Formulating liquid detergents for multiple enzyme stability; incorporating enzymes in your liquid laundry detergent can be a difficult task. Cognis researchers explain how.

ALTHOUGH USE of enzymes in detergents was first described by Otto Rohm in a 1913 patent application, (1) the use of enzymes in laundry detergents was not widespread until the 1960s due to poor stability under typical wash conditions. Development of proteolytic enzymes resistant to alkali and high temperature and improved encapsulation techniques have accelerated the use of enzymes in powder formulations during the past three decades.

Stabilization of enzymes in liquid laundry detergents is more difficult than in powders. In liquid systems, enzymes are easily denatured by detergent ingredients. Alkalinity, high water content and surfactant interactions are all capable of changing the three dimensional conformation of the protein. Protease enzyme is also prone to auto-digestion effects leading to poor formulation stability. Since the early 1990s, the use of enzymes in liquid detergent systems has become more common due to advances in stabilization technology.

Early patents by Procter & Gamble disclosed the use of polyacids and free calcium ions to stabilize protease enzymes. (2-4) It is postulated that these materials are able to bind with specific amino acids in the protein to retain the tertiary structure of the molecule in aqueous solutions. Other early patents by Unilever disclosed the use of polyols in combination with poly-functional amino compounds or alkali metal borates to stabilize mixed enzyme systems. (5-6) It is likely that all of these materials function in a similar fashion.

It is also known that the types of surfactants used in a liquid detergent formulation can have a significant impact on enzyme stability. Studies performed at Shell Development Company in 1980s showed that enzymes were considerably more stable in formulations containing alcohol ethoxylates and alcohol ether sulfates than those containing alcohol sulfates and sulfonates. (7) Enzyme stabilizer systems were only partially effective in reducing enzyme destabilization by alkyl sulfated and sulfonated surfactants.

To better understand surfactant/ enzyme interactions in liquid detergents systems, Cognis studied the effects of different surfactant types on enzyme stability and performance in a model liquid laundry detergent formulation. Using an experimental design approach, quaternary surfactant mixtures composed of fatty alcohol ether sulfate (FAES), linear alkylbenzene sulfonate (LAS), alcohol ethoxylate (AE) and alkyl glucoside (AG), were formulated with protease, lipase and cellulase enzymes. Amylase enzymes, which are commonly used in laundry detergents, were not included in this study because stability in liquid systems is typically not a concern. The effect of each surfactant type on enzyme stability and performance was determined. From this information, liquid detergent surfactant compositions can be optimized for multiple or specific enzyme stability as well as detergency.

Design and Experiment

The surfactants used in these studies are given in Table 1. All of the surfactants are commercially available and were used as supplied without further purification.

In our studies, three different types of enzymes were used. These enzymes, listed in Table 2, are also commercially available and were used as supplied. For liquid detergent systems, the enzymes are typically dissolved in propylene glycol and stabilized with calcium chloride.

The activity of each enzyme in the formulation was determined initially and as a function of time at 40[degrees]C. Elevated temperature was used to accelerate the denaturation process and to simulate common storage conditions. Standard spectrophotometry assay methods determined enzyme activity. In general, the stability of the enzyme is expressed in percentage units based on initial enzyme activity.

The activity of protease enzyme was determined using the degradation reaction of suc-L-Ala-L-Ala-L-Pro-L-Phe-p-nitro anilide (AAPF-pNA) to liberate p-nitroaniline. (8) The activity of lipase was determined using the esterolytic degradation of p-nitrophenyl palmitate by lipase to generate colored p-nitrophenol. (9) The activity of cellulase was determined by following the hydrolysis reaction of carboxy methyl cellulose to glucose and reacting it with p-hydroxy benzoic acid hydrazide (PAHBAH) to form a yellow dye. (10-11)

Enzyme stability and performance were determined in a formulated laundry detergent blend as a function of time at an elevated temperature (i.e., 40[degrees]C) using an experimental mixture design. The model formulation and percent weight ranges for each component is given in Table 3. The total surfactant concentration was held constant at 26% actives.

In general, laundry detergents are formulated using a combination of anionic and nonionic surfactants as the surfactant system. Nonionics give good detergency on oily soil while anionics give good detergency on particulate soils and contribute to formulation stability. In U.S. formulations, anionics are typically present at higher levels than nonionics. In addition to the surfactants, other ingredients were added to the model formulation. Coconut fatty acid acts primarily to control foam and soften water. Monoethanolamine (MEA) neutralizes the fatty acid and acts as an alkalinity source to help remove fatty acid/ester based soils. Ethanol is added as a hydrotrope to improve formulation stability and lower viscosity. A propylene glycol/borax mixture helps stabilize the enzymes.

We found in preliminary studies that the minimal level of propylene glycol/borax (ratio 7:1) required to stabilize enzymes at the mid-range surfactant concentration of the design study was 15%. To prepare addition of the enzymes, sodium borate (borax) is first dissolved in warm propylene glycol, the mixture was cooled and the enzyme preparations were added. The enzymatic/stabilizer mixture was added to the surfactant base buffered at the appropriate pH. In our experiments, all of the formulations were buffered with MEA to a pH between 8 and 9.

To assess enzyme performance, the Terg-o-tometer method was employed. Standard enzymatic soil cloths were washed in each formulation under carefully controlled use conditions. The optical reflectance of the soil cloth was measured before and after washing. The reflectance of the cloth was measured using a Hunter Labscan colorimeter. The difference in reflectance before and after washing gave a measure of enzyme performance.

Results and Discussions

To determine the effect of these four surfactants in our model formulation (Table 3) on enzyme stability, a total of 19 samples with different surfactant blend ratios were prepared. The enzyme stability and the performance were determined, and the raw data was analyzed using experimental design software to obtain a better understanding of how enzyme stability changes as a function of surfactant blend ratio. (12)

The effect of each surfactant on protease stability at 28 days is shown in the perturbation plot in Fig. 1. The perturbation plot indicates how protease stability varied as a function of surfactant type and concentration while holding all other surfactants constant at the mid-point concentration of the design range. For protease, stability increased with increasing concentration of FAES (represented by line A) and decreased with increasing concentration of LAS (represented by line B). AG and AE (represented by lines C and D, respectively) have a neutral effect on stability over the region covered in the design study. Another finding was that at high levels of FAES, acceptable stability can be obtained even in the presence of LAS.

[FIGURE 1 OMITTED]

To test the performance of protease in these formulations, detergency on EMPA 116 and 117 enzymatic soil clothes was used. These EMPA test cloths were stained with a combination of blood, milk and carbon black on cotton and polyester/cotton blends respectively. The cloths were washed with the 19 model formulations using the Terg-o-tometer method. The reflectance of the cloths before and after washing was used as a measure of enzyme performance. The perturbation plot for EMPA detergency is shown in Fig. 2. The performance increased with increasing concentration of FAES and decreased with increasing concentration of LAS. Again, AG and AE have a neutral effect. These results are qualitatively similar to that observed for protease stability. In general, better stability translates to better stain removal.

[FIGURE 2 OMITTED]

The effect of surfactant composition on lipase stability is shown in the perturbation plot in Fig. 3. Increasing concentration of AG or AE increases lipase stability while both FAES and LAS have a strong destabilizing effect on lipase. It was found that samples high in FAES or LAS showed very high initial enzyme activity which decreased rapidly with storage time. Other investigators have found a similar effect with LAS and SDS. (13) Circular dichroism measurements confirm that the interaction caused by LAS is due to conformational changes in the protein structure.

[FIGURE 3 OMITTED]

The effect of surfactant composition on cellulase stability is shown in Fig. 4. Increasing concentration of FAES, AG or AE improved cellulase stability while increasing concentration of LAS resulted in decreasing stability. At high levels of LAS, poor enzyme stability was obtained regardless of surfactant blend composition.

[FIGURE 4 OMITTED]

The experimental design results can be used to predict the optimum surfactant composition for multiple enzyme stability. The stability results for each enzyme were optimized simultaneously to obtain the composition that gives the best overall stability for all three enzymes. The result of this treatment is shown in Fig. 5.

[FIGURE 5 OMITTED]

The orange area in Fig. 5 represents the composition giving the best stability for all three enzymes. This composition was defined as AE 13%, FAES 6.5% and AG 6.5% with an optimum surfactant blend ratio of 2:1:1 for AE:FAES:AG, respectively. In addition, this result indicated that the best stability in our model formulation was obtained when LAS was minimized.

We have discussed the impact of surfactant type and concentration on enzymatic activity. Using this information we can predict surfactant mixtures with good stability for multiple enzymes or specific enzymes. Examples are given in Table 4.

From this information, we have also developed various complete enzymatic liquid laundry formulations that exhibit good performance and stability. An example of one composition with premium performance is found in Table 5 on the next page. Using a FAES rich formulation and AE/AG as co-surfactants, this formulation gave very good cost performance against market leading products and exhibits excellent enzyme stability.

Conclusions

After determining the minimum level of stabilizer required in our model formulation, an experimental mixture design was used to determine the effect of different surfactants on enzyme stability. For protease enzyme, stability increased with increasing concentration of FAES. LAS had a strong destabilizing effect on the enzyme. This indicated that LAS can bind to the enzyme and change its conformation. AE and AG showed a neutral effect on protease stability. Qualitatively, the same trend was observed for protease enzyme performance as determined by detergency measurements.

For the lipase enzyme, both FAES and LAS as anionic surfactants had an adverse effect on long-term stability. It appears that anionic surfactants can bind to the protein changing the three dimensional conformation of the polymer. The stability of lipase increased with increasing concentration of AG and AE nonionic surfactants in the formulation.

For cellulase, LAS again has a strong destabilizing effect on the enzyme. Cellulase stability increased with increasing concentration of FAES, AE or AG. The difference in behavior between LAS and FAES may relate to the difference in the anionic strength of each surfactant.

The above stability results, when taken together, can be used to determine the optimum surfactant compositions for multiple enzyme stability or stability for specific enzymes. These results indicated that optimal stability for multiple enzymes can be achieved with a surfactant blend ratio of 2:1:1 for AE, FAES, and AG. For specific enzymes, such as protease or protease/amylase system, FAES-rich compositions should be considered with AG and AE as co-surfactants.

Table 1: Surfactants used in the present studies

Surfactant Class                     Chemical description

Fatty Alcohol Ether Sulfate (FAES)   Sodium salt of 3 mole ether
                                     sulfate based on [C.sub.12-16]
                                     fatty alcohol

Alkyl Benzenesulfonate (LAS)         Sodium salt of a high 2-phenyl
                                     alkylbenzene sulfonate

Alcohol Ethoxylate (AE)              [C.sub.12-15] oxo alcohol with
                                     7 moles EO

Alkyl Glucoside (AG)                 [C.sub.12-16] fatty alcohol
                                     with a DP of 1.4; DP = avg.
                                     glucose units

Table 2: Liquid enzymes used in the present studies

Enzyme      Description

Protease    Serine-type protease characterized
            by excellent performance at elevated
            pH that hydrolyzes peptide bonds
            in proteins; good for blood, egg
            and grass stain removal.

Lipase      1,3-specific lipase with broad sub-
            strate specificity that hydrolyzes
            ester bonds in fats and oils; good
            for removal of salad oils, sebum
            and lipstick removal

Cellulase   A mono-component cellulase that
            hydrolyzes cellulose; good for
            removing micro-fibrils from cotton
            resulting in brighter colors and
            less pilling.

Table 3: Experimental laundry detergent--model formulation

Ingredients                    % wt.    Function

FAES                           0-19%
LAS                            0-19%    Surfactants
AE                             0-13%    (26% total active)
AG                             0-13%

Coconut fatty acid             4%       Co-builder/foam control
Monoethanolamine               1%       Buffer
Sodium sulfate                 0.1%     Viscosity
Ethanol                        4.5%     Hydrotrope
Propylene glycol/borax (7/1)   15%      Enzyme stabilizer

Protease                       0.75%
Lipase                         0.75%    Enzyme system
Cellulase                      0.75%

DI water                       QS 100

Table 4: Examples of surfactant systems with good enzyme
stability in premium liquid laundry formulations

              Protease/Amylase   Multiple Enzyme
Ingredients         % Wt              % Wt

FAES                16.0               6.5
AE                  5.0               13.0
AG                  5.0                6.5

Protease            0.75              0.75
Amylase             0.40              0.40
Lipase                                0.75
Cellulase                             0.75

Table 5: A premium liquid laundry formulation with excellent
enzymatic stability and performance compared to market
leaders.

Ingredients          % Wt        Function             Trade Name

Water                53.36
Boric acid           1.10        Enzyme stabilizer
Sodium gluconate     0.70        Enzyme stabilizer
Propylene glycol     3.00        Enzyme stabilizer
EtOH 3A              3.00        Hydrotrope

AG (50%)             5.80        Surfactant           Glucopon 625 UP
AE                   5.20        Surfactant           Neodol 25-7
FAES (70%)           25.00       Surfactant           Texapon N-70

Optical brightener   0.14        UV whitening agent
Sodium hydroxide,    0.50        Neutralizing agent
  50%
Monoethanolamine     0.50        Buffer

Protease             0.75        Enzyme               Savinase 16.0L
Amylase              0.95        Enzyme               Termylase 300L

Preservative/        as needed
Optical brightener

Note: Add in order listed.

Acknowledgements

Compiling, understanding and interpreting the results for this article required significant effort from many contributors. The authors would like to especially thank former colleagues Dr. George Smith and Dr. Karl Heinz Maurer, whose contributions made this comprehensive study possible.

References

(1.) O Rohm, Germaon Patent DE 283,923 (1915).

(2.) Letton, et al, U.S. Patent 4,318,818 (March 9, 1982).

(3.) Barrat, C.R., et al, U.S. Patent 4,111,855 (Sept. 5, 1978).

(4.) Boyer, S.L., Farwick, T.J., U.S. Patent 5,476,608 (Dec. 19, 1995).

(5.) Tai, Ho T., U.S. Patent 4,404,115 (Sept. 13, 1983).

(6.) Boskamp, J.V., U.S. Patent 4,462,922 (July 31, 1984).

(7.) Shell Technical Bulletin SC: 814-84.

(8.) Del Mar, E.G., Largman, C., Brodrick, J.W., Geokas, M.C., Anal. Biochem. 99:316-20 (1979).

(9.) Winkler and Stuckmann, J. Bacterial. 138:663-70 (1979).

(10.) Lever, M., A new reaction for colorimetric determination of carbohydrates, Anal. Biochem. 47:273-79 (1972).

(11.) Lever, M., Carbohydrate determination with 4-hydroxybenzoic acid hydrazide (PAHBAH): Effect of bismuth on the reaction, Anal. Biochem. 81:21-27 (1977).

(12.) Design Expert 4 (STAT-EASE Inc., Minneapolis, MN)

(13.) Gormsen, E., Stentebjerg-Olesen, B., Lipolase, A Microbial Lipase for Detergents, 21st Detergent Symposium, JOCS (1989).