A rheological characterization of synthetic detergent formulations

Solid bar cosmetic products are a valuable alternative to their functionally equivalent liquid counterparts. They are quickly growing in popularity as a response to the increasing commitment to the Green Economy and demand for sustainable products.¹ As a global movement toward plastic-free products and packaging is gaining momentum,^2,3 solid cleansing bars offer a solution to these problems as they can easily be stored and distributed in simple cardboard boxes or even in the absence of any packaging whatsoever, which would be impossible with liquid products. Moreover, high quality cleansing bars are concentrated products that could potentially be dissolved in water before their use; thus, further decreasing their environmental footprint, both in terms of packaging and transportation costs, due to the reduction in size and weight.

It is commonly thought that any solid cleansing bar could be referred to as soap, but this is in fact a misnomer. There are two distinct kind of products that are available on the market. Traditional soap, also commonly referred to as salt of fatty acid, has been produced for centuries and is the result of a reaction called “saponification,” which occurs between fat molecules, either of animal or vegetable origin, and a strong alkali base.⁴ The resulting soap is a detergent (defined as a surfactant or mixture of surfactants that have cleaning properties in dilute solutions,⁵ such as the ability to remove dirt from a solid surface) that also has foaming properties.⁶ Additionally, soaps belong to a class of surfactants called emulsifiers, which are substances that help the formation and stability of emulsions.⁶ In essence, soap has amphiphilic properties as it can form chemical bonds with both water (hydrophilic) and fats (lipophilic). Synthetic detergents, commonly referred to as syndets, are an alternative to traditional fatty acid-based soaps.^4,6 They also are based on fats, but they are produced synthetically through a series of chemical reactions rather than a single saponification process.⁷ As suggested by their name, syndets are also detergents as well as wetting and foaming agents; thus, their primary use is as cleansing bars, although they are also used in laundry products. One of the key features that differentiates the two products is the way they are chemically manufactured. In particular, the reaction used to produce syndets does not involve the use of a strong base, and their pH is much closer to the slightly acidic pH of the skin, which is typically around 5.5.⁸ In contrast, soaps usually have pH of circa 11, which can cause skin dryness and irritation after frequent use.^7–10 For this reason, syndets can be considered more appropriate than traditional soaps for general skin cleaning products, especially for sensitive skins. An additional advantage of syndets over classical soaps is their increased compatibility with additives such as colorants, fragrances, and other property-enhancing compounds,^8,11 which are commonly used in the formulation of different cosmetic products but can be unstable in the alkaline pH conditions of soap.¹² For this reason, syndets are also better suited for a wider range of products such as facial bars, shampoo bars, and shaving bars.⁶ Finally, a further advantage of syndet formulations over traditional soap is that their cleansing performance does not decrease with increasing levels of electrolytes,^8,13 such as calcium carbonate (CaCO₃).

For the reasons described above, the demand for syndet cosmetic bars has been growing rapidly over the recent years and it is expected to keep increasing in the years to come.^14–17 Therefore, cosmetic bar manufacturers have been trying to increase the production capacity by using existing production lines originally designed to produce classical soap bars; a process that has revealed many difficulties associated with different processabilities of the materials. In the authors' experience,¹⁸ syndet manufacturing has proven to be significantly more difficult to control and scale up from a lab extruder to a production line. Moreover, we have observed syndet manufacturing to be extremely sensitive to external factors that are not yet well documented or understood in literature and require further studies. One of the key issues associated with the extrusion process of syndet bars is the observed occurrence of blockages inside the extruder and/or at the conical part of the extruder. In addition, we have identified that moisture and chemical contents of the processed formulation have a significant impact on the material processability. However, even when materials' moisture content is carefully controlled, production issues still occur. This leads to the need for further investigations to achieve a better understanding of the correlation between the materials' viscoelastic properties and their processability.

In this study, we present a rheological characterization of a set of samples that include syndet base, syndet formulations sampled at different stages of mixing, as well as a classic basic vegetable soap formulation. Oscillatory shear rheology results are presented for these samples as a function of both moisture content and a compressional stress applied perpendicularly to the shear deformation in order to mimic the compressional forces that occur within the extruder during the production process. However, the range of stress and strain achieved on the rotational rheometer are lower than those typically observed in standard extrusion lines.¹⁹ For this reason, additional samples of pure syndet base and the reference vegetable soap were tested using a capillary rheometer, which better represents the extrusion process conditions.

We demonstrate that the increase in compressive stress in the axial direction, as occurring inside the extruder, has an impact on the rheological behavior of syndet and soap samples, and therefore, on their processability due to the induced stiffening of the material because of the imposed compressive stress; similar to what has already been shown for soft solids.^20–22 The relative changes of the pressure at the head of the extruder can be controlled dynamically by driving the screw speed, or statically by the choice of refining mesh or screw design. In this work, we also provide experimental evidence that the moisture content has an impact on the yield stress of the bars.²³ 

A summary of all samples used in this work and their key characteristics is presented in Table I. Samples were prepared by collecting extruded billets (i.e., cut blocks of predefined length) of pure syndet base Syndopal 300MB (Stephenson Group Ltd., Leeds, UK) from five different batches (SB1–5) with varying moisture content. Syndopal 300MB is a synthetic detergent base that is composed of sodium cocoyl isethionate, hydrogenated vegetable oil, water, polyglyceryl-4 laurate, glycerin, and tetrasodium glutamate diacetate, with the palm derivatives in compliance with the Roundtable on Sustainable Palm Oil (RSPO) Mass Balance (MB) certification.²⁴ In addition, two syndet bar formulations were tested: an aloe facile bar (AB1–2) and an oat body bar (OB1–2), both based on Syndopal 300MB. For the two syndet formulations, samples were collected at two stages of the production process. One set of samples was prepared from the coated base pellets after they had passed through the mixer with the other formulation ingredients but before extrusion. The second set of samples was collected from the billets after the mixture had passed through an extruder. Finally, a reference set of samples (BS1–2) were collected from the extruded billets of a basic soap formulation containing minimal additives (OPAL SP-SG soap base, Stephenson Group). The variation of the moisture content between samples, as reported in Table I, is due to different factors: (i) batch-to-batch variability of the base; (ii) the different formulation composition, which requires various additives to be mixed with the base; (iii) storage and (iv) transportation conditions. For the purposes of oscillatory shear rheology measurements, all samples were shaped into disks of $∼$2–2.5 mm thickness and 15 mm in diameter to match the geometries of the rheometer's tool. The disc-shaped samples were prepared by heating the material mass to 40 $°C$ in an incubator and then pressing it into a custom-made mold. The excess material was manually trimmed to leave a clean disc-shaped sample. Due to the small sample-to-sample variability introduced in the latter step, the initial height of the samples was determined by reading the gap size between the two rheometer plates at which the normal force transducer was reading a value equal to or slightly higher than its threshold value of 0.02 N. In addition to those described above, two additional sets of samples (i.e., pure syndet base and basic soap formulation) were prepared for the purposes of capillary rheometer measurements. In particular, a set of cylindrical shape samples was prepared by directly extruding the materials though a 15 mm diameter plate, thus matching the geometries required for loading the capillary rheometer. Samples for rotational rheology measurements were also taken from the same batches in the same way as described above.

The linear viscoelastic (LVE) properties of a generic material can be described by means of its shear complex modulus,

which is a frequency-dependent (⁠$ω)$ complex number whose real [$G′ω$] and imaginary [$G″ω$] parts, respectively, capture the elastic and viscous response of the material under investigation (these are also commonly referred to as storage and loss moduli, respectively).²⁵ Bulk rheology is the conventional approach to measure the LVE properties of a material at the macro-scale, which consists in applying an oscillatory shear stress of the form

and measuring the resulting oscillatory shear strain,

where $τ0$ is the amplitude of the stress function applied at frequency $ω$⁠, $γ0(ω)$ is the amplitude of the resulting frequency-dependent strain function, and $ψω$ is the frequency dependent phase lag between the applied stress and the resulting strain, which defines the sample's viscoelastic response [0 $<ψω<π2$⁠, where 0 represents an ideal elastic solid and $π2$ an ideal viscous liquid, respectively]. In the specific case of sinusoidal stresses and strains, it can be shown that

To quantify the LVE properties of the material of interest, two kinds of tests were performed:

1. Amplitude sweep: this test consists in applying gradually increasing amplitudes of shear strain at a fixed frequency of oscillation and it is used to determine the deformation response of a material. The materials' LVE regime is the region of strains within which $G′ω$ and $G″ω$ are independent of the applied strain.²⁶ From these tests one could obtain a measure of the yield stress either by taking the cross over point between the two moduli²⁷ or like in this paper by identifying the ordinate (i.e., the stress value) corresponding to the maximum value of the function obtained by multiplying the measured storage modulus by the applied strain (i.e., $G′γ$⁠, which represents the elastic stress) vs the strain itself (⁠$γ$⁠), as shown in Fig. 3(b) and discussed in the next section.

2. Frequency sweep: this test consists in applying a fixed shear strain amplitude within the materials' LVE regime at a gradually increasing frequency of oscillation and are therefore used to determine the frequency-dependent response of a material.²⁸ 

The above concepts are described graphically in Figs. 1 (b) and 1 (c).

In this work, bulk rheology measurements were performed using a single head rheometer (MCR302, Anton Paar) equipped with a parallel plate geometry [Fig. 1(a)], with an upper plate diameter of 15 mm, and at a temperature of 40 $°C$⁠. Before testing, samples were left on the heated plate at 40 $°C$ for a minimum time of 5 min to allow the sample temperature to become uniform. Strain sweeps in the range of 0.001%–0.1% and angular frequency 10 rad/s were performed at different axial stresses (induced by decreasing the gap h between the parallel plates). For all samples, strain sweeps were performed at compressive stresses ranging from a few kPa to $∼$ 300 kPa (i.e., within the rheometer's limit given a plate of 15 mm), with one strain sweep per level of compression and a delay of a few seconds to allow the sample to stabilize. In addition, for each tested sample two frequency sweep measurements were performed at low and high compressional axial stress. Measurements were performed at the University of Glasgow rheology lab in the UK.

Furthermore, two samples, SB5 and BS2, were tested using a twin bore capillary rheometer (Rosand 2000, Netzsch, Germany). Sample volume was approximately 40 cm³ for each barrel and extruded under isothermal conditions maintaining temperature of 40 °C. Die geometries used for the measurements were 1.0 × 16 and 1.5 × 24 mm² with 180° angle for the BS2 sample and 2.0 × 32 mm² with 180° angle for the SB5 sample. The capillary rheometer measurements were performed at NETZSCH-Gerätebau GmbH, Germany. A schematic of a capillary rheometer and its operating principles are shown in Fig. 2. The apparent wall shear stress (⁠$τwapparent)$ and shear rate $(γ̇wapparent)$ are given by the following equations:²⁹ 

where (P₀ – P_l) is the pressure drop along the capillary, R and L are its radius and length, Q is the volumetric flow rate.

In order to take into account the non-Newtonian behavior of the tested samples, the Rabinowitsch correction was applied to the results from the experiments to yield the true shear rate (⁠$γ̇true)$⁠, as given by the following equation:³⁰ 

where n is the power law constant given as

The shear moduli for all samples were measured during amplitude sweep tests as discussed above. Figure 3(a) shows representative strain response curves for the storage and loss moduli of a single syndet base sample (SB2, 8.1% moisture). Both moduli show an increase of up to an order of magnitude in response to the applied compressive stress in the axial direction. Interestingly, this behavior was common to all the tested samples as corroborated by the results shown hereafter. In Fig. 4, we present a comparison between the storage modulus, loss modulus and yield stress as functions of the applied compressive stress for four samples having different moisture content. For the viscoelastic moduli, each data point was obtained by taking their average within the apparent linear viscoelastic regime of the material, where the moduli are independent of the strain amplitude [i.e., over the first five points in the LVE region of the strain sweep, as schematically shown in Fig. 1(c)]. The yield stress was determined by identifying the ordinate (i.e., the stress value) corresponding to the maximum value of the function obtained by multiplying the measured storage modulus by the applied shear strain (⁠$G′γ$⁠) and plotting it vs the shear strain (⁠$γ$⁠) itself, as shown in Fig. 3(b). Notably, this approach is more conservative than the one in which the crossover point between the two moduli (occurring within the materials' non-linear regime) is often taken as a reference for identifying the materials' yield stress.²⁶ Indeed, the latter occurs at higher strain values than those measured here, which are instead closer to the materials' LVE regime.³¹ It is important to highlight that, several methods of estimating the yield stress have been reported in the literature, many of which use values of $G′ω$ and $G″ω$ obtained outside the LVE regime, like the approach used in this work; nonetheless, these values represent a leading order description of the materials' elastic response to the imposed oscillatory perturbation.²⁶ From Fig. 4, it is apparent that there is a strong dependence of the mechanical properties of the syndet base material on the applied compressive stress regardless of the moisture content, with all the measured parameters spanning more than two orders of magnitude as the compressive stress is varied by $∼$ 300 kPa.

From Fig. 4(c), it is possible to infer that the moisture content of the tested samples has a minimal impact on the material properties at relatively low compressive stresses, whereas at high compressive stresses, a higher moisture content appears to result in a lower yield stress. At this point, it is important to highlight that these results are limited by the maximum normal force that the instrument can apply to the sample (i.e., 50 N), which translates to $∼$ 0.3 MPa, given the geometry of the used tool. The maximum compressive stress in the axial direction achieved in this work is significantly lower than that typically observed on cleansing bar production line, where we observed pressure values of 1–5 MPa.

Base pellets are not typically extruded on their own and syndet cleansing bar formulations typically include additional ingredients such as moisturizers, coloring, fragrance, and other additives.^8,11 Therefore, in order to gain a better understanding of the effect of these additives on the rheological properties of the materials, and thus provide valuable feedback to the production process of these complex materials, measurements similar to those presented in Fig. 4 were taken on samples collected at different stages of the production process. These measurements were performed for each syndet formulation and the results for the aloe bar are presented in Fig. 5, whereas those for the oat bar formulation are shown in Fig. 6. Furthermore, Fig. 7 shows an overall comparison between the pure syndet base, the two syndet bar formulations and the basic vegetable soap formulation taken as a reference. It is interesting to notice that the results shown in both Figs. 5 and 6 indicate a reduction of the mechanical properties of the pellets as soon as they are “fully mixed” with the additives during the extrusion process. Once again, the compressive stress dependence of the results is clear as both moduli and the yield stress span more than two orders of magnitude for the applied compressive stress range. Notably, while for both the aloe bar and the oat bar samples, the presence of additives in the formulation lowers the value of both moduli, the yield stress appears to be almost unaffected. Interestingly, although crucial for the quality of the final product, the homogenization achieved in the extruder does not appear to have a significant impact on the mechanical properties of the materials in both formulations.

Similarly, by comparing the different materials in Fig. 7, it can be seen that all bar formulations have lower moduli than the pure syndet pellets with the biggest difference being observed in the storage modulus $G′.$ Moreover, it is interesting to notice that at higher stress loadings, the basic soap formulation exhibits higher yield stresses values than the syndet samples.

As discussed earlier, the ranges of shear rate and pressure typically measured during the extrusion process of cosmetic bars are much higher than those achieved in the rotational rheometer. A more realistic range of shear rates and stresses can be achieved by means of capillary rheometer measurements. Figure 8 shows a comparison between capillary and rotational rheometer measurements performed on both pure syndet base samples (SB5) and basic soap samples (BS2). In particular, the results obtained from the two rheometers are compared on the same plot on the basis of the Cox–Merz empirical rule,^32,33 where the dynamic viscosity $η′=G″/ω$ obtained via small amplitude oscillatory measurements (performed at both the lowest and the highest applied axial stresses) are compared with the capillary (shear) viscosity (⁠$η$⁠). From Fig. 8, it is clear that, for both the examined samples, the disparity between the shear and the dynamic viscosities grows as the compressional stress applied perpendicularly to the shear deformation during the oscillatory measurements increases. Thus, leading to a plausible speculation suggesting a potential agreement between the two methods if the oscillatory measurements were performed in the absence of a normal load. Nonetheless, it is interesting to note that, in the case of capillary measurements at relatively high shear rates, the viscosities of both the samples are very similar; possibly inferring the existence of a critical shear rate (here at $γ̇true≈20 s−1$⁠), below which the shear hardening phenomenon reported in Figs. 3–7 plays a major role.

Being the preferred choice for providing cleansing, antimicrobial protection, and overall skin hygiene to healthy, compromised, and diseased skin alike, solid bar cosmetic products based on synthetic detergents (“syndets”) are potentially a valuable alternative to their functionally equivalent liquid counterparts as well as classical vegetable soap alternatives. However, despite having a positive impact on the Green Economy globally in several aspects, syndets' complex (solid-like) nature has proven to be a hurdle to overcome in their manufacturing process. In this work, we have performed a rheological characterization of a set of syndet formulations as a function of both their moisture content and of a compressional axial stress applied perpendicularly to the shear deformation. Within the range of explored parameters, our investigation reveals that syndets show a compressional shear stiffening by up to two orders of magnitude and a slight reduction in yield stress as a function of increasing moisture content. Moreover, when these results are compared to nonlinear shear measurements performed with a capillary rheometer (i.e., by appealing to the Cox–Merz empirical rule), there exist a disparity between the outcomes of the two approaches that widens as the compressional stress applied during the oscillatory measurements increases; thus, corroborating the complex rheological character of these (solid-like) systems and potentially explaining the difficulties encountered by the manufacturing companies either during their production or while attempting to predict their flow behavior. Nevertheless, to fully understand this complex behavior and devise solutions to overcome the production issues caused by it, further work is needed. Additional factors affecting the behavior of syndet-based formulations, such as temperature variations and the role of individual ingredients, need to be further investigated.

The authors thank Adrian Hill and Shona Marsh for helpful conversations. The authors acknowledge financial support from Soapworks Ltd., EPSRC Impact Acceleration Account (No. EP/R511705/1) and Knowledge Transfer Partnerships (KTPs) (No. 12905).

The authors have no conflicts to disclose.

Giuseppe Ciccone: Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Visualization (equal); Writing – original draft (equal); Writing – review andediting (equal). Simeon Skopalik: Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Visualization (equal); Writing – original draft (equal); Writing – review and editing (equal). Claire Smart: Conceptualization (equal); Project administration (equal); Resources (equal); Supervision (equal). Senol Gezgin: Investigation (equal). David Ridland: Supervision (equal); Writing – original draft (supporting); Writing – review and editing (supporting). Manosh C Paul: Funding acquisition (equal); Project administration (equal); Supervision (equal). Maria del Pilar Noriega Escobar: Conceptualization (equal); Formal analysis (equal); Resources (equal). Manlio Tassieri: Conceptualization (lead); Data curation (lead); Funding acquisition (equal); Project administration (lead); Supervision (equal); Writing – original draft (equal); Writing – review and editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.