOC solvents. However, it quickly became apparent that the desired cleaning performance on heavy duty soils could not be achieved with classic low residue surfactants. So, the problem was approached from the opposite direction: by investigating whether the high-streaking of a good degreaser could be reduced by adding something else to the formulation.  An earlier study (“New Technologies in Surface Care,” Julia Wates, 101st AOCS Annual Meeting, Phoenix, May 2010) had shown that adding a dispersion of colloidal silica to a cleaner leaves behind a hydrophilic layer on the surface after drying. This made the surface easier to clean the next time, especially in the case of greasy soil.

The breakthrough in the current project came when it was demonstrated that hydrophilic surface modification by colloidal silica can also reduce the appearance of streaking without adversely affecting cleaning performance.  This is important because cleaning and streaking typically oppose each other, so a change in composition that improves one of these parameters is usually detrimental to the other.  Experiments in which cleaning, streaking, and formulation stability were simultaneously optimized demonstrated that a successful formulation needed at least two carefully selected surfactants plus the correct type of colloidal silica combined in the right ratios. A prototype low-streak blend was developed that met those requirements.

The specific requirements for a low-streak degreaser exemplified by the prototype (WO 2012/080197: “Low-streak degreaser composition,” J. Wates, M. Dery, A. Slikta and O B. Ho) are that it must contain at least one nonionic surfactant having a critical packing parameter of >0.95, inorganic nanoparticles such as colloidal silica, and a second surfactant having a critical packing parameter of <0.85. 

The critical packing parameter or CPP of a surfactant molecule is a number that indicates the relative sizes of the hydrophilic head group and the hydrophobic tail.  The larger the head group is relative to the tail, the lower the CPP. CPP is important because it is an objective measure that can be used to distinguish between the primary nonionic surfactant (usually an alcohol ethoxylate), which is largely responsible for the degreasing performance of a cleaner, and secondary or co-surfactants (such as alkyl glucosides, amine oxides, or amphoterics) that are not very effective degreasers on their own but provide other functions in a formulation. There is a relationship between CPP and the  shape of a surfactant molecule which, in turn, determines how a surfactant self-assembles or forms aggregates in aqueous solutions and on surfaces. Table 1 summarizes the connection between CPP and the shapes of surfactant molecules and their aggregates.  Secondary or co-surfactant molecules that have low CPP values are shaped like cones or truncated cones, and they tend to form spherical or cylindrical micelles with the large head groups on the outside and the smaller tails on the inside. On the other hand, when the CPP is close to 1, as is the case for many alcohol ethoxylates, the cross sections of the head and tail are similar in size. This results in the individual molecules being cylindrical, and their aggregates are lamellar phases or planar bilayers.

Table 1 

Streaking is a dynamic process that occurs as water and other solvents in a cleaner evaporate from a treated surface and the surfactant concentration in the formulation gets higher and higher. Streaks are simply visible residues of surfactants and other ingredients that are left behind on a surface after the application and drying of a cleaning product. Nonionic surfactants with high CPP values that form lamellar phases or other large aggregates at high concentrations are most likely to leave visible residues or streaks. In contrast, secondary surfactants with low CPP values that tend to form smaller aggregates are lower streaking, suggesting that streaking can be reduced by controlling surfactant self-assembly on the surface.  The aggregation behavior of a single surfactant can be predicted by its CPP, but self-assembly of mixed surfactants will depend on the ratios and CPP values of all the components and may be quite different from that of any individual surfactant.  By combining surfactants with different CPP values in the right way, it should be possible to manipulate the phase behavior of the mixed systems and minimize formation of lamellar regions at high surfactant concentration without sacrificing good degreasing performance in the dilute cleaner.

To determine whether this theory is valid in practice, an investigation was conducted into the mechanism of streaking for two dilute cleaning formulations.  The high-streak formulation was a classic aqueous degreaser based on a primary nonionic surfactant with CPP ~1, while the second formulation was the low-streak prototype containing an optimized blend of nonionic and secondary surfactants with colloidal silica.  Both formulations were applied to various surfaces and allowed to dry. Their appearances were then compared with different techniques.  Figure 1 shows the two cleaners spread on glass mirror tiles and photographed in a light box against a black background.

 Figure 1


When the high-streak formulation was viewed under a light microscope at low magnification, the streak appeared as a lane with spots in it.  A scanning electron microscope (SEM) image of the streak at 1000 x magnification (Figure 2) revealed that these spots are evenly spaced “blisters” having an average diameter of 2-5µm.  The structure of the blisters was studied in more detail with atomic force microscopy (AFM), a technique used to investigate the shape and features of surfaces at very small scales. 

Figure 2 

Information is gathered by “feeling” the surface with a fine-tipped cantilever that is deflected by surface forces in a manner that is analogous to a stylus moving over the grooves of a vinyl record.  In this way, a picture is built up of what the surface looks like.  Figure 3 shows an AFM image looking down at a single blister on an oxidized silicon wafer spin-coated with the high-streak formulation.  The scale of the image is 5µm across, and the blister appears as a flat spot surrounded by empty space.  Closer inspection shows that the blister is built up from layers of lamellar structures perpendicular to the treated surface. These are assumed to be surfactant, and this observation is consistent with the proposed mechanism for streaking.


Figure 2


A simple drying experiment was run using a rotary evaporator to simulate what happens as the cleaner loses water and dries on a treated surface.  Phase separation was observed at ~50% solids content and although the mixture remained fluid and pourable, the separation was irreversible and the mixture could not be diluted back with water to a single phase. This suggests a mechanism for streaking in which phase separation upon drying results in the formation of multiple nucleation centers on the surface leading to the 2–5µm blisters with almost no material in between.  The blisters are large enough to scatter light, making them visible to the naked eye as streaks.


Figure 4


Moving next to the low-streak formulation, nothing was observed under the light microscope at low magnification where the cleaner was applied.  Even with the SEM at much higher magnification, the material appeared to be evenly distributed on the surface and there were no remarkable structures visible.  At the highest SEM magnification of 25K (Figure 4), a fine pattern emerged with darker regions that looked like holes where there was no material present. The AFM images for the low-streak formulation are consistent with this interpretation.  Figure 5(a) has a scale of 5µm across and confirms that the surface is covered by sub-micron holes.  Figure 5(b) has a smaller scale of 1µm across and it shows that the holes are separated by material made up of rod-like structures with a repeat dimension of 10-20nm.  The most likely explanation for the generation of these structures is that the colloidal silica particles in the formulation interact with any remaining water in the surface film on drying, and are attracted by capillary forces into the gutters between the wormlike micelles that form at high surfactant concentration.  Consequently, the micelles are coated with silica and become more rigid and rod-like. In contrast with the behavior of the high-streak formulation, there are no “large” (micron-scale) structures present on the surface that can scatter light.


Figure 5

The study was completed by running the drying experiment for the low-streak formulation.  In this case, the evaporated material was highly viscous and not pourable which is consistent with the presence of wormlike micelles.  However, the mixture remained transparent and was easily diluted back to the original water level, indicating that there was no phase separation.  What this means is that the material forms an evenly distributed film on the surface as it dries; there are no structures large enough to scatter light and therefore no streaks.  Interestingly, when the silica is removed from the formulation, the behavior moves toward that of the high-streak formulation and there is some phase separation at very high solids content.  So, streaking can be reduced by balancing the critical packing parameters of the surfactants, but to obtain zero streaking the silica is needed to stabilize the wormlike mixed micelles formed during drying and prevent the system from eventually transitioning to a lamellar phase. 

The graph in Figure 6 compares streaking data for the prototype low-streak blend formulated at different concentrations with commercial ready-to-use household cleaners applied to glass mirror tiles and photographed in a light box.  Average streak intensities were generated by image analysis. A low number indicates low-streaking (anything below around 30 is invisible to the naked eye).  Only two commercial cleaners demonstrated less streaking than the low-streak prototype formulations. At the other extreme, an “all-purpose cleaner with orange action” was very streaky.

Figure 6 

Although the main focus of this article has been on streaking, the low-streak prototypes also had to meet stringent targets for degreasing performance.  Figure 7 shows the results of a non-mechanical test in which the cleaners are poured onto a painted metal panel coated with mineral grease and rinsed with water.

 Figure 7

The prototype low-streak formulations were able to remove more than 90% of this very tenacious soil. The only commercial cleaner that gave comparable degreasing performance was the “all-purpose cleaner with orange action.” But, as previously demonstrated, this cleaner was also the highest streaking product.  Figure 8 shows the performance of four formulations in mechanical cleaning tests in which a greasy kitchen soil was baked onto stainless steel panels and removed with a four-lane scrub tester. The commercial all-purpose and multi-surface cleaners performed quite well after 20 scrub cycles, while a glass cleaner barely removed any soil.  The prototype low-streak formulation gave excellent soil removal under these test conditions.


Figure 8 Finally, it should be mentioned that although the reference products in this study were household cleaners, the same technology has been successfully used in industrial and institutional applications where a combination of excellent grease removal and low-streaking is required, such as in heavy-duty window cleaners for external use on buildings in high-traffic areas.


Julia Wates is principal research chemist in the Cleaning Applications Technical Service & Development Group of Akzo Nobel Surface Chemistry LLC.  She can be contacted at