Computer-based method for determining sunscreen compositions containing multiple UV filter materials

Abstract

To provide a computer-based method for determining an optimal sunscreen composition comprising multiple UV filter substances on the basis of constraints and objectives.SOLUTION: A computer-based method for determining a sunscreen composition comprising multiple UV filter substances comprises the steps of: selecting a constraint for a characteristic of the composition including a sunscreen performance target; selecting an optimization objective; and automatically determining the sunscreen composition as a composition of filter substances from a set of filter substances, such that the composition meets the constraint and is optimized with respect to the selected optimization objective. The automatically determining step comprises the steps of: generating multiple candidate compositions; determining sunscreen performance of the candidate compositions using a performance simulation tool; and comparing the determined sunscreen performance of the candidate compositions with the sunscreen performance target.SELECTED DRAWING: None

Description

[Technical Field]

[0001] The present invention relates to a computer-based method for determining a sunscreen composition comprising multiple UV filter substances. The present invention further relates to a computer program product for determining a sunscreen composition and a method for producing the sunscreen composition. [Background technology]

[0002] Today, sunscreen developers have several computer tools available that can predict, to some extent, the sunscreen performance of UV filter combinations (see, for example, B. Herzog, U. Osterwalder: "Simulation of sunscreen performance", Pure Appl. Chem. 2015; 87(9-10): 937-951). Some tools are available online for general use (e.g., BASF® Sunscreen Simulator or DSM® SUNSCREEN OPTIMIZER™). Based on the input of a UV filter combination, these predict, among other things, the sun protection factor (SPF) and the ratio between the UVA protection factor and the SPF. These tools therefore allow developers to quickly evaluate and compare various UV filter combinations in silico.

[0003] However, the vast number of possible choices still means that searching for the best combination of filter factors by trial and error can be tedious, time consuming, and may not lead to true optimization. Summary of the Invention [Problem to be solved by the invention]

[0004] The object of the present invention is to create a method related to the first-mentioned technical field that allows for the efficient determination of optimal sunscreen compositions. [Means for solving the problem]

[0005] The solution of the present invention is specified by the features of claim 1. According to the invention, a computer-based method for determining a sunscreen composition comprising a plurality of UV filter substances comprises: a) selecting at least one constraint for at least one property, including a sunscreen performance target, of the composition to be determined; b) selecting an optimization objective from a plurality of optimization objectives; c) automatically determining, from the set of filter materials, a sunscreen composition as a composition of filter materials that satisfies at least one constraint and is optimized with respect to the selected optimization objective, said automatic determination comprising: - generating a plurality of candidate compositions; - determining the sunscreen performance of a candidate composition using a performance simulation tool; - comparing the determined sunscreen performance of the candidate composition with a sunscreen performance target; Includes:

[0006] As mentioned above, performance simulation tools for sunscreen compositions are available, see, for example, B. Herzog, U. Osterwalder: "Simulation of sunscreen performance", Pure Appl. Chem. 2015; 87(9-10): 937-951.

[0007] The method of the present invention allows for the automated determination of optimal sunscreen compositions based on constraints and objectives, thereby avoiding manual trial and error. As will be described in more detail below, even if sunscreen performance simulation tools are readily available, the effort required to optimize a composition of, for example, five or more filter substances is prohibitive for a manual approach. Therefore, by integrating performance simulation into an automated determination that includes determining the performance of multiple candidate compositions, it becomes possible for the first time to systematically and truly find optimality.

[0008] The possibility of selecting at least one constraint and optimization objective allows the individual formulation needs to be respected. The desired solution can be tailored to his or her specific requirements. As will be explained in more detail below, the computer-based method according to the present invention is fast enough for today's personal computers, which allow an interactive work process and provide feedback to the user in a reasonable time, preferably within a few seconds. This allows the user to quickly learn from the results and adjust the set of constraints, boundaries, or filter substances considered as needed and / or desired.

[0009] Preferably, the user is requested to select at least one constraint, which allows for a user-defined interactive process. In this case, the constraints can be selected according to the user's general requirements and according to the properties of the sunscreen to be optimized. In a simple embodiment of the invention, the user selects one value for sunscreen performance (e.g., the desired SPF). Nevertheless, instead of a target value, the constraint can be provided in the form of a range, minimum value, or maximum value. Furthermore, target values ​​or ranges for one or more properties of sunscreen performance can be provided.

[0010] Alternatively, the constraints are selected automatically by a computer, in particular by a fixed or variable constraint sequence in successive optimization steps to find a preferred sunscreen composition.

[0011] Preferably, the user is requested to select an optimization objective, which allows for a user-defined, interactive process, where the objective can be selected according to the user's general requirements and according to the properties of the sunscreen to be optimized, and the objective can be changed during the method, particularly in order to iteratively improve the sunscreen composition.

[0012] Alternatively, the optimization objectives are selected automatically by a computer, in particular, objectives according to a fixed or variable sequence of constraints in successive optimization steps to find a preferred sunscreen composition.

[0013] The constraints and objectives have a predetermined interrelationship. For example, the cost of a composition can be provided as a constraint (a maximum value that cannot be exceeded) or an objective (minimization of cost). Each of the other constraints and objectives is selected appropriately in each case.

[0014] Preferably, the sequence comprising steps a) to c) is repeated. The user iteratively adjusts at least one constraint and / or optimization objective. This allows for incremental refinement of the sunscreen composition, taking into account information obtained from previous optimization steps. The iterative process is feasible due to the fact that a sunscreen composition can be determined within a few seconds using the method of the present invention when implemented on standard hardware.

[0015] Alternatively, the method is designed so that the desired sunscreen composition can be found in a single step, and / or the adjustment of constraints and / or optimization objectives is performed automatically by the computer based on the initially selected parameters and the results of previous decisions. In a further embodiment, the computer automatically provides suggestions regarding criteria for further improvement of the composition, and the user decides whether to follow these suggestions fully or partially for the next step in the iterative process.

[0016] Preferably, the user selects the actual set of filter substances to be considered from a base set, or superset, of filter substances. In particular, this selection is based, for example, on the user's knowledge of the properties of the available substances and / or the availability of filter substances. This makes it easier to find the optimal composition for each task at hand. Furthermore, reducing the number of filter substances to be considered can significantly facilitate automated determination of sunscreen compositions.

[0017] Alternatively, all available filter materials are considered, or the selection of the actual set is made automatically by a computer, for example based on a database of filter properties and / or availability.

[0018] In a preferred embodiment, the user provides a maximum amount of at least some of the selected filter substances, and in particular a maximum amount of all selected filter substances, thereby ensuring that regulatory approved use levels are not exceeded. Furthermore, the sunscreen composition can be tailored to the user's needs.

[0019] Alternatively, the maximum amount is taken into account automatically, for example based on a database containing regulatory information.

[0020] In some embodiments, the user provides a minimum amount of at least some of the selected filter materials, which can influence the determination of the sunscreen composition to ensure that the user's needs are met.

[0021] Advantageously, the method includes the step of selecting at least one further constraint (in addition to the sunscreen performance target) for at least one property of the composition to be determined.

[0022] Preferably, at least one further constraint is one of the following properties: a) total amount of filter material; b) a quantity of one or more separate filter materials; c) the cost of the composition of the filter material; d) weighting of the composition of the filter or filter material; e) Environmental considerations; f) the amount of excess solvent; g) Oil Load, is a range or boundary value (upper or lower limit) associated with one of the following:

[0023] In principle, the range can be infinitesimal, ie correspond to a target value that is required to be met (exactly or within a given general tolerance).

[0024] The filter weighting is a number that can represent the price of the filter, a score for how environmentally friendly the filter is, a score for how easy the filter is to formulate, a score for the filter's impact on the sensory perception of sun protection, etc. The composition weighting of a filter material is the sum of the weightings of the individual filters multiplied by their corresponding percentages in this combination, and can thus represent, for example, the total cost of the filter combination, the total sensory impact, etc.

[0025] The oil load is the sum of all the filters in the oil phase plus any excess solvent that may be needed to dissolve the solid filters. This is a good measure of how much flexibility the formulator has for a particular filter combination in terms of adding additional ingredients in the oil phase that may be relevant to, for example, the skin feel of the final product.

[0026] Environmental considerations or ecotoxicity considerations relate to the environmental impact of filter materials. Methods for determining ecotoxicity values ​​are described, for example, in U.S. Patent No. 7,096,084 (SC Johnson & Son, Inc.), U.S. Patent No. 9,595,012 (Johnson & Johnson Consumer Inc.) and WO 2019 / 207129 (BASF SE).

[0027] The above list of possible constraints is not exhaustive, and further constraints are possible, such as those relating to the sensory properties of the sunscreen composition.

[0028] Preferably, the sunscreen performance goal is: a) internal or external sun protection factor (SPF); b) UVA Protection Factor (UVAPF) internally or externally; c) critical wavelength; d) the ratio of UVA protection to UVB protection; and e) Blue Light Protection, is selected from one of the following:

[0029] The goal may be provided as a target value or a target boundary (eg, a minimum value).

[0030] The internal or external sun protection factor (SPF) can be expressed as an absolute SPF value, as a recommended labelled sun protection factor or as a protection category according to, for example, EU Commission Recommendation 2006 / 647 / EC of 22 September 2006 or other (national or multinational) regulations on the labelling of sun protection products. Protection categories can be, for example, -Low protection: SPF is less than 15; - Medium protection: SPF is between 15 and 29; -High protection: SPF is between 30 and 49; - Very high protection: SPF is over 50, It can be defined as follows:

[0031] The in vivo UVA protection factor (UVAPF) can be expressed, for example, as a UVAPF determined in accordance with ISO 24442:2001, as a UVAPF protection grade according to the Japan Cosmetic Industry Association (JCIA) Voluntary Industry Standard for Measuring UVA Protection Effectiveness, or as a persistent pigment darkening (PPD) or immediate persistent darkening (IPD) value.

[0032] As usual, the critical wavelength refers to the wavelength at which the sunscreen transmits 10% of the light. For example, sunscreens with a critical wavelength above 370 nm are considered by the FDA to provide excellent UVA protection.

[0033] The ratio of UVA protection to UVB protection can be given as UVAPF / SPF or Boots Star® rating, a proprietary in vitro method introduced by Boots Company in 2011 to describe the protection offered by sunscreen products.

[0034] Blue light (often called high-energy visible light (HEV light)) protection refers to protection from light in the blue-violet portion of the visible spectrum, found in sunlight as well as LED and fluorescent lighting.

[0035] In principle, quantities relating to sunscreen performance can be determined according to any standard, as long as the performance simulation tool used is capable of providing sunscreen performance in accordance with the standard, i.e. the available targets may depend on the underlying simulation tool.

[0036] The above list of performance goals is not exhaustive and further goals are possible, including suitable combinations of the above.

[0037] Preferably, the multiple optimization objectives are: a) Cost-effectiveness; b) weighting; c) filtering efficiency; d) environmental considerations; e) the amount of excess solvent; f) Minimum oil load; g) the most homogeneous protection; h) highest sun protection factor and / or UVA protection factor; i) Best blue light protection; and j) the similarity of the filter material to the provided composition; Includes at least two of the following:

[0038] These objectives can be minimized or maximized.

[0039] Sunscreen simulation tools facilitate quick and easy side-by-side comparisons of various filter combinations. However, direct comparisons of characteristics such as efficiency or cost are only meaningful if all the compared combinations result in more or less the same performance. Comparing many combinations requires significant effort to accurately calibrate each combination.

[0040] In the context of the present invention, this adjustment process is greatly simplified by selecting a filter material similar to the (user) provided composition as the objective. The filter material composition is determined to be as close as possible to the (user) provided composition while achieving performance goals and satisfying constraints. Thus, the resulting adjustment will quickly adjust the filter concentration to meet the desired performance goals and constraints while remaining as close as possible to the user's underlying ideas.

[0041] Preferably, in this case, the objective to be minimized is the Euclidean distance between the provided composition and the determined composition, satisfying the constraints.

[0042] The list of objectives is not exhaustive and additional objectives are possible, including any suitable combination of the above properties or sensory properties of the composition.

[0043] Due to the vast number of possible UV filter combinations, a simple combinatorial brute-force approach is infeasible. Assuming an optimal combination of six UV filters is found (a very common number in the sunscreen industry), and the amount of each filter is quantified in discrete increments from 0.1 wt% up to 5.0 wt%, over 15 billion possible filter combinations result. Additionally, selecting these six filters from the 24 approved in Europe involves 134,596 options, giving this setup a total of approximately 2,100 trillion possible combinations. Typical personal computers available commercially today typically have CPU clock frequencies of 2–4 GHz, and it takes approximately 2–4 ms to determine the sunscreen performance of a given composition using a performance simulation tool. Therefore, the computational time to calculate all performance values ​​would exceed 130,000 years. Optimized code can reduce the computational time by a factor of 100, and the use of multiple cores can reduce the computational time by a factor of 8–10. However, even with all these measures, the computational time would exceed 100 years. Furthermore, extending the number of filters beyond the six that would be desirable to meet all requirements increases the computation time significantly.

[0044] Thus, in a first preferred embodiment, the automated determination of a sunscreen composition comprises the numerical optimization of an objective function with respect to a selected optimization objective, the variables of the objective function comprising the proportions of filter substances of the sunscreen composition to be determined.

[0045] Numerical optimization algorithms eliminate the need to determine the performance of a large number of candidate compositions and are powerful tools for finding optimal solutions in high-dimensional spaces.

[0046] Advantageously, the numerical optimization involves the application of sequential quadratic programming, in particular the application of interior point methods.

[0047] Numerical optimization algorithms require the constraint objective function to be convex in order to find a global optimum instead of a local optimum. Furthermore, these algorithms are more or less sensitive to the selection of a good starting point to ensure overall convergence. There is no general-purpose optimization algorithm; rather, there is a collection of algorithms, each tailored to a specific type of optimization problem. It is impossible to predict whether a problem will be solved quickly or slowly, or whether a solution will actually be found.

[0048] Surprisingly, a numerical optimization algorithm that evaluates the first and second derivatives of an objective function for the selection of the search direction and step size to be adopted at each iteration for the optimization of sunscreen compositions is not only capable of finding a global solution to the above objective, but is also fast enough on a regular personal computer to allow for an interactive work process.

[0049] The complexity of current algorithms used in sunscreen performance prediction tools requires estimation of first derivatives, for example, by finite difference methods for the calculation of the Jacobian and quasi-Newton methods, e.g., BFGS or SR1, to approximate the Hessian at each iteration.

[0050] When using numerical optimization to determine sunscreen composition, there is no limit to the number of UV filters, and therefore, filter preselection is not necessary for calculation reasons.In fact, high-speed calculation is possible even with more than 20 filters on a normal personal computer.In addition, there is no resolution limit, which means that the "true optimal" composition can be found.

[0051] Nevertheless, it must be considered that the numerical optimization approach may not work for all constraint objectives, i.e., it is not guaranteed that a global optimum will be found for all constraint objectives. Furthermore, the results will not be rounded to a reasonable number of decimal places. To obtain results at the desired resolution level (e.g., 0.1 wt. % increments), it may be necessary to use methods such as branch and join.

[0052] In a second preferred embodiment, for the automatic determination of a sunscreen composition, a plurality of candidate compositions is automatically defined, and the sunscreen performance of at least some of the plurality of candidate compositions is determined using a performance simulation tool.

[0053] Using a combinatorial approach, optimal (or near-optimal) sunscreen compositions can be found for any objective, with the solution guaranteed to be near the global optimum. Furthermore, a list of best solutions can be obtained without extra effort (see below).

[0054] A combinatorial approach can be realized by applying various means to reduce the number of calculations, in particular: 1) Reducing the number of UV filters by allowing the user to pre-select up to six filters; 2) Selecting an increment of the amount of filter to be 0.5-1.0 wt.%, most preferably 0.5 wt.%; 3) Reducing the search space, 3a) a. Conservatively estimating the total amount of filters needed to achieve a performance target (e.g., SPF / 2 wt% filters) and calculating only combinations with the corresponding total amount of filters (e.g., 6 filters for SPF 30 requires 312,620 calculations for 0.5 wt% increments and a maximum use level of 5 wt% per filter); b. Enumerating the corresponding filter combinations so that the most efficient filter combinations are tested first; c. Stopping the calculations once the target performance is met (typically less than 1000 calculations); d. Re-estimating the total amount of the filter according to the over-performance or under-performance of the first estimation, and re-calculating only the combination with the corresponding total amount of the filter; e. Iteratively decreasing the total amount of filters by a fixed number, e.g., 1% by weight, and recalculating all these combinations until the goal is no longer achievable (backward search); f. increasing the total amount of filter by increments (e.g., 0.5% by weight) (forward search) until the performance target is again achieved, stopping the search as soon as the target is achieved, thereby obtaining the lowest total amount of filter (most efficient filter combination) that still achieves the target performance; The minimum total amount of filters that can achieve the target performance is searched for by 3b) Starting with the lowest feasible total filter amount, search for the optimal solution by increasing the total amount by one increment, e.g., 0.5 wt. %, with each search; thereby reducing the search space; is.

[0055] Therefore, it is preferred that in a first sub-step a minimum total amount of filter substances is determined for a composition that achieves the sunscreen performance goal, and in a subsequent second sub-step the constraints on the value of the total amount of filter substances of candidate compositions are gradually increased, starting from the determined minimum total amount, until a stopping criterion is met.

[0056] As noted above, preferably, the minimum total amount of filter material is determined by gradually decreasing the total amount of filter material in a candidate composition tested for sunscreen performance until the sunscreen performance target is no longer attainable, and then gradually increasing the total amount of filter material until the performance target is again met, where the incremental increase increments are smaller than the incremental decrease increments.

[0057] The determination is started from a conservative estimate of the total amount of filter material. All candidate compositions tested in one step have the same total amount. If one of the candidate compositions is found to meet the sunscreen performance target, the corresponding step is terminated, and the next step is continued with a reduced total amount. This is repeated until no candidate composition is found in the step that meets the sunscreen performance target. The increments for the stepwise reduction and stepwise increase can be fixed or variable, for example, depending on the underperformance or overperformance of the tested composition.

[0058] Preferably, the candidate compositions to be tested are sorted according to the efficiency of the filter material they contain, such that candidate compositions with high expected efficiencies are tested first, and the testing sequence is stopped as soon as the sunscreen performance target is met by one of the candidate compositions, thereby significantly reducing the number of candidate compositions to be tested in stages of gradually reducing the total amount of filter material.

[0059] Many objectives, such as cost, excess solvent, and minimum oil load, typically decrease (become better) with larger total amounts of filters until an optimum is reached. If the total amount of filters is further increased (by imposing the constraint that the total amount of filters used equal a predetermined value), some objectives, such as cost, excess solvent required, or minimum oil load, will again increase (become worse). Thus, a stopping criterion can be the increase (worsening) of one or more objectives when the total amount of filter material is increased.

[0060] Linear objective functions (e.g., the most efficient or cost-effective combination of UV filters) can be evaluated by a simple dot product and thus can be calculated very efficiently using matrix operations on a standard computer. In these cases, all combinations can be efficiently evaluated at once, followed by successive block-by-block performance predictions. Using the current optimal value (e.g., lowest cost) in the previous block, performance predictions from the remaining combinations with worse values ​​compared to this optimal value can be skipped. Typically, this allows the optimal value to be found for a given total amount of filters with fewer than 3,000 calculations.

[0061] Preferably, the method includes providing a list of the best candidate compositions with the highest value for the optimization objective, for example, a list of 10 or 20 compositions.Using a combinatorial approach, this list can be obtained without excessive trial and error.Based on the list, the user can select a preferred composition, especially a composition that is not only close to the top value for the optimization objective, but also has some other one or more properties that make it better than the top composition on the list.Advantageously, the list includes not only the values ​​for the composition and the objective, but also other relevant properties that characterize the composition.

[0062] Compared to numerical optimization approaches, in the context of combinatorial methods, filters must be preselected, as the maximum number of filters would typically be 6-7 on a typical computer. Similarly, the minimum increment would be approximately 0.5% by weight. Nevertheless, the algorithm is slower (or requires significantly more computing power) than numerical optimization approaches.

[0063] Often, optimizing one particular characteristic (e.g., efficiency, etc.) forces another characteristic (e.g., cost, etc.) into an unacceptable range. Therefore, preferably, the method of the present invention includes optimizing a trade-off within a plurality of optimization objectives, providing an acceptable range of values ​​for each of the objectives, providing a relative importance factor between the objectives, and minimizing one of the values ​​using the relative importance factor in a linear constraint for the minimization.

[0064] The tolerance range may be provided by the user or automatically. In particular, this range may be based on the results of a previous optimization step (for a single objective). The relative importance factors may be set to equal values, e.g., 1, for all paired objectives, or may be provided by the user or automatically.

[0065] In particular, the trade-off optimization involves the following substeps: a) using the above method to determine the best possible values ​​of properties for two or more optimization objectives of interest; b) from the results obtained, obtain the maximum and minimum values ​​for each of these characteristics or use the information obtained in the previous optimization to select the individual boundaries for the characteristics related to the optimization objective; c) selecting the relative importance between characteristics as a factor F (for equally important characteristics F=1); and d) minimizing one of the properties using a relative importance factor F in the linear constraint or constraints; may include:

[0066] Instead of a compromise optimization, inequality constraints can be used in the optimization to prevent optimizing one particular characteristic from forcing another characteristic into an unacceptable range, although this does not allow for precise setting of the relative importance of the characteristics.

[0067] Preferably, the method includes automatically determining an optimal solvent composition for the determined sunscreen composition. The choice of solvent is highly relevant to various aspects of the final formulation, including its cost and skin feel. See, for example, B. Herzog, J. Giesinger, M. Schnyder: "Solubility of UV Absorbers for Sunscreens is Essential for the Creation of Light Feel Formulations", SOFW Journal, 139, 7-2013, 7-14. Therefore, optimizing the solvent composition can improve the properties of the final formulation.

[0068] Advantageously, the optimum solvent composition is determined from the minimization of excess solvent, subject to the constraint that all filter materials of each sunscreen composition are dissolved.

[0069] To determine the optimal solvent composition, a number of solvents (eg, 4-6 substances) are predetermined.

[0070] For minimization, numerical optimization using a suitable algorithm, for example, recursive least squares programming (SLSQP) or the Simplex algorithm, can be utilized.

[0071] Alternatively, a combinatorial approach is used to determine the optimal solvent composition.

[0072] When the actual solvent composition is relevant to the optimization objective of optimizing the sunscreen composition (as is usually the case), the determination of the optimal solvent composition is preferably incorporated into the overall determination of the sunscreen composition, thereby allowing for the identification of an overall optimum, including the selection of filter and solvent composition.

[0073] The computer program product of the present invention contains instructions that, when executed by a computer, cause the computer to perform each of the steps of the method of the present invention described above.

[0074] In the method for producing a sunscreen composition of the present invention, the composition is determined as a composition of UV filter substances according to the method of the present invention described above. Then, the sunscreen composition is obtained by combining the UV filter substances.

[0075] Other advantageous embodiments and feature combinations will become apparent from the following detailed description and the claims as a whole.

[0076] The drawings used to explain the embodiments are as follows: [Brief explanation of the drawings]

[0077] [Figure 1] 1 is a flow chart that schematically illustrates a method for determining a sunscreen composition containing multiple UV filter materials. [Figure 2] 1 is a flow chart that schematically illustrates a combinatorial approach to finding optimized sunscreen compositions. DETAILED DESCRIPTION OF THE INVENTION

[0078] In the figures, identical components are given the same reference numerals.

[0079] 1 is a flow chart illustrating a method for determining a sunscreen composition containing multiple UV filter materials. The method is computer-based and is performed by executing dedicated software running on a local computer (e.g., a personal computer) or on a server connected to a local user terminal. A user interacts with the local computer or terminal in an interactive manner. The software may include several modules running on different processors. In this case, the processors may be co-located or remote from each other. In particular, a local user client (including a web browser or dedicated client application) can interact with the server software, and / or more computationally intensive tasks, such as numerical optimization or sunscreen performance simulation, can be performed by a dedicated processor (e.g., a GPU) or server.

[0080] Essentially, the objective of the described embodiment of the method of the present invention as applied by a user is to find a sunscreen composition, including a solvent composition, that meets the sunscreen performance target and possibly other constraints and is optimal with respect to one or more optimization objectives.

[0081] First, the user selects the actual set of filter substances to be considered (step 10). For this purpose, a list containing a basic set of filter substances is displayed, and the user selects the desired substance. Furthermore, the user has the opportunity to provide minimum and maximum amounts for at least some of the filter substances (step 15). This is not mandatory; the user can leave the corresponding input fields empty, which means that the amount of each filter substance can be as low as 0 (minimum not indicated) or as high as 100% (maximum not indicated), or the maximum amount is automatically obtained from regulatory limits in a user-specified or automatically detected area, for example, via language and / or location settings on the user's computer, or via positioning data, if available.

[0082] Next, the user provides information regarding the required sunscreen performance target (step 20). For this purpose, the user selects the following available properties related to sunscreen performance: - internal or external sun protection factor SPF (e.g. dedicated SPF or protection category); - UVA protection index UVAPF internally or externally (e.g. dedicated IPD or JCIA); -critical wavelength; - the ratio of UVA protection to UVB protection (e.g. based on Boots Star® ratings); - Blue light protection, Select one of the following.

[0083] Additionally, the user provides a target value for the selected property, for example, SPF≧30 inside the body.

[0084] The user then has the opportunity to select further constraints on at least one property of the combination to be determined (step 30). These constraints may be on the following properties: - Value or range for the total amount of filter material; - Values ​​or ranges for the amount of a particular filter substance; - maximum value for the cost of the composition of the filter material; - Values ​​or ranges for the weighting of the composition of the filter material; - values ​​or ranges for the environmental friendliness parameters of the composition; - a value or range for the maximum amount of excess solvent; - Maximum oil load, It can be related to.

[0085] The constraints are displayed on the screen, and the user selects the desired additional constraints. Depending on the constraint, input fields for the set value, minimum value, and / or maximum value are displayed. The user is free to select no additional constraints, one additional constraint, or multiple additional constraints.

[0086] Next, the user selects multiple optimization objectives: - Cost-effectiveness; -weighting; - filtering efficiency; - Environmental considerations; - the amount of excess solvent; -Minimum oil load; - Homogeneous protection; - Sun protection factor and / or UVA protection factor; - Blue light protection; - the similarity of the filter material to the offered composition, Select one or more optimization goals from (step 40).

[0087] The filter weighting is a number that can represent the price of the filter, a score for how environmentally friendly the filter is, a score for how easy the filter is to formulate, a score for the filter's impact on the sensation of sun protection, etc. The weighting of a filter combination is the sum of the weightings of the individual filters multiplied by their corresponding percentages in the combination, and can thus represent, for example, the total cost of the filter combination, the total sensory impact, etc.

[0088] The minimum oil load is the sum of all the filters in the oil phase, plus any excess solvent that may be required to dissolve the solid filters.This is a good measure of how much flexibility the formulator has for a specific filter combination.This is due to the fact that the specific total oil load of the final product should not usually be exceeded.This is because a very high oil load will cause the product to have an undesirably heavy feel.If the minimum oil load due to the filter material is already very high, the flexibility to add additional oil-based materials will be very limited.

[0089] After the objective is selected, it is checked whether the selected objective is compatible with the further constraints previously provided (decision 50). If not, for example if the property to be optimized was subject to constraints, a warning message is displayed. The user then has the opportunity to release the corresponding constraint or select another objective.

[0090] Finally, the user selects the actual set of solvent substances to be considered (step 60). For this purpose, a list containing a basic set of solvent substances is displayed and the user selects the desired substances. Furthermore, the user has the opportunity to provide minimum and maximum amounts for at least some of the solvent substances (step 65). This is not mandatory and the user can leave the corresponding input fields empty, which means that the amount of each solvent substance can be as low as 0 (minimum not indicated) or as high as 100% (maximum not indicated).

[0091] Once the user has provided all the information, an optimal sunscreen composition is automatically determined (step 100) as a composition of materials from the selected actual set of filter materials and an optimized composition of solvents compatible with each composition of filter materials. This composition achieves the performance goals and satisfies all possible additional constraints. This composition is optimized with respect to the selected objectives.

[0092] The output to the user (step 70) includes the filter substances and their corresponding amounts, and the solvents and their corresponding amounts, and further includes numerous properties characterizing the corresponding compositions, including performance achieved, UVA / SPF ratio, total amount of filter substance, efficiency (see below), weighting, minimum oil load, and cost.

[0093] As described in more detail below, optimization involves calculating the sunscreen performance of various candidate compositions using available performance simulation tools.

[0094] Based on the results, the user can decide whether the composition found is the desired one, or whether the input parameters, such as the selection of filter and / or solvent materials, ranges corresponding to their amounts, or further constraints, should be adjusted for the next optimization step (decision 80). Alternatively or additionally, optimization can be performed for another objective or combination of objectives. In this case, the input parameters can be adjusted based on the results of the previous optimization step.

[0095] In a preferred embodiment, the optimal sunscreen composition is automatically determined using numerical optimization of an objective function for a selected optimization objective, the variables of the objective function including the proportion of filter substances in the sunscreen composition to be determined.

[0096] In the described example, the optimization is performed using the "trust-constr" method available in SciPy (release 1.4, December 19, 2019) (see SciPy.org), an open-source library for the Python programming language. This method is based on the EQSQP algorithm (Lalee, Marucha, Jorge Nocedal, and Todd Plantega: "On the implementation of an algorithm for large-scale equality constrained optimization", SIAM Journal on Optimization 8.3: 682-706, 1998).

[0097] The objective function depends on the optimization objective. Some examples are given below:

[0098] One possible objective is to find the most efficient combination of UV filters, that is, the combination that achieves the desired performance goal for the minimum total amount of UV filters. In the following equation, the number of individual filters is given by k, and the concentration of filter i is x i is given as:

[0099] The purpose is to:

number

number

number

[0100] Furthermore, the amount of UV filter must be within the following minimum and maximum limits:

number

[0101] Another object relates to weighting UV filter combinations, where the filter weight is a number that can represent the price of the filter, a score of how environmentally friendly the filter is, a score of how easy it is to compound the filter, a score of the filter's impact on the sensation of sun protection, etc., in the following formula:

number

[0102] The weighting of a filter combination is calculated by multiplying the weighting of each individual filter by the corresponding proportion in this combination (

number

[0103] The purpose is to:

number

[0104] Another objective is the minimum oil load of the composition. This is the sum of the concentrations of all m filters in the oil phase plus any excess solvent or solvent mixture that may be required to completely dissolve all solid filters. This is a good measure of how much flexibility the formulator has with a particular filter combination.

[0105] The purpose is to:

number

number

[0106] (Candidate) Filter Composition

number

number

number

number

number

[0107] Therefore, the resulting vector

number

[0108] Here, the conditions for completely dissolving the solid UV filter are:

number

number

number

[0109] Here, the minimum amount of excess solvent is used.

number

number

[0110] Another objective is the similarity of the composition to a (user-provided) composition, which allows for the tuning of known compositions so that performance goals are achieved and constraints are met.

number

[0111] The optimization provides adjustments that quickly adjust the filter density to meet desired performance goals and constraints while preserving the user's basic ideas as closely as possible.

[0112] Up to this point, we have assumed that the user has selected a single optimization objective. However, optimizing one particular characteristic, e.g., efficiency, can often force another characteristic, e.g., cost, into an unacceptable range. A possible solution to this problem is to set an inequality constraint that corresponds to this characteristic.

[0113] A very versatile method is to search for the optimum compromise between two or more properties by following the procedure described below.

[0114] First, the properties of interest are identified. The sunscreen composition is successively optimized for each of these properties, as described above, resulting in maximum and minimum values ​​for each property of interest.

[0115] Next, the individual property boundaries p min , p max These bounds may be set automatically to the minimum and maximum values ​​obtained from the optimization, or may be set by the user, taking into account the minimum and maximum values.

[0116] In a further step, the relative importance between the characteristics is selected as a factor F. For equally important characteristics, F=1.

[0117] Finally, the property of interest, p i One of them is General:

number

number

number

number

number

number

number

[0118] The usual performance and performance constraints discussed above also apply.

[0119] Instead of numerical optimization, in a further embodiment, the optimal sunscreen composition is determined automatically using a combinatorial approach. Figure 2 is a flow chart that schematically illustrates such a combinatorial method for finding an optimized sunscreen composition (step 100 in the method of Figure 1). In this context, a plurality of candidate compositions is automatically defined, and the sunscreen performance of at least some of the plurality of candidate compositions is determined using a performance simulation tool.

[0120] Such an approach is feasible when various measures are applied to reduce the number of calculations. 1) reducing the number of UV filters by pre-selecting a maximum of 6 filters, and before proceeding, checking whether the number of filters exceeds 6 (decision 101), and if so, the user is asked to reduce the number of selected filters or switch to numerical optimization; 2) selecting an increment for automatic definition of the composition of 0.5 to 1.0 wt.%, preferably 0.5 wt.%; 3) Reducing the search space, 3a) a. Conservatively estimating the total amount of filters required to achieve a performance target, e.g., SPF / 2 wt.% (step 102), and calculating only combinations with the corresponding total amount of filters (using six filters, 0.5 wt.% increments, and a maximum use level of 5 wt.% per filter, limiting the total concentration to 15 wt.% (SPF30 / 2), requires 312,620 calculations for SPF=30); b. Enumerating the corresponding filter combinations (step 103) so that the combinations with the most efficient filters are tested first; c. Next, checking whether the target performance can be met with the current amount of filtering; for this purpose: - determining the performance of the candidate composition using a performance simulation tool (step 104); - checking whether the determined performance meets the performance target (decision 105), and if not, checking whether further candidate compositions remain (decision 106), and if so, checking the next candidate composition (according to the enumeration); d. If the determined performance of one of the candidate compositions meets the target, stop the calculations (this typically occurs within less than 1000 calculations), reduce the amount of filter to a fixed number, for example 1% by weight (step 107), and enumerate and check the resulting candidate compositions as described above (steps 103-106), repeating this (backward search) until the target is no longer achievable, i.e., until none of the compositions meets the target (no further candidate compositions at decision 106); e. Next, increase the total amount of the filter by one increment (e.g., 0.5% by weight) (step 110), list the resulting candidate compositions (step 111), and check whether the target performance can be met with the current total amount of the filter; for this purpose, - determining the performance of the candidate composition using a performance simulation tool (step 112); - checking whether the determined performance meets the performance target (decision 113), and if not, checking whether further candidate compositions remain (decision 114), and if so, checking the next candidate composition (according to the enumeration); f. If the determined performance of one of the candidate compositions meets the target, stop the calculations (this typically happens within less than 1000 calculations); if none of the candidate compositions meets the target, increase the amount of filter again by a fixed number, for example 0.5% by weight (step 110), and enumerate and check the resulting candidate compositions as described above (steps 111-114), repeating this (forward search) until the target is achieved, i.e., at least one composition meets the target (no further candidate compositions at decision 114), thereby obtaining the lowest total amount of filter (most efficient filter combination) that allows achieving the target performance; The minimum total amount of filters that can achieve the target performance is searched for by 3b) Search for an optimal solution by starting with the lowest feasible total filter amount and increasing the total amount by one increment, e.g., 0.5 wt. % for each search, where: - Reducing the search space (step 120), - starting from the determined lowest total filter amount, calculating the relevant properties for the remaining candidate compositions with a fixed total filter amount (step 121); - checking whether the performance with respect to the selected optimization objective has improved compared to the previous run (decision 122), and if so, increasing the total filter volume by the mentioned increment (step 123) and repeating the decision with the increased value (steps 120 to 122); - if there is no further improvement (decision 122), repeat the best composition (or list of best compositions) for the selected optimization objective for further processing. reducing the search space by achieving is.

[0121] Many properties, such as cost, excess solvent, and minimum oil load, typically decrease as the total filter volume increases until an optimum is reached. Thus, in many cases, the forward search can continue as long as the property value decreases, stopping when the total filter volume begins to increase again from the increased total filter volume. Otherwise, the search must continue until the maximum total UV filter volume is reached.

[0122] Linear objective functions (e.g., the most efficient or cost-effective combination of UV filters) can be evaluated by a simple dot product and therefore can be calculated very efficiently using matrix operations on a typical computer. In these cases, all combinations can be efficiently evaluated at once, followed by successive block-by-block performance predictions. The current optimum value (e.g., lowest cost) in the previous block can be used to skip all performance predictions from the remaining combinations that have values ​​worse than this optimum value. Typically, this reduction of the search space (step 120) allows the optimum to be found for a given total amount of filters in fewer than 3000 calculations.

[0123] Using a combinatorial approach, it is possible to display not only a single optimized composition, but also a list of the best candidate compositions, e.g., 10 or 20 compositions, along with their properties. This provides the user with additional useful information. In particular, the user can identify whether the best solutions feature very similar compositions of filter material or whether they involve substantially different compositions. In the latter case, the user can, based on their knowledge, select only the second or third best compositions, or decide to proceed with the next optimization step based on the second or third best compositions, which may provide a better starting point for further improvement.

[0124] The user may be able to choose between numerical optimization and a combinatorial approach at each iteration step. As an example, once a promising composition is found using a combinatorial approach, further improvement can be achieved by applying a numerical optimization step, starting from the identified candidate composition. This is because the numerical approach provides more accurate quantities for the components of the composition than the combinatorial approach, where possible values ​​are discrete. Conversely, in optimization using a combinatorial approach, it is also possible to double-check whether the numerical approach has actually found the true global minimum.

[0125] Switching to a combinatorial approach can also be useful for obtaining a ranked result list.

[0126] In all embodiments of the present invention, preferably the UV filter substance is octyl methoxycinnamate (PARSOL® MCX), isoamyl methoxycinnamate (Neo Heliopan® E 1000), homosalate (3,3,5 trimethylcyclohexyl 2-hydroxybenzoate, PARSOL® HMS), ethylhexyl salicylate (also known as ethylhexyl salicylate, 2 ethylhexyl 2-hydroxybenzoate, PARSOL® EHS), octocrylene (2 ethylhexyl 2-cyano-3,3-diphenylacrylate, PARSOL® 340), polysilicone 15 (PARSOL® SLX), diethylhexyl 2,6-naphthalate (Corapan® TQ), syringylidene malonate, e.g. diethylhexyl Syringe lidene malonate (Oxynex® ST Liquid), benzotriazolyl p-cresol (Tinoguard® TL) and benzophenone-3 and drometrizole trisiloxane, bis-ethylhexyl-oxyphenol methoxyphenyl triazine (PARSOL® SHIELD), butyl methoxydibenzoylmethane (PARSOL® 1789), methylene bis-benzotriazolyl tetramethylbutylphenol (PARSOL® MAX), diethylamino hydroxybenzoyl hexyl benzoate (UVINUL® A PLUS), ethylhexyl triazone (UVINUL® T150), diethylhexyl butamido triazone (Uvasorb® HEB), tris-biphenyl triazone (Uvasorb® A2B), 4-methyl-benzylidene camphor (PARSOL® 5000), and 1,4-Di(benzoxazol-2'-yl)benzene bis-ethylhexyloxyphenol methoxyphenyl triazine, phenylbenzimidazole sulfonic acid (PARSOL® HS) and disodium phenyldibenzimidazole tetrasulfonate (Neoheliopan® AP), fine (preferably coated) titanium dioxide (e.g., PARSOL® TX) and zinc oxide (e.g., PARSOL® ZX).

[0127] Using the methods described above, optimized compositions of UV filter materials were calculated, subject to various constraints and optimization objectives.

[0128] Example 1 In the first example, the objective was to find the most efficient combination of UV filters, i.e., to achieve the desired performance target with the minimum total amount of UV filters.

[0129] The following filter materials are used with the following boundaries and weights: [Table 1] Selected with.

[0130] Performance constraints were SPF ≥ 30 and ratio UVAPF / SPF ≥ 0.33. No property constraints were applied.

[0131] The optimized composition was determined using the optimized combinatorial approach described above (0.5 wt% increments, up to 5.0 wt%, resulting in 1.28 million possible combinations) and numerical optimization. [Table 2] and has the following properties: [Table 3] It has.

[0132] Example 2 In the second example, the objective was to find the most weighted efficient combination of UV filters.

[0133] The filter materials, their boundaries and weightings, and constraints were the same as in Example 1.

[0134] The results of the two methods are [Table 4] and has the following properties: [Table 5] It has.

[0135] Example 3 In a third example, the objective was to find the UV filter combination with the lowest oil load.

[0136] The filter materials, their boundaries and weightings were the same as in Examples 1 and 2. The performance constraints were SPF ≥ 30 and the ratio UVAPF / SPF ≥ 0.33. The property constraints were a maximum weighting of 6.5 and a maximum total amount of filter of 17%.

[0137] The results of the two methods are [Table 6] and has the following properties: [Table 7] It has.

[0138] Example 4 In the fourth example, the objective was to find the most weighted efficient combination of UV filters.

[0139] In contrast to Examples 1 to 3, instead of six filter materials, the following boundaries and weightings were used: [Table 8] Ten filter materials with the following properties were selected.

[0140] Performance constraints were SPF ≥ 50 and ratio UVAPF / SPF ≥ 0.33. No property constraints were applied.

[0141] Optimization using a combinatorial approach was infeasible: even a 0.5% weight increment would have required checking 98 billion possible combinations, which is infeasible on a conventional computer.

[0142] The results of the numerical optimization are [Table 9] and has the following properties: [Table 10] It has.

[0143] Example 5 In the fifth example, the objective was to find the most efficient combination of UV filters.

[0144] The filter materials, their boundaries and weightings, and constraints were the same as in Example 4.

[0145] Again, a combinatorial approach was not feasible for the same reasons. Numerical optimization results showed that [Table 11] and has the following properties: [Table 12] It has.

[0146] Example 6 In the sixth example, the objective was to find a compromise between weighting and efficiency (by weighting the two objectives equally).

[0147] The filter materials, their boundaries and weightings, and constraints were the same as in Examples 4 and 5.

[0148] Again, a combinatorial approach was not feasible for the reasons mentioned. Numerical optimization results showed that [Table 13] and has the following properties: [Table 14] It has.

[0149] Comparison of the properties with those of Examples 4 and 5, in which the composition is optimized for one purpose, reveals that: [Table 15] is derived.

[0150] As can be seen from the results, the compromise is

number

[0151] Example 7 In the seventh example, the objective was to find a compromise between weighting and efficiency, similar to the sixth example, but aiming for a 100 / 50 weighting / efficiency compromise (F=2).

[0152] The filter material and its boundaries were the same as in Examples 4 to 6. In addition to the constant constraints on SPF and UVAPF / SPF ratio, the following property constraints were applied: Maximum weighting w max ≦18.0 Maximum total amount of filters a max ≦25% (i.e., E min ≧2.0) was imposed.

[0153] Again, a combinatorial approach was not feasible for the reasons mentioned. Numerical optimization results showed that [Table 16] and has the following properties: [Table 17] It has.

[0154] Again, checking the results, the compromise is the desired relative weighting of the two objectives (F=2), i.e.

number

[0155] The present invention is not limited to the above-described embodiments. In particular, the number and sequence of method steps may vary, and the user may be offered more or less options for interacting with the process or inputting information. Part of the required information may be provided automatically, such as by retrieving it from a database. In contrast to the described embodiments, even the iterative process may be computer-assisted or computer-guided, i.e., successive optimizations with different input parameters and / or objectives may be proposed to the user or performed automatically.

[0156] Other algorithms are available for numerical optimization. One readily available method includes the recursive least squares programming (SLSQP) algorithm, available in the SciPy library mentioned above. Additional methods such as line search or penalty / augmented Lagrangian algorithms may also be optimizable.

[0157] In summary, it should be noted that the present invention creates a method that allows for the efficient determination of optimal sunscreen compositions.

Claims

1. 1. A computer-based method for determining a sunscreen composition comprising a plurality of UV filter materials, the method comprising: a) selecting at least one constraint for at least one property, including a sunscreen performance target, of the composition to be determined; b) selecting an optimization objective from a plurality of optimization objectives; c) automatically determining, from a set of filter materials, a sunscreen composition that satisfies said at least one constraint and is optimized with respect to the selected optimization objective, as a composition of filter materials; The automatically determining step includes: - generating a plurality of candidate compositions; - determining the sunscreen performance of said candidate composition using a performance simulation tool; - comparing the determined sunscreen performance of said candidate composition with said sunscreen performance target; Including, the step of automatically determining the sunscreen composition comprises a numerical optimization of an objective function related to the selected optimization objective, a variable of the objective function comprising a proportion of a filter substance in the sunscreen composition to be determined; A numerical optimization algorithm is used that evaluates the first and second derivatives of the objective function for selecting the search direction and step size to be adopted at each iteration; method.

2. the user is asked to select said at least one constraint; The method of claim 1.

3. The user is asked to select the optimization objective; 3. The method according to claim 1 or 2.

4. the sequence comprising steps a) to c) is repeated and a user iteratively adjusts the at least one constraint and / or the optimization objective; 4. The method according to claim 2 or 3.

5. The user selects the actual set of filter materials to be considered from the basic set of filter materials; 5. The method according to any one of claims 1 to 4.

6. The user provides a maximum amount of filter material for at least some of the selected filter materials; The method of claim 5.

7. The user provides a minimum amount of filter material for at least some of the selected filter materials; 7. The method according to claim 5 or 6.

8. The method includes selecting at least one further constraint on at least one property of the determined composition.

8. The method according to any one of claims 1 to 7.

9. The at least one further constraint may be the following property: a) total amount of filter material; b) a quantity of one or more separate filter materials; c) the cost of the composition of the filter material; d) weighting the composition of said filter material; e) Environmental considerations; f) the amount of excess solvent; g) oil load; is a range or boundary value relating to one of 9. The method of claim 8.

10. The sunscreen performance target is: a) internal or external sun protection factor SPF, b) UVA protection factor UVAPF, either internal or external; c) critical wavelength; d) the ratio of UVA protection to UVB protection, and e) Blue Light Protection, selected from one of 10. The method according to any one of claims 1 to 9.

11. The plurality of optimization objectives are: a) cost-effectiveness, b) weighting, c) filtering efficiency; d) Environmental considerations; e) the amount of excess solvent; f) minimum oil load; g) the most uniform protection; h) highest sun protection factor and / or UVA protection factor; i) Best blue light protection, and j) similarity of the filter material to the provided composition; at least two of 11. The method according to any one of claims 1 to 10.

12. The numerical optimization involves the application of sequential quadratic programming, in particular the application of interior point methods.

12. The method according to any one of claims 1 to 11.

13. For the step of automatically determining a sunscreen composition, a plurality of candidate compositions are automatically defined, and the sunscreen performance of at least some of the plurality of candidate compositions is determined using the performance simulation tool.

12. The method according to any one of claims 1 to 11.

14. In a first substep, a minimum total amount of filter material is determined for a composition that achieves said sunscreen performance goal; In a subsequent second substep, the constraint on the value of the total amount of filter material of said candidate composition is gradually increased, starting from the determined minimum total amount, until a stopping criterion is met.

14. The method of claim 13.

15. the minimum total amount of filter material is determined by gradually decreasing the total amount of filter material in a candidate composition tested for sunscreen performance until the sunscreen performance target is no longer attainable, and then gradually increasing the total amount of filter material until the performance target is again met, wherein the incremental increase increments are smaller than the incremental decrease increments; 15. The method of claim 14.

16. the candidate compositions to be tested are sorted according to the efficiency of the filter substances they contain, such that candidate compositions with high expected efficiencies are tested first, and the testing sequence is stopped as soon as the sunscreen performance goal is met by one of the candidate compositions; 16. The method according to claim 14 or 15.

17. providing a list of best candidate compositions; 17. The method according to any one of claims 13 to 16.

18. optimizing a trade-off among a plurality of optimization objectives, said optimizing comprising: providing a tolerance range for values ​​associated with each of said plurality of optimization objectives; providing a relative importance factor between said plurality of optimization objectives; and minimizing one of said values ​​using said relative importance factor in a linear constraint for minimization; 18. The method according to any one of claims 1 to 17.

19. a step of automatically determining an optimal solvent composition for the determined sunscreen composition is performed; 19. The method according to any one of claims 1 to 18.

20. The optimum solvent composition is determined by minimizing excess solvent while ensuring that all filter materials of each sunscreen composition are dissolved.

20. The method of claim 19.

21. The user is provided with an option to participate in the interactive work process or input information in at least one step; 21. The method according to any one of claims 1 to 20.

22. 1) Reducing the number of UV filters by user pre-selection of up to six filters; 2) selecting filter amount increments between 0.5 and 1.0 wt.%; 3) Reducing the search space, 3a) searching for the lowest total amount of filters that achieves the performance goal, a. conservatively estimating the total amount of filters required to achieve the performance goal and calculating only combinations that include the corresponding total amount of filters; b. Enumerating the corresponding filter combinations such that the combinations containing the most efficient filters are tested first; c) stopping the calculation as soon as the performance goal is met; d. Re-estimating the total amount of the filter according to the over-performance or under-performance of the first estimation, and re-calculating only the combinations including the corresponding total amount of the filter; e) Iteratively reducing the amount of said filters by a fixed number and recalculating all these said combinations until the goal is no longer achievable; f) increasing the total amount of filtering by said increment until said performance target is again achieved, and stopping the search as soon as the target is achieved; searching for the minimum total amount of filters that can achieve the performance goal by 3b) searching for an optimal solution, starting with the lowest feasible filter total and increasing said total by one increment in each search; reducing the search space by further comprising:

22. The method of any one of claims 1 to 21.

23. comprising instructions which, when executed by a computer, cause the computer to carry out the method of any one of claims 1 to 22, Computer program.

24. 1. A method for making a sunscreen composition, the method comprising: determining the composition of the UV filter substances according to a method according to any one of claims 1 to 22; combining said UV filter substances; A method comprising:

Images