COMPUTER AIDED FORMULATION DESIGN EXPERT SOFTWARE CASE STUDY

COMPUTER AIDED FORMULATION DESIGN EXPERT SOFTWARE

1. Computer aided formulation development

Roshan Bodhe PH-02 Page 1

INTRODUCTION TO COMPUTER AIDED FORMULATION AND DEVOLOPMENT:
Formulation and development is a process of selection of component and processing.Now days
computer tools used in the formulation and development of pharmaceutical product.Various
technique such as design of experiment are implemented for optimization of formulation and
processing parameter.Traditionally optimization in pharmaceuticals refer changing one variable
at a one time, so to obtain solution of a problematic formulation.Many times finding the correct
answer is not simple and straight forward in such cases use of computer tools (optimization
procedure) for best compromise is the smarter way to solve problem.
OPTIMIZATION TERM: -
 Optimization means:-
Optimization means choosing the best element from some set of available alternatives.
To optimize means, to make as perfect as possible, to make as functional as possible, to
make as effective as possible, so as possible is important then optimization means what
1. Perfect by whose method?
2. For what characteristics?
3. For what condition?

 Factor: -It is an assigned & independent variables, Such as temperature and
concentration. Those are quantitative & qualitative.
 Level:-Those are designation assigned to the factor. Level indicated as high, low,
medium,Eg:-polymer ratios.
 Response:-It is an outcome of an experiment. Eg:- Disintegration time
 Response surface:-Response surface representing a relationship between the
independent variables X1&X2 and dependent variables Y.

OPTIMIZATION IS OPTIONAL THING OR NOT:-
Before you submit the application for the approval they make inspection of your provision and
facilities known as pre approval provision.Sodue to this approach optimization is not a optional
thing. In earlier days approvalwas given first then they conduct the inspection.Now the things are
change. There is provision known as pre- approval inspection,before they grant the approval they
will inspect the
1)Inspect the facilities.
2)Inspect and check the formulation.
 Why this additives is there?
 Why this excipient added in to this concentration?
 Why the weight of tablet?

So, due to this each and every thing you will justify and defend masterformula record and batch
manufacturing records.so you have justify your formulation justify your formulation.

2. Computer aided formulation development

Roshan Bodhe PH-02 Page 2

For Example: - Why you are heating at 60 degree not on 90 degree.Why you are adding this
preservative or antimicrobial agent to this concentration.So it is notoptional thing, we cannot
escape this, because optimization is mandatory and regulatory need of formulation and
development of pharmaceutical product.
REGULATORY NEED OF OPTIMIZATION TECHNIQUE :
Optimization is a regulatory or mandatory thing required during the formulation and
development because of following reason.Provide in depth understanding to explore and further
defend in the formulation development.
So it is also benefit of manufacturing of formulation. Because you know each and every detail
and probably it’s his expansion and review and all things will be in his hand. So, optimization is
not screening technique,but it is regulatory thing.
OPTIMIZATION IS A TRIAL AND METHOD :-
Early optimization is done on the basis of trial and error method.
What is mean by trial and error?
A series oflogical steps carefully controlling the variables and changing one variable at a time till
satisfactory result produced. So gross variables carefully understood by changing one variable at
atime and making a small change in to it then series ofthis experiment has been done.
If the experiment is done with sufficient length of time, at adequate resources, sufficient time and
help, then time will come and he will come with perfect product.
Golden principles required during the formulation and development because in pharmaceutical
field,those principles decide the things regarding for formulation and development. Because in
pharmaceutical industry you may have thousands of chemical and excipient available,then you
should put more and more quantity.
1) Minimum and absolute quantity of ingredient in formulation:-
Those related to quantity and number of excipient in the formulation. Because we added more
and more ingredient,then those putting burden on labor, and process of manufacturing.
2) Minimum number of steps:-
More will be the unit process then more will be the energy is consumed, more will be the
parameter required at each step.So minimum and absolutely small no. of unit operation required
during formulation and development.
OPTIMIZATION PARAMETER :-
A) Problem types:-
There are two types of problem which are usually addressed in the optimization technique.
a) Unconstrain problem:-
In unconstrain there is no restriction place on the system, then that called as unconstrain
problem.

3. Computer aided formulation development

Roshan Bodhe PH-02 Page 3

In pharmaceutics unconstrain problem are not exist.
Example:- Make the hardest tablet as possible.

2)Constrain problem:-In that some restriction place on the system,
Those problem are important in the pharmaceutical so we think about the constrain
problem.
Example: -Make the hardest tablet as possible, but it disintegrate within 15 min.
Variables types:-
There are certain variables in optimization technique regarding the pharmaceutical
formulation.
These are the measurement values, whichare characteristics of data that called veriables.
There are two types of variables, are as fallow.

a) Dependent variables
b) Independent variables

1) Independent variables in optimization technique:-
That means things you can operate on your own choice,called as independent
variables
These variables under the control of formulator.E.g Mixing time,Concentration of
binder,Drug to polymer ratio,Compression force. These does not depend upon any
other values.

2) Dependent variables:-
These are variables which are not directly under the control of formulator. These
variables are response or chartecteristics of independent variables
VARIOUS EXPERIMENTAL DESIGN
1. Completely randomized designs
2. Randomized block designs
3. Factorial designs
4. Full Fractional Response surface designs
5. Central composite designs
6. Box-Behnken designs
7. Adding center points
8. Three level full factorial designs

1. Completely randomized Designs:-
Theseexperiment compares the values of are response variable based anddifferent
levels of that primary factor.
For example:- if there are 3levels of the primary factor with each level to be run 2
times then there are 6 factorial possible run sequences.
2. Randomized block designs

4. Computer aided formulation development

Roshan Bodhe PH-02 Page 4

For this there is one factor or variable that is of primary interest. To control non-
significant factors, an important technique called blocking can be used to reduce or
eliminate the contribution of these factors to experimental error.
3. Factorial design:-Full Used for small set of factors.

4. Fractional:-It is used to examine multiple factors efficiently with fewer runs
than corresponding full factorial design.

TYPES OF FRACTIONAL FACTORIAL DESIGNS :
1. Homogenous fractional
2. Mixed level fractional
3. Box-Hunter
4. Plackett-Burman
5. Taguchi Latin square
Homogenous fractional:-
Useful when large number of factors must be screened.



Design of Experiment For Formulation And Development:
All pharmaceutical products are formulated to specific dosage form drugs to be effectively
delivered to patient typical pharmaceutical dosage form include tablets, capsules, solution
suspension, etc
Different dosage form required different technology usually present different technological
challenge for formulation & development .Due to complex challenges, formulations scientist
used effective methodology like as design of experiment and statistical analysis for formulation
and development .
Formulation scientist used this method for process optimization and process validation .
EXPERIMENTAL DESIGN :-
The designs used for simultaneous methods are frequently referred to as response surface
designs. Various experimental designs frequently involved in the execution of RSM can broadly
classified as:
A. Factorial design and modifications
B .Central Composite design and modifications
C .Mixture designs
D .D-optimal designs
A. Factorial design and modifications :-
Factorial designs (FDs; full or fractional) are the most frequently used response surface designs.
These are generally based upon first-degree mathematical models. Full FDs involve studying the
effect of all the factors (n) at various levels (x), including the interactions amongst them, with the
total number of experiments as x
n
. The simplest FD involves study of two factors at two levels,
with each level coded suitably.
FDs are said to be symmetric, if each factor has same number of levels, and asymmetric, if the
number of levels differs for each factor.

5. Computer aided formulation development

Roshan Bodhe PH-02 Page 5

Besides RSM, the design is also used for screening of influential variables and factor influence
studies. Representsa 22 and 23 FD pictorially, where each point represents and individual
experiment.The mathematical model associated with the design consists of the main effects of
each variable plus all the possible interaction effects, i.e., interactions between the two variables,
and in fact, between as many factors as are there in the model.
The mathematical model generally postulated for FDs is given as
Y = b0 + b1X1... + b12X1X2... + b123X1X2X3... + e …
where,
bi, bij and e represent the coefficients of the variables and the interaction terms, and the random
experimental error, respectively. The effects (coefficients) in the model are estimated usually by
multiple linear regression analysis selection'. Their statistical significance is determined and then
a simplified model can be written
B Central composite design and its modifications:-
Also known as Box-Wilson design, it is the most often used design for second-order models ,
Central composite design (CCD) is comprised of the combination of a two-level factorial points
(2n),axial or star points (2n) and a central point.
Thus the total number of factor combinations in a CCD is given by 2n +2n + 1. The axial points
for a two-factor problem include, (± a, 0) and (0, ± a), where a is the distance of the axial points
from the center. A two factor CCD is identical to a 32 FD with square experimental domain .
a) face centered cube design (FCCD):-
Results when the same positive and negative distance is taken from the center in a CCD .
A rotatable is identical to FCCD except that the points defined for the star design are changed to
[± (2n )1/4,… 0] and those generated by the FD remain unchanged. In this way, the design
generates information equally well in all the directions.
e.g. the variance of the estimated response is same at all the points on a sphere centered at the
origin. The second-order polynomial for two factors, generally used for the composite designs.

b)Box-Behnken Design:-
Is a specially made design that requires only 3 levels (-1, 0, 1). It overcomes the inherent pitfalls
of CCD, where each factor has to be studied at 5 levels (except for 2 factors with a = ± 1, where
the number of levels per factor is 3), thus the number of experiments increases with rise in the
number of factors.
A BBD is an economical alternative to CCD. Also called as orthogonal balanced incomplete
block design, these are available for 3 to 10 factors. Because the design involves study at three
levels, the quadratic model is considered to be most appropriate.
C ) Mixture designs:-
In FDs and the CCDs, all the factors under consideration can simultaneously be varied and
evaluated at all the levels. This may not be possible under many situations. Particularly, in
pharmaceutical formulations with multiple excipients, the characteristics of the finished product
usually depend not so much on the quantity of each substance present but on their proportions.
Here, the sum total of the proportions of all excipients is unity and none of the fractions can be
negative.
Therefore, the levels of the various components can be varied with the restriction that the sum
total should not exceed one. Mixture designs are highly recommended in such cases. In a two-
component mixture, only one factor level can be independently varied, while in a three-
component mixture only two factor levels, and so on. The remaining factor level is chosen to

6. Computer aided formulation development

Roshan Bodhe PH-02 Page 6

complete the sum to one. Hence, they have often been described as experimental designs for the
formulation optimization. For process optimization, however, the designs like FDs and CCDs
are preferred.There are several types of mixture designs, the most popular being the simplex
design.
Scheffé s polynomial equation:-
components are given as under:
Linear : Y = b1X1 + b2X2 + b3X3 … (
Quadratic : Y = b1X1 + b2X2 + b3X3 + b12X1X2 + b13X1X3 + b23X2X3 ...
Special cubic model: Y = b1X1 + b2X2 + b3X3 + b12X1X2 + b13X1X3 + b23X2X3 +
b123X1X2X3 …
The mathematical model of mixture designs does not have the intercept in its equations. As a
consequence,these Scheffé models are not calculated by linear regression.
D) D-optimal designs:
If the experimental domain is of a definite shape, e.g., cubic or spherical, the standard
experimental designs are normally used. However, in case the domain is irregular in shape, D-
optimal designs can be used.
These are non-classical experimental designs based on the D-optimum criterion, and on the
principle of minimization of variance and covariance of parameters. The optimal design method
requires that a correct model is postulated, the variable space defined and the number of design
points fixed in such a way that will determine the model coefficients with maximum possible
efficiency.
One of the ways of obtaining such a design is by the use of exchange algorithms using
computers. These designs can be continuous, i.e., more design points can be added to it
subsequently, and the experimentation can be carried out in stages. D-optimal designs are also
used for screening of factors
Depending upon the problem, these designs can also be used along with factorial, central
composite and mixture designs.

COMPUTER SOFTWARE USED IN THE FORMULATION AND
DEVELOPMENT:-
Choice of Computer Software Package:-
Many commercial software packages are also available, which are either dedicated to a set of
experimental designs or are of a more general statistical nature with modules for select
experimental design(s). The dedicated computer software is frequently better as the user pays
only for the DoE capabilities. In contrast, the more powerful, comprehensive and expensive
statistical packages like SPSS, SAS, BBN, BMDP, MINITAB, etc. are geared up for larger
enterprises offering diverse facilities for statistical computing, support for networking and client-
server communication, and portability with a variety of computer hardware. When selecting a
DoE software, it is important to look for not only a statistical engine that is fast and accurate but
also the following:
• A simple graphic user interface (GUI) that's intuitive and easy-to-use.
• A well-written manual with tutorials to get you off to a quick start.
• A wide selection of designs for screening and optimizing processes or product formulations.
• A spreadsheet flexible enough for data entry as well as dealing with missing data and changed
factor levels.

7. Computer aided formulation development

Roshan Bodhe PH-02 Page 7

• Graphic tools displaying the rotatable 3-D response surfaces, 2-D contour plots, interaction
plots and the plots revealing model diagnostics
• Software that randomizes the order of experimental runs. Randomization is crucial because it
ensures that "noisy" factors will spread randomly across all control factors.
• Design evaluation tools that will reveal aliases and other potential pitfalls.
• After-sales technical support, online help and training offered by manufacturing vendors


Software Salient feature Source
Design Expert Powerful, comprehensive and
popular package used for
optimizing pharmaceutical
formulations and processes;
allows screening and study of
influential variables for FD,
FFD, BBD,CCD, PBD and
mixture designs; provides 3D
plots that can be rotated to
visualize the response surfaces
and 2D contour maps;
numerical and graphical
optimization
www.statease.com
MINITAB Powerful DoE software for
automated data analysis,
graphic and help features, MS-
Excel compatibility, includes
almost all designs of RSM
www.minitab.com
JMP DoE software for automated
data analysis of various
designs of RSM, graphic and
help features
www.jmp.com
CARD Powerful DoE software for
automated data analysis,
includes graphic and help
features
www.s-matrix.com
DoE PRO XL &
DoE KISS
MS-Excel compatible DoE
software for automated data
analysis using Taguchi, FD,
FFD and PBD. The relatively
inexpensive software, DoE
KISS is, however, applicable
only to single response
variable.
www.sigmazone.com
MATREX Excel compatible optimization
software with facilities for
various experimental designs
http://www.rsd-
associates.com/
matrex.htm

8. Computer aided formulation development

Roshan Bodhe PH-02 Page 8



HOW TO USED DESIGN EXPERT SOFTWARE VERSION 7.0
How to install design expert software
1) Install the exe file
2) Do not run design expert now
3) Copy the crack from crack folder to C/ program file / DX 7 trail run
4) If you want help then press F1for help


SOP OF DESIGN EXPERT SOFTWARE : -
1) File new design
2) Click response surface
3) Add numeric factor
4) write name of independent variables
5) Add low limit and high limit
6) Continue
7) Add response
8) Run will be generated

APPLYING DESING OF EXPERIMENT:-
DOE is an important tool for formulation scientist,Because those gives intelligent and important
decision at every stage of formulation and development. For example, In the formulation of
tablet consist different excipient ,And those affected on final formulation.
Factor Low level (mg) High level (mg) Effect on
Diluent ratio 10 15 Disintegration time
Disintegration level 5 10 Dissolution time
Binder level 4 6 Friability
Lubricant level 6 8 Weight uniformity

Major Technical Challenges in Tablet Formulation Development & Role of DOE:-

Major challenge Potential process Technologies
Uniformity Fluid Bed Granulation
Compatibility Highly Compressible excipient
Flow ability Free Flowing excipients
Dissolution Tablet Matrix containing polymer


and Taguchi design.

9. Computer aided formulation development

Roshan Bodhe PH-02 Page 9

STEP IN FORMULATION AND DEVELOPMENT
1) Excipient compability:-
i. First step in formulation and development is campactability study those select the
excipient.
ii. Those are physically and chemically compatible with the API
iii. It should be biodegradable and compatible
iv. By applying doe we can understand interaction effect with API over a time.
v. By applying DOE we can understand the interaction effect of excipient with API.

Feasibility Study:
i. Excipients are selected from excipient compability study, and next step is the feasibility
study.
i. Those conducted to determine the manufacturing process that enable the formulation to
achieve TPP
ii. Those evaluated technical challenges associated with the formulation and development
iii. In table no. 1 potential process are given to overcome the challenges in formulation and
development.
iv. As technical challenges overcome the next step is selection of manufacturing process.

Selection Of Manufacturing Process:

1) Formulation preliminary study:-
Those gives idea about selection of final excipient
2) Formulation optimization study:-
 Those define the optimum level of excipient in the each formulation.
 During this many formulation factor and response are evaluated in tablet formulation and
development. table 3
 These problems can be solve by applying doe.
 Those gives idea to understand formulation system and optimize the formulation by
choosing best combination of excipient to achieve the TPP

Initial Formulation System for Tablet Formulation
FACTOR API LEVEL & EXCIPIENT CHOICE
API 5-10%
DILUENT MICROCRYSTALINE CELLULOSE
STARCH, LACTOSE
DISINTEGRANT SODIUM STARCH GLYCOLATE
LUBRICANT MAGNECIUM STEARATE

Design and Conduct a Formulation Optimization Study

10. Computer aided formulation development

Roshan Bodhe PH-02 Page 10

FACTOR EXCIPIENT LOW STRENGTH
( % )
HIGHSTRENGTH
(%)
API - 5 10
Diluent MCC 59 30
Disintegrant Crospovidone 5 5
lubricant Magnesium
stearate
1 1

VARIOUS VARIABLES :-

VARIABLES NO. OF LEVELS
API% 2
DILUENT 3
DISINTEGRANT 2
LUBRICANT 2



CASE STUDY

Experimentaldesign manufacturing of Atenolol tablet:-
Atenolol, heavy magnesium carbonate, maize starch and sodium lauryl sulphate , were mixed in
Rapid Mixer Granulator and pneumatics, for 10 min
s
at different impeller and chopper speed.
Purified water (50%) was heat and add gelatin in heated water withconstant stirring until
dissolve. 3.57 % w/w maize starch paste (granulating agent) was prepared with boiling purified
water. Add the gelatin solution to starch paste and mixed properly. Add the granulatingagent to
the material over a period of 2 min at different impeller and chopper speed followed by kneading
for about 2 min to get a good granular mass. Wet granular mass was dried in fluidized bed dryer ,
at an internal temperature of 60 ± 5°C, outlet temperature 40±5°C till a loss on drying of 1.5.3.3
% was achieved on IR moisture balance in auto mode at 105°C.
Dried granules were sifted through 18 mesh on vibratory sifter and mill the retentions of granules
through 1 mm screen of multimill , with knives forward direction at slow speed. The dried
granules were blended with 1% magnesium stearate in Octagonal.
In the presented study, granulation process wasoptimized by taking three different lots, in which
dependent variables were impeller speed, binder addition time, chopper speed, impeller mixing
timeblender for 20 min. Tablets (390mg) were compressed on a 16 station rotary tablet
compression machine using a 9 mm standard flat-face punch.
The prepared tablets were round and flat with an average diameter of 9.0 ± 0.1 mm and a
thickness of 4.75 ± 0.2 mm. and their effect on bulk density, true density, Carr.s index, and
hausner ratio.

11. Computer aided formulation development

Roshan Bodhe PH-02 Page 11


Table 1 Presentation of 3 experiments with variables for granulation process:-

Lot no Impeller
Speed
Binder
addition
time(min)
Impeller
mixing
time(min)
AMP.
Reading
Chopper
Speed
1 Fast 1 12 3.7 Slow
2 Slow 2 8 3.9 Fast
3 Fast 2 6 3.9 Fast

Table3.CompressionprocessvariablesdesignandResponsedataofbatchesintheBoxBehnkend
esign

Batch X1(Ton) X2
(ton)
X3(rpm) Hardness(N) Friability (%) DT(sec)
BB1 -1 -1 0 28.34 1.8 122
BB2 -1 -1 0 57.7 0.3 362
BB3 -1 -1 0 35.39 0.2 128
BB4 -1 -1 0 70.30 0.8 96
BB5 0 0 -1 42.66 0.8 215
BB6 -1 0 -1 59.09 0.23 303



Variables
Level
Low(-1) Medium(0) High (1)
X1
(Precompression
Force)
0 1 2
X2
(Compression Force)
2 4 6
X3
(Compression Speed)
25 30 35

For compression process three levels Factorial Box- Behnken experimental design was used to
evaluateeffect of selected independent variables on the responses, to characterize physical
properties oftablets and to optimize the procedure. This design is suitable for exploration of
quadratic response surfaceand for construction of polynomial models, thus helping to optimize
process by using a small numberof experimental runs. For the three levels three factor Box and
Behnken Experimental design, a total of 15 experimental runs, shown in Table 3, are needed.
The generated models contain Quadratic term explaining the non linear nature of responses This
design also resolves the three factor interaction effect of individual terms and allow a mid level
setting (0), for The design consists of replicated center points and a set of points lying at the mid

12. Computer aided formulation development

Roshan Bodhe PH-02 Page 12

points of each edge of multidimensional cube that defines the region ofinterest .The model is of
the following form:

y = b0 + b1x1 +b2x2 +b3x3 +b4x1x2 +b5x2x3+b6x1x3 +b7x12+b8x22+b9x32 + E

Where;
y is the selected response, b0-b9 are the regression coefficients, X1, X2 and X3 are the
factorsstudied and E is an error term. The Box-Behnken experimental design is an orthogonal
design.
Therefore, the factor levels are evenly spaced and coded for low, medium and high settings; as -
1, 0, +1 Factors studied in the Box and Behnken experimental design where precompression
force (X1), compression force (X2) and compression speed(X3). The factors levels are shown in
Table 3. The selected responses were Hardness (Y1), friability (Y2)and Disintegration time (Y3).
The responses studied and the constraints selected considering AtenololPhysical properties and
regarding U.S.FDA guidelines, presented in Table 4.

Table 4: Responses selected and the constraints used in Box-Behnken design

Code Parameter Constraints
Y1 Hardness 30-40 N
Y2 Friability NMT 0.5 % W/w
Y3 Disintegration time NMT 10 min




EVALUATION OF TABLETS :-
1. Hardness:-
The hardness of the tablets was tested for 10 tablets by pharma hardness teste and average
hardness (N) was being taken and compared with that of standard one.

2. Friability :-
Friability test was performed in accordance with USP , 5 tablets wereselected randomly, their
individual weight was taken and then kept in the friabilator and rotated for 4 minat a speed of 25
rpm the tablets were taken out and any loose dust from them was removed, the weightwas
registered and friability was calculated as a percentage weight loss.

3. Disintegration time :-
The disintegration of the tablets was tested in a disintegration tester , sixtablets were put in to a
basket that was raised and lowered in a beaker containing preheated water at37 °C. The
disintegration test was calculated as the mean value and as the range.

4. In-vitro dissolution studies :-
The release rate of atenolol from tablets (n=3) was determined according to British
Pharmacopoeia (ref)using the Dissolution Testing Apparatus 2 , fitted with paddles. The

13. Computer aided formulation development

Roshan Bodhe PH-02 Page 13

dissolution test was performed using 900 mL of 0.1 N HCl, 37±0.5°C and 50 rpm. A 5 ml
sample waswithdrawn from the dissolution apparatus at predetermine time interval, and the
samples werereplaced with fresh dissolution medium. The samples were filtered through a
0.45µm membrane filter and diluted to a suitable concentration with 0.1 N HCl. Absorbance of
these solution was measured at 275nm using UV/VIS spectrophotometer Cumulative drug
release was calculated usingthe equation generated from Beer Lamberts calibration curve in the
linearity range of µg/mL.

5. Statistical analysis
Statistical analysis of the Box-Behnken design batches was performed by multiple regression
analysis using Microsoft Excel. To evaluate the contribution of each factor with different levels
to theresponse, the two-way analysis of variance (ANOVA) was performed using the DESIGN
EXPERT 7.0.1, demo version software. To graphically demonstrate the influence of each factor
on theresponse, the response surface plots were generated using DESIGN EXPERT 7.0.1
(STAT-EASE) demoversion software.

RESULTS AND DISCUSSION :-
In the present investigation, combinations of three variables were studied using the Box-
Behnkenexperimental design. The mathematical models developed for all the dependent
variables usingstatistical analysis software are shown in equations (1)-(3):

Hardness = 48.43 + 11.92 X1 + 5.60 X2 - 0.37 X3 + 1.38 X1 X2 - 0.58 X1 X3 + 2.19 X2 X3 -
1.41 X12+ 0.91 X22 - 5.10 X32 R2 0.6794 ------ (1)

Friability = 0.90-0.45 X1 - 0.19 X2 + 0.13 X3 + 0.19 X1X2 - 0.43 X1 X3 - 0.17 X2 X3 + 0.016
X12 - 0.16 X22 .
0.21 X32 R2= 0.8202 ------ (2)

Disintegration Time = 262.00 + 62.25 X1 + 53.25 X2 + 85.25 X3 - 68.00 X1 X2 + 28.50 X1X3
+ 168.50X2
X3 - 72.75 X12 - 12.25 X22 + 59.75 X32 R2 = 0.6043 ------- (3)

The hardness of all tablets was found to be below 61 N

Table 5: Box-Behnken Experimental Design


Batch X1(Ton) X2
(ton)
X3(rpm) Hardness(N) Friability (%) DT(sec)
BB1 -1 -1 0 28.35 1.5 122
BB2 -1 -1 0 57.52 0.2 362
BB3 -1 -1 0 35.39 0.5 128
BB4 -1 1 0 70.23 0.61 96
BB5 -1 0 -1 42.26 0.2 215
BB6 - 0 1 59.36 1.8 303

14. Computer aided formulation development

Roshan Bodhe PH-02 Page 14


Variables
Level
Low(-1) Medium(0) High (1)
X1
(Pre-compression
Force)
0 1 2
X2
(Compression Force)
2 4 6


Table 6: Analysis of variance for dependent variables from the Box-Behnken design

Sources SS DF MS F value Probability
Hardness
Regression 1526 9 169.56 1.18 0.0045
Residual 720.21 5 144.14
Total 2246.22 14

Friability
Regression 3.40 9 0.38 2.33 0.0015
Residual 0.74 5 0.15
Total 4.14 14
Disintegration
Regression 2.830 9 3.144 0.85 0.6097
Residual 1.853 5 3.700
Total 4.683 14 4.683


Table 7: Granulation Optimization : Analysis data





Lot No. Initial air
drying
time(min)
Total hot air
drying
time(min)
Inlet temp
o
C Out late temp
o
C
LOD % w/w
1 10 50 50 44 1.50
2 10 60 60 44 1.80
3 10 70 55 38 2.20


Results and Discussion
In The Present investigation granulation and compression process were optimized.

Results and Discussion for Granulation:-
Batch No 1

15. Computer aided formulation development

Roshan Bodhe PH-02 Page 15


Process :-
In granulation process, impeller speed, binder addition time, chopper speed, impeller mixing
timewere studied by taking three different lots for dependent variables i.e. bulk density, true
density,Carr.s index, and hausner ratio (Table 2).
Effect of impeller and chopper speed
Various granulation batches prepared to study the effect of impeller and chopper speed are listed
intable 1
RESULTS AND DISCUSSION FOR COMPRESSION : -
Process:-
Combinations of three independent variables werestudied using t he Box Behnken design.
Themathematical models developed for all the dependent variables using design expert software
areshown in equation (1)-(3)
EFFECT OF PRE- COMPRESSION FORCE : -
Compression force and Compression speed :-
15 batches had been prepared to study the effect of pre compression force is listed in table no 3.
so, wetook three different pre compression force -1 indicates there is no pre compression force,0
indicates pre compression force is 1tonne,+1 indicates pre-compression force is 2 tonnes,after
preparing all the batches results showed that ,after applying pre compression force hardness
becomeshigher than limit in combination with lower compression force and lower compression
force showed cracking tableting defect (BB 4). So, result showed that Precompression force
should not begiven ,when we are applying higher compression force and ,and when Pre -
compression force is notgiven and compression force is also lowered cause lower hardness and
friability problems with highcompression speed.(BB 7) All three independent variables also
affects invitrodissolution studies ,if pre compression force is 2 tonnes and compression force 6
tonnes with lower compression speed 25 rpm ,takes more time to dissolve and vice versa.
Dissolution profiles for two Optimized batches BB 5 and BB 7 were analysed and compared
with innovator product and calculated for similarity factor showed result given in table no. 7


Table No 8
Formulation F2(Similarity Factor) F1 (Difference factor)
BB6 67 03
BB7 29 26


COMPARISION DATA USING DESIGN EXPERT SOFTWERE : -

16. Computer aided formulation development

Roshan Bodhe PH-02 Page 16

17. Computer aided formulation development

Roshan Bodhe PH-02 Page 17








CONCLUSIONS:-
 Design of experiment & statistical analysis have been used in the formulation
development
 Using design of experiment formulation scientist evaluate the all formulation factors in
systematically and timely manner to optimize the formulation and manufacturing
process
 When the pharmaceutical process and product are optimize by systematic approach then
process validation & scale up can be efficient because of the robustness of the
formulation and manufacturing process

18. Computer aided formulation development

Roshan Bodhe PH-02 Page 18


REFERENCES :-
1. Bolton S, Bon C. Pharmaceutical Statistics Practical & Clinical Application, 5TH ED.
New York London ;InformA Healthcare Publishing ; 2010.Pg No 239
2. JAIN N K “Pharmaceutical Product Development” ,CBS Publisher ; New Delhi2010.
Pp 295-340
3. Ganesh. R. Godage Advance Drug Delivery System, Tech Max Publication Pune 2017
Pp 6.1-6.24
4. Kubinyi H. Drug Research: Myths, Hype And Reality. Nature Reviews Drug Discovery.
2003 Aug;2(8):665.
5. Hussain As, Shivanand P, Johnson Rd. Application Of Neural Computing In
Pharmaceutical Product Development: Computer Aided Formulation Design. Drug
Development And Industrial Pharmacy. 1994 Jan 1;20(10):1739-52.