The defatted seed meal powder of C. oleifera (15 g) was extracted with boiling water (450 mL) three times and the water extract was extracted successively with ethyl acetate and n-butanol. The percentage of crude saponins extract that was obtained from the defatted seed meal of C. oleifera was 8.34. The total saponins content in the crude saponins extract was 39.5% (w/w). Xi et al. [16] extracted the crude saponins from 11 commonly used saponin-rich traditional Chinese medicines (Acanthopanax senticosus, Anemarrhena asphodeloides, Aralia taibaiensis, Asparrausi officinalis, Astragalus membranaceus, Ophiopogon japocicus, Panax ginseng, Panax notoginseng, Polygala tenuifolia, Polygonatum odoratum and Poria cocos). The yields of crude saponins extract were all less than 6% (1.12%, 1.57%, 0.78%, 5.48%, 3.42%, 1.08%, 0.96%, 0.95%, 3.63%, 0.86% and 2.35% (w/w), respectively). The total saponins content were all less than 46%. These results show that the saponins content in the defatted seed meal of C. oleifera is higher than other traditional Chinese medicines.

As shown in Figure 1, foam heights increased with increase in the aqueous solution concentrations. Among the three tested substances, sodium lauryl sulfate (SLS) ranked the highest for the foam height of the 0.5% solutions. The foaming height of the 0.5 crude saponins extract solution from C. oleifera was 37.1 of 0.5 SLS solution and was 51.3% with that of the 0.5% polyoxyethylene sorbate monooleate (Tween 80) solution. These results showed that 0.5% crude saponins extract solution from the defatted seed meal of C. oleifera possessed moderate foam power ability. Although foam generation has little to do with the cleansing ability of the detergents, it is extremely important for the user and is therefore an important criterion in evaluating detergents [15].

Foam produced by mechanical agitation is typically an unsteady thermodynamic system. When the foam set is resting, it will decay. The rate of weakening defines the stability of the foam [17]. The change of the foam height versus the time is shown in Table 1. The general aspect of the curves obtained shows little difference, which reflects a good foam stability versus time for 0.5% SLS, Tween 80 and the crude saponins extract solutions. Foam with R5 values higher than 50% can be regarded as metastable [18]. The R5 value of 0.5% saponins solution was 86.0%, representing the good foam stability of the saponins from the defatted seed meal of C. oleifera.

A detergent is something that increases the ability to displace air from a liquid or solid surface. The wetting phenomenon has aroused considerable commercial interest and plays a vital role in the removal of soil, dye, lubrication and printing by washing. The wetting ability of surface-active agent is commonly used to determine their comparative detergent efficacies. The Draves wetting data in Figure 2 shows that the wetting time decreased sharply with a solution concentration between 0.01–0.1%, followed by a slight drop after 0.1%. In terms of the wetting time of the 0.5% aqueous solution, the crude saponins extract solution needed 10.9 minutes to penetrate the cotton yarn while the SLS took just 0.1 minute. These data indicate that the wetting ability of the crude saponins extract of the plant was weaker than the general detergent, SLS.

Wetting phenomena are complex and depend upon several processes and factors such as diffusion, surface tension, concentration and the nature of the surface being wet. Each wetting agent has to reduce surface tension. The reduction in the surface tension of water from 72 mN/m to 32–37 mN/m with the use of commercial shampoo is considered as a good detergent [15]. As shown in Figure 3, the surface tension of 0.5% aqueous solution dropped from 72.0 mN/m to 35.6 mN/m by SLS, to 41.7 mN/m by Tween 80 and to 50.0 mN/m by crude saponins extract, respectively.

The data indicated that crude saponins extract has the potential to be an effective detergent. However, for commercial application, a more direct examination of detergent ability should be made. The sebum removed from 0.5% SLS, Tween 80 and crude saponins extract solutions were 90.4%, 77.6% and 53.8%, respectively (Figure 4). These results indicate SLS has excellent detergency. Crude saponins extract and Tween 80 also show moderate detergency. According to the data from Yang et al [19], the crude saponin from Sapindus mukorossi also possessed moderate detergency (60%). It seems that the detergent ability of the saponin from S. mukorossi is superior to the saponin from C. oleifera. However, the saponins contents of them are different. The saponins content of the S. mukorossi crude saponins (85%) is double than the saponins content of the C. oleifera crude saponin (39.5%)

The defatted seed meal of C. oleifera was purchased from Golden-Flower Tea Oil Company (Miaoli, Taiwan). This sample was originally collected from the mountains of central Taiwan in Miaoli County. The sample was milled to 100-mesh size in a Cyclone Mill (Tecator AB, Hoganas, Sweden). Quillaja saponin, inorganic salts and all other chemicals were from Sigma (St. Louis, MO, USA).

The total saponins content of extract was determined by the vanillin-sulfuric acid method [20]. This extract was mixed with vanillin (8%, w/v) and sulfuric acid (72%, w/v). The mixture was incubated at 60 °C for 10 min, cooled in an ice water bath for another 15 min, followed by absorbance measurement at 538 nm. Quillaja saponin was used as a reference standard [21] and the content of total saponins was expressed as Quillaja saponin equivalents (QS μg/mg extract).

The method used for measuring foaming power and foam stability was developed by Ross and Miles [22]. A portion of the test solution was placed in a jacketed cylinder. Foam developed when a stream of a second portion of 200 mL test solution was added to the first portion of the test solution through a standard orifice from a 90 cm height. This resulted in turbulence and foam. The height of the foam generated was measured immediately and again after 5 min. The foam height at the initial stage indicates the foam power of the surfactant solution. The parameter R5, defined as the ratio of the height of the foam at 5 min to that at the initial stage, is proposed as an evaluation of foam stability [18].

The method used to measure wetting ability was developed by Draves [23]. In the test, a five gram skein of gray cotton yarn is submerged in the test solution. The time for the air in the yarn to be replaced by penetration of the solution is recorded. The end point is observed as the moment when the skein sinks. The unit of wetting ability is minute.

Surface tension of the prepared surfactants was measured at room temperature (25 °C) using a Du Nouy tensionmeter (Sigma 703, KSV Instruments, USA), along with a 0.5% solution (w/v) in distilled water [24]. The surface tension of the added distilled water was 72.0 mN/m. The surfactants were aged for 30 min before further measurements.

The detergent ability was evaluated by a slightly modified version of the method developed by Thompson [15,25]. Hair tresses were obtained from a beauty salon. The tresses were prewashed with 5% SLS solution, dried and cut into 10 inch, 3 g swatches. The hair swatch (3 g) was suspended in 20 mL of 10% sebum solution (olive oil 20%, coconut oil 15%, stearic acid 15%, oleic acid 15%, paraffin wax 15% and cholesterol 20%) in hexane for 15 min with intermittent shaking. The swatch was removed, the solvent evaporated at room temperature and the dried hair swatch weighed to determine the sebum load. Each swatch was then split into two equal samples of 1.5 g each: one for the surfactant treatment and the other to act as an internal control to overcome the tress-to-tress variation in soil levels. The control swatch was left untreated. The test swatch was washed with 100 mL of the surfactant solution by the Finger Method described by Thompson et al. [25]. It was then dried using a hair dryer and further dried in an oven at 60 °C for 4 h to ensure uniform moisture content. The sebum remaining in the test swatch after surfactant treatment and that in the unwashed control one was then extracted using 20 mL of hexane in a stopper flask for 30 min on a rotary shaker. The hexane solution was then evaporated to dryness and the sebum extract from the test and control swatches was weighed. The detergent ability was evaluated as a percentage of sebum removed after surfactant treatment.

All analytic measurements were taken at least in triplicate. Data are expressed as mean ± SD. Statistical comparison of means and simple correlation coefficients were conducted using the Student’s t-test in a general linear model (GLM) procedure of an SAS system (SAS Institute, Cary, NC).