`
`SCIENCE'S TOWER OF BABEL
`
`We are all familar with the Biblical story concerning
`the Tower of Babel and the considerable confusion
`which resulted when the builders' "tongues were
`confounded" (Genesis, 11 : 4- 9) .
`
`On the other hand, we usually think of science as
`being the one area which readily crosses the barriers
`of language and which is fully intelligible in any
`tongue . . Unfortunately, however, this is not always
`the case, and the irony of the matter is that the con(cid:173)
`fusion to which we here refer arises even in the
`individual's native language. This confusion
`is
`primarily due to carelessness on the part of authors.
`Frequently authors are not nearly as precise or con(cid:173)
`sistent as scientists should be in their choice of
`symbols, abbreviations, and terminology.
`
`A writer will refer to the "activity" of a compound
`without making it clear whether he means its
`chemical activity, biological activity, or optical
`activity.
`In other instances D - or L- will be used
`when d- or l- is meant, and vice versa. The ad(cid:173)
`jectives angular, optical, specific, and observed are
`often used either incorrectly or inconsistently in
`discussing the ability of a substance to rotate
`polarized light.
`In other instances, for no good
`reason, different systems of measurement are
`sometimes used in the same article, such as centi(cid:173)
`grade and Fahrenheit temperatures, metric and
`English lengths, or metric and apothecary weights.
`
`A less obvious but more perplexing area of in(cid:173)
`consistency pertains to terminology used to ex(cid:173)
`press spectrometry nomenclature. This situation
`has been due, in part, to evolutionary changes in
`accepted or approved terms and definitions by re(cid:173)
`sponsible nomenclature bodies as well as careless(cid:173)
`ness on the part of some authors. As a guide to
`our authors and readers we are presenting on page
`VIII of this issue, a list of the terms and definitions
`adopted for use in THIS JouRNAL. This information
`is based upon preferred terminology approved by
`the American Society for Testing and Materials,
`"Analytical Chemistry," and "Chemical Abstracts."
`In the interest of accuracy and clarity, we strongly
`encourage regular use of this preferred terminology
`by all of our readers and authors, whenever spectrom(cid:173)
`etry terminology is employed.
`
`JOURNAL OF
`
`Pharmaceutical
`Sciences
`
`February 1965
`volume 54
`
`number 2
`
`EDWARD G. FELDMANN,
`editor
`SAMUEL w. GOLDSTEIN'
`associate editor
`J ACQUELINE BORAKS,
`assistant editor
`S ANDRA SuE SMITH,
`editorial associate,
`R HEA LORIS TALLEY'
`contributing editor
`
`Editorial Advisory Board
`JOSEPH P. BUCKLEY
`JAC~ COOPER
`TROY C. DANIELS
`GEORGE P. HAGER
`GORDON H. SVOBODA
`JOSEPH V. SWINTOSKY
`
`Committee on Publications
`LLOYD M. PARKS, chairman
`WILLIAM S. APPLE
`GROVER C. BOWLES, JR.
`THOMAS J. MACEK
`HUGO H. SCHAEFER
`
`The Journal of Pharmaceutical Sciences is
`published monthly by the American Pharma(cid:173)
`ceutical Association, at 20th & Northampton
`Sts., Easton, Pa. Second-class postage paid
`at Easton, Pa., and at additional mailing office.
`All expressions of opinion and statements of
`supposed fact appearing in articles or editorials
`carried in THIS JoURNAL are published on the
`authority of the writer over whose name they
`appear, and are not to be regarded as neces(cid:173)
`sarily expressing the policies or views of the
`American Pharmaceutical Association.
`Offices-Editorial, Advertising, and Sub(cid:173)
`scription Offices: 2215 Constitution Ave.
`N. W., Washington, D. C. 20037. Publication
`Offices: 20th & Northampton Streets, Easton,
`Pa.
`Annual Subscriptions (effective January 1,
`1965)-United States and foreign: industrial
`a nd government institutions $50, educational
`institutions $30, individuals for personal use
`only $20;
`single copies $3 . Subscription
`rates are subject to change without notice.
`Members of the American Pharmaceutical
`Association may elect to receive the Journal
`of Pharmaceutical Sciences as a part of their
`a nnual $30 A.Ph.A . membership dues.
`Claims-Missing numbers will not be sup(cid:173)
`plied if claims are received more than 60 days
`after the date of the issue, or if loss was due to
`failure to give notice of change of address.
`T he Association cannot accept responsibility
`for foreign delivery when its records indicate
`shipment has been made.
`Change of Address-Members and sub(cid:173)
`scribers should notify at once both the Post
`Office and
`the American Pharmaceutical
`Association, 2215 Constitution Ave. N. W.,
`Washington, D. C. 20037, of any change of
`address.
`
`©Copyright 1965, American Pharmaceutical
`Association, 2215 Constitution Ave. N. W.,
`Washiugton, D. C. 20037.
`
`APOTEX 1032, pg. 1
`
`
`
`Effect of Emulsifier Concentration and Type on the
`Particle Size Distribution of Emulsions
`
`By E. L. ROWE
`
`Two series of mineral oil/water emulsions containing varying amounts of emulsifier,
`either sodium dodecyl sulfate or polysorbate 80, were studied over a 2-year period.
`The Coulter counter was used to determine particle size distributions at various
`time intervals. For each emulsifier, there is a minimum concentration above which
`the emulsions are stable for over 2 years with little change in particle size distribu(cid:173)
`tion.
`Increasing surfactant concentration causes decreasing median particle size
`in both series according to a logarithmic relationship. Some of the theoretical
`aspects of surfactant adsorption at the oil/water interface are discussed.
`
`EMULSIONS AS disperse systems have an ad(cid:173)
`
`vantage over solid-in-liquid dispersions be(cid:173)
`cause the particles are spherical and easier to
`measure. However, the particle size distribution
`of emulsions is changed easily by adjustment of
`the phase volume ratio, method of manufacture,
`temperature, and viscosity (1). Another im(cid:173)
`portant variable is the emulsifier concentration.
`There have been few emulsion studies where the
`emulsifier concentration was varied deliberately
`and the mean particle size measured carefully.
`The earliest work in this area was done by Lange(cid:173)
`vin (2), in 1933, who varied the proportion of
`emulsion ingredients and found microscopically
`that " ... increase in the· proportion of acacia is
`accompanied by a decrease in the size of the oil
`globules .... " The lack of experimental work in
`this area since that time is possibly due to the
`tedium involved in the measurements.
`Recent advances in instrumentation have made
`the quantitative measurement of particle size
`tedious. The
`distributions of emulsions less
`Coulter counter1 is becoming a popular instru(cid:173)
`ment for the measurement of particle size distribu(cid:173)
`In the present study,
`tions of emulsions (3, 4).
`the Coulter counter was used to determine the
`size distributions of two series of emulsions.
`In
`each case, the concentration of emulsifier was
`varied from 0 to 5.0%. The effect of emulsifier
`type and concentration on the distribution and
`the changes in particle size with time were o b(cid:173)
`served over a 2-year period.
`
`EXPERIMENTAL
`Preparation of Emulsions.-Two series of ten oil(cid:173)
`in-water emulsions were made with varying amounts
`of surfactant.
`In both series, 100 ml. of mineral
`oil U.S.P. (viscosity, 40 centistokes) and 300 ml. of
`
`Received July 29, 1964, from the Pharmacy Research
`Unit, The Upjohn Co., Kalamazoo, Mich.
`Accepted for publication October 16, 1964.
`Presented to the Scientific Section, A.PH.A., New York
`City meeting, August 1964.
`The author thanks Drs. E. N. Hiestand, W. I. Higuchi,
`and L. C. Schroeter, for constructive comments and Mr.
`W- Woltersom for experimental assistance.
`~ Coulter Industrial Sales, Elmhurst, Ill.
`
`aqueous surfactant solution were mixed in a Waring
`Blendor for 10 min. Fifty milliliters of each emul(cid:173)
`sion was poured into a 50-ml. graduated cylinder
`for observation of creaming, coalescence, and volume
`ratio of the phases at 25°. The remainder was
`stored in bottles for particle size distribution studies.
`In the series labeled 61A through 61J, sodium
`dodecyl sulfate2 (SDS) was used as the sole emulsifier
`with the concentration varying from 0 to 5.0%.
`(See Table I.) The 65 series was made with poly(cid:173)
`sorbate 80 3 as the sole emulsifier at the same concen(cid:173)
`tration levels as in the SDS series.
`Particle Size Distributions.-The Coulter counter
`(model A) was used with the 100-tt aperture for all
`particle size determinations. The instrument is
`kept in constant calibration by frequent checks with
`well-characterized monodisperse polystyrene latices
`and puffball spores. The manufacturer does not
`recommend measurement of particles with diameter
`less than 1.5% of the aperture diameter (5). How(cid:173)
`ever, in most of the present emulsions, few or no
`particles were observed below 2-tt diameter.
`In the
`few cases where a significant number of the par(cid:173)
`ticles were smaller than 2 tt. the measured contribu(cid:173)
`tion to the total weight distribution was less than
`2%. Also, the frequency distribution curves gen(cid:173)
`erally included at least one point on the small side
`of the modal (peak) diameter.
`In any event, the
`conclusions are not affected by small errors in this
`region. The emulsions were shaken gently so that a
`homogeneous 2-ml. sample could be withdrawn by
`pipet. The 2-ml. sample was diluted to 1 L. with a
`conductive vehicle (0.1% SDS for the 61 series and
`standard conductive vehicle4 for the 65 series) and
`stirred with a magnetic stirrer in a 25° water bath.
`A 15-ml. sample of this dilute emulsion then was
`diluted to 1 L. with the standard conductive vehicle.
`Solvation of the oil upon dilution was considered to
`be negligible. The effect, if any, is probably con(cid:173)
`stant for all the emulsions involved, and the relative
`results would still be meaningful. Cockbain ( 6) has
`in SDS-stabilized paraffin
`reported aggregation
`emulsions where the SDS concentration exceeded
`0.45%. Higuchi (3) found aggregation in 1%
`hexadecane emulsions containing more than 0.1%
`dioctyl sodium sulfosuccinate and that the rate of
`deaggregation was a slow process. However, in the
`
`2 l\1arketed as Duponol C by E. I. du Pont de Nemours
`and Co., Wilmington, Del.
`a Marketed as Tween 80 by Atlas Chemical Industries,
`Wilmington, Del.
`4 Standard vehicle is 0.1% polysorbate 80 and 0.75%
`NaCl in aqueous solution.
`260
`
`APOTEX 1032, pg. 2
`
`
`
`Vol. 54, No. 2, February 1965
`
`261
`
`TABLE I.-COALESCENCE OBSERVED IN Two SERIES OF MINERAL 0IL/W ATER EMULSIONS
`
`Emulsion
`No.
`61A
`61B
`61C
`61D
`61E
`61F
`61G
`61H
`61!
`61}
`65A
`65B
`65C
`65D
`65E
`65F
`65G
`65H
`65I
`65}
`
`Emulsifier
`
`Type
`SDS
`SDS
`SDS
`SDS
`SDS
`SDS
`SDS
`SDS
`SDS
`SDS
`Polysorbate 80
`Polysorbate 80
`Polysorbate 80
`Polysorbate 80
`Polysorbate 80
`Polysorbate 80
`Polysorbate 80
`Polysorbate 80
`Polysorbate 80
`Polysorbate 80
`
`Concn.,%
`0
`0.001
`0.005
`0.01
`0.05
`0.1
`0.5
`1.0
`2.0
`5.0
`0
`0.001
`0.005
`0.01
`0.05
`0.1
`0.5
`1.0
`2.0
`5.0
`
`,----Visible Coalescence of Oil Phase, %-----.
`1 Yr. Old
`2 Yr. Old
`1 Mo. Old
`100
`100
`94
`95
`100
`64
`95
`97
`88
`88
`82
`79
`81
`82
`67
`<1
`<1
`<1
`0
`<1
`<1
`Negligiblea
`N egligiblea
`0
`Negligible
`0
`0
`Negligible
`0
`0
`100
`92
`75
`100
`100
`92
`100
`95
`67
`84
`34
`92
`11
`12
`<3
`7
`7
`<3
`14
`6
`0
`Negligiblea
`3
`0
`Negligible
`0
`<1
`Negligiblea
`Negligible
`0
`
`a One or 2 small drops.
`
`present systems only a negligible amount of aggrega(cid:173)
`tion was noted microscopically in the diluted emul(cid:173)
`sions prior to particle size determination. Both
`weight and number distributions were calculated
`from the raw data.
`
`RESULTS
`
`The emulsions tend to cream rather quickly to the
`theoretical cream phase volume of 34%6 but are
`easily redispersed. Creaming, of course, is not a
`criterion of physical stability in the colloidal sense.
`Growth of the oil particles by coalescence which
`leads to a separated oil layer (commonly called
`breaking) is a true measure of physical stability (7).
`The increase in free coalesced oil which occurs with
`time (even in the most stable emulsions) is expected
`since the emulsified state is thermodynamically un(cid:173)
`stable and, at best, represents a pseudoequilibrium.
`The two series of emulsions were observed over a 2-
`year period for visible coalescence. Table I lists
`three sets of observations made during this time
`interval. There is a minimum surfactant concentra(cid:173)
`tion necessary for physical stability.
`In the SDS
`series, this concentration is near the critical micelle
`concentration ( CMC) of 0.18-0.25% w /v reported
`by Rehfeld (8). The minimum polysorbate 80
`concentration for stability is not so clear-cut.
`In
`this series, the emulsions that are fairly stable at 1
`month have a greater tendency than the SDS sta(cid:173)
`bilized emulsions to coalesce.
`The particle size distributions of the 65 series are
`plotted in Fig. 1 on log probability paper. The 61
`series is similar. While the log normal fit is not
`perfect over-all, the distributions are more closely
`log normal than normal. The number median
`(dn) and mass median diameters (dm) were deter(cid:173)
`mined from the log probability plots and are re(cid:173)
`ported in Table II. Several anomalous results are
`marked with an asterisk. The cause is difficult to
`ascertain since the technique of using the Coulter
`counter in this study was being developed at the
`time of the first run. Data acquired in subsequent
`
`5 The true volume' fraction of oil is 0.25; but in the form
`of close-packed spheres, it would have an apparent volume
`fraction of-1/0.74 X 0.25 = 0.34.
`
`determinations did not contain deviations of this
`magnitude.
`Polydispersity or broadness of the distribution
`can be expressed by the slope of the linear distribu(cid:173)
`tion plot on probability or log probability paper.
`The slope is related to the standard deviation of the
`median particle size. With curves that deviate
`somewhat from linearity, the true slope value is diffi(cid:173)
`cult to determine. Another easy method to ex(cid:173)
`press the polydispersity is the ratio of mass median
`diameter to number median diameter .. A mono(cid:173)
`disperse system has a dm/dn ratio of 1. The broader
`the distribution, the larger the ratio. As shown in
`Table II, polydispersity of the SUS-stabilized emul(cid:173)
`sions decreases with increasing emulsifier, while
`polysorbate 80 emulsions have increasing poly(cid:173)
`dispersity with increasing surfactant concentration.
`This is a result of the manner in which the number
`
`GJ
`;i 99
`
`()
`(f)
`~ 95
`~ 90
`D:l
`;ii 80
`~ 70
`!::. 60
`~50
`::;: 40
`>- 30
`D:l
`~ 20
`w1
`>
`i=
`:5
`:J
`::2:
`:J
`()
`
`0.1 '------'----'---'----'----"'-'-.L...I...J'-----'----'---~
`2
`3 4 5 6 7 8 910
`1
`20 30 40
`DIAMETER, p.
`Fig. 1.-Particle size distributions (log probability
`plot) of the 65 series of emulsions.
`
`APOTEX 1032, pg. 3
`
`
`
`262
`
`Journal of Pharmaceutical Sciences
`
`( (..J'<:OLOLOC\It-
`/:f~c-i~~,...j
`'d
`0
`0~~~~~~
`~'l::l ..ql (!) <:0 <:0 <:0
`
`......
`OOC\IOO':>Ot-
`
`. . . . . .
`01:--(!)....-i..ql,..-i
`OOLO..qiCQC\IC\1
`
`'<:!'
`
`~ l ~ ~..-;~~~
`r ~J:-.:~~..-;~
`
`'l::l
`
`...-i<:OLQ..qi...-i
`C\1....-i....-i....-i....-i
`
`J<:OC\IC\IC\1....-i
`
`'d
`0
`1.0...-iCQO':>...-i
`•
`.
`•
`•
`•
`•
`0 ~
`~'<:'I co(!)(!)(!) t-
`
`median and mass median diameters change. A
`large increase in the number of fine emulsion par(cid:173)
`ticles with a smaller relative change in the rest of the
`distribution will decrease the number median diam(cid:173)
`eter more than the mass median diameter because of
`the small mass contribution. This is apparently the
`case with the polysorbate 80 emulsions and is sub(cid:173)
`stantiated by the increasing turbidity of the aqueous
`phase of the creamed emulsion as surfactant concen(cid:173)
`tration increases.
`In the SDS-stabilized emulsions,
`the aqueous (after creaming) phase becomes clearer
`with increasing surfactant concentration. The num(cid:173)
`ber median diameters of emulsions 61G through 61J
`do not differ significantly, which can be interpreted
`to mean there is little change or at least no increase in
`the small particle end of the distribution with in(cid:173)
`creasing SDS concentration. Decreasing mass me(cid:173)
`dian diameters of the same set is an indication of de(cid:173)
`creasing relative number of large particles.
`
`DISCUSSION
`
`Although there is some indication of growth of the
`particles by coalescence with time (Table II), the
`change is small and possibly within the error of the
`measurements. Therefore, there is no point in
`trying to quantitate the kinetics of growth. Void
`(9) also found the rate of change of interfacial
`area, in 50% Nujol-water emulsions stabilized by
`SDS, to be too slow to afford a criterion for emulsion
`stability. Experimental evidence by Fischer and
`Harkins (10) showed that the specific interfacial area
`of paraffin oil-in-water emulsions decreased by over
`30% in the first 50 hr. and thereafter decreased at a
`much slower almost constant rate. van den Tempel
`( 11) observed a similar effect. His data indicate
`first-order kinetics with respect to the number of oil
`particles.
`The concept of a fast initial coalescence rate
`followed by a slow steady rate seems to be rea(cid:173)
`sonable. The high shearing force applied to the oil
`in the manufacture of these emulsions produces oil
`droplets of varying stability, depending on the
`amount of surfactant available for adsorption. As(cid:173)
`suming that a condensed monolayer of surfactant is
`necessary for stability, the -initial coalescence rate can
`be attributed to instability caused by incomplete oil
`surface coverage. When the oil/water interfacial
`area is reduced by particle growth (coalescence) so
`that the available surfactant can form a complete
`monolayer, a slow rate of coalescence becomes
`dominant.
`In a practical sense, this is an ideal situa(cid:173)
`tion
`if generally applicable. The particle size
`distribution could be observed over a short period of
`time after manufacture.
`In the absence of com(cid:173)
`plicating factors, such as chemical instability, this
`should give enough data to establish the value and
`constancy of the slow coalescence rate, ·allowing a
`fairly accurate prediction of the long range physical
`stability.
`Tables I and II show that a minimum of about
`0.1% SDS (3.5 X 10-3 moles/L.) is needed for
`stability. The interfacial oil/water area in an
`emulsion can be calculated by
`
`6 X 1020
`Sv = --=-=---
`dvs
`
`where Sv is the specific surface area in square
`Angstroms per milliliter and d,s is the ~surface-
`
`~ l LQ(!)(!)CQ<:O
`
`·.3 .jc-Qt0t0~....i
`
`C\1....-i....-i....-i....-i
`
`~
`>.
`-~
`~r~~:--:~~0:~
`-~ J-.qiC\IC\1....-i....-i
`],;g
`~~ ~ 0':>-.qiO':>(!)....-i
`]~'<:'I~~~~~
`ol
`::t~
`
`.s l 00 lQ 0':> ..ql t-
`
`C\1....-i....-i....-i....-i
`
`~ J ~tOtO~,...;
`..,
`Cl)
`El
`ol
`~(.J...-i(!)J:--LQO':>
`§ l)oc-Q~~,...;
`~I '<:'I
`...-i
`Q)'"(j
`~0
`•
`...-iC\Il:--LQ...-i
`~.J ~~tO tO~
`
`00 l ~ ~~~c:~
`
`,..-ilQLQ..ql...-i
`C\1....-i....-i....-i....-i
`
`'<:'I
`
`. . . .
`
`CQ..qlt-OCO...-i
`.
`00(!)-.qiCQC\IC\1
`
`t- 0':>0 O':>C\Il''-
`
`0 • • • • •
`
`O':>CO<:OC\ICO...-i
`t-<:O..q!CQC\IC\1
`
`0 • • • • •
`
`cx:c:c:~~~
`COC\IO':>O':>l'-LO
`...-i...-i
`
`......
`<:000000<:00
`
`......
`LQLOJ:--OC\10
`
`.... '
`CQ..ql(!)...-iO...-i
`.
`CO...-iOOO':>OO<:O
`...-i...-i
`
`*
`C:CX:~t--:~..-:
`00...-i..q!C\IC\IC\1
`
`. . . . . .
`000-.qi(!)C\1-.qi
`
`000 000
`000000000000
`()) Q) Q) Q)
`Q) Q)
`
`tdtd~tdtdtd
`'8'8'2"8'8'2
`'O'Oaoo'O
`
`000000
`( f l ( f l ( f l ( f l ( f l ( f l
`;>,P.,P.,P.,P.,P.,
`
`~~~~~~
`lQ
`. . . . . .
`0...-ilOOOO
`000...-iC\IlO
`
`l:l
`
`.s 0 ~ 0 tiL--l h ~ ~ 0 pj H h
`a <:0(!)(!)<:0<:0 CO<:O<:O<:O(!)(!)
`~z ...-i...-i...-i...-i...-i
`
`LQLQLQLQLQLQ
`
`fil
`
`APOTEX 1032, pg. 4
`
`
`
`Vol. 54, No. 2, February 1965
`
`weighted mean diameter in microns. The mass
`median diameter, dm, may be substituted for dv• with
`little error. Taking 21.4 M as the mass median
`diameter of emulsion 61F, Sv is 0.28 X 1020 A. 2 /ml.
`In 100 ml. of emulsion, there are 25 ml. of oil, and the
`total oil/water interfacial area is 7 X 1020 A. 2.
`There is sufficient SDS present at 0.1% concentra(cid:173)
`tion, assuming 25 A. 2 surface coverage per molecule,
`to cover an area of 39 X 10 20 A. 2• Thus, there is
`more than :five times the amount necessary to form
`a condensed monolayer at the interface.
`By comparison, the polysorbate 80 stabilized
`emulsions are fairly stable after 1 month's aging
`when the aqueous surfactant concentration is at
`least 0.05% ( 4.1 X 10-4 moles/L. ). This concentra(cid:173)
`tion represents one-tenth the number of molecules
`.~ as are in 0.1% SDS; but, as shown by Fig. 2, the
`polysorbate 80 molecule is very bulky and con(cid:173)
`ceivably could cover 10 times as much surface as a
`SDS molecule.
`In addition, the surface activity of
`polysorbate 80 is greater than SDS at low concentra(cid:173)
`tions (12, 13); thus, a greater percentage of poly(cid:173)
`sorbate 80 molecules are adsorbed at the interface.
`The polysorbate 80 emulsions haveagreatertendency
`to form free oil by coalescence than the SDS-sta(cid:173)
`bilized emulsions (Table I), but the particle size
`distributions remain essentially constant, indicating
`perhaps uniform particle growth of all sizes with the
`number of smallest measured particles being re(cid:173)
`plenished by growth of those too small to measure.
`Another point of interest is that stable SDS emul(cid:173)
`sions have larger median diameters (consequently
`smaller total interfacial area) than the stable poly(cid:173)
`sorbate 80 emulsions at the same emulsifier concen(cid:173)
`tration. This may be due to a dependency on the
`kinetics of droplet formation during manufacture,
`especially at concentrations below
`the CMC.
`Smaller droplets may be formed more easily in the
`polysorbate 80
`formulations because of
`lower
`dynamic interfacial tensions. On the other hand,
`coalescence kinetics may be the most important fac(cid:173)
`tor. Assuming that equally small droplets are
`formed by the applied shearing forces regardless of
`the emulsifier, the polysorbate 80 :film could be more
`resistant to rupture or desorption by shear-induced
`collisions during formation than the SDS :film. The
`third, and perhaps most logical, explanation is .that
`
`Oil
`
`Water
`
`c(-0H
`/"
`~
`t>o
`~) .
`/0 ~·
`
`rf 0
`
`0__(0
`
`Sodium Dodecyl
`Sulfate
`
`'lo
`o..../O
`lfiO
`~o/'V Polysorbate 80 ~
`0~
`
`263
`
`SDS
`
`10
`8.0
`6.0
`
`4.0
`
`2.0
`
`~
`1.0
`z
`0.8
`u
`o.6
`z
`0 u
`
`0.4
`
`0::
`IJJ
`ii:
`())
`-I
`:::>
`:2:
`IJJ
`
`0.2
`
`0.1
`0.08
`0.06
`
`0.04
`
`0.02
`
`0. 01 [_'--'--'---'----'----'---'---'---'----'---'----'--'--'
`0 2 4 6 8 10 12 14 16 18 20 22 24 26
`MASS MEDIAN DIAMETER, p,
`
`Fig. 3.-Mass median diameter of mineral oil/
`water emulsions as a function of emulsifier con(cid:173)
`centration (data from Table II).
`
`SDS can stabilize larger droplets than polysorbate
`80. The ionic character of SDS creates an electrical
`double layer at the oil/water interface adding an
`electrostatic repulsion factor in addition to the :film
`barrier which reduces effective particle-particle
`collisions and
`inhibits coalescence. With
`this
`premise, the upper stable diameter limit is higher
`when SDS is the emulsifier.
`As the emulsifier concentration is increased, the
`mass median diameters of both series tend to de(cid:173)
`crease (Table II).
`Increasing surfactant concentra(cid:173)
`tion leads to greater adsorption at the oil/water
`interface (8, 9).
`Increased surfactant adsorption
`stabilizes 'more total oiljwater interfacial area; as a
`result, the particle size distributions have smaller
`median diameters. Logarithmic plots of surfactant
`concentration against mass median diameters are
`linear6 for both the 61 and 65 series (Fig. 3).
`The general equation for these curves is
`
`log C = log Co -
`
`kdm
`
`where k is the slope, C is the surfactant concentra(cid:173)
`tion, and log Co is the intercept. When C = Co, the
`mass median diameter, dm, is zero. 7 The relation(cid:173)
`ship holds for the emulsions at all ages. For SDS,
`the equation of the least-squares regression line is
`log C = 2.68 - 0.174 dm
`The correlation coefficient for the series is 0.985.
`In the case of polysorbate 80, the equation is
`log C = 2.01 - 0.237 dm
`with a correlation coefficient of 0.991.
`
`HOS
`Fig. 2.-Schematic representation of a typical
`polysorbate 80 molecule and a sodium dodecyl
`sulfate molecule at the oil-water interface.
`
`6 A plot of log surfactant concentration against log inter(cid:173)
`facial. area (related to 1/ dvs) is also linear and may be a more
`general empirical relationship.
`7 Co could have significance in terms of complete solubiliza(cid:173)
`tion of the oil phase, but the significance is not obvious from
`the data.
`
`APOTEX 1032, pg. 5
`
`
`
`264
`
`Journal of Pharmaceutical Sciences
`
`Emulsion data taken from the work of Riegelman
`and Pichon (14) also fit this type of plot (Fig. 4).
`They stabilized 50% mineral oil/water emulsions
`with a homologous series of cetyl alcohol poly(cid:173)
`ethylene glycol ethers over a surfactant concentra(cid:173)
`tion range of 0.25 to 4.0%. Using the midpoint of
`their surface average diameter ranges, the equation
`of the least-squares regression line is
`
`log C = 2.14 - 0.329 ds
`
`with a correlation coefficient of 0.999.
`It may be assumed safely that the increase of
`stabilized oil/water interfacial area, as surfactant
`concentration is increased above the CMC, is due
`to increased adsorption of dodecyl sulfate ion. The
`problem is whether the additional surfactant is
`adsorbed (a) as single ions forming a close-packed
`monolayer as illustrated in Fig. 5a or (b) as aggre(cid:173)
`gated or micellar groups (Fig. 5b ). Micellar ad(cid:173)
`sorption has been proposed, especially at the solid(cid:173)
`liquid interface, and experimental evidence has in-.
`dicated this (15, 16). However, micelles of SDS or
`polysorbate 80 have an ionic or polar surface, and
`it seems reasonable that they are not strongly ad(cid:173)
`sorbed at the hydrophobic oil surface.
`On the other hand, the generally recognized
`physical model (Fig. 5a) of a condensed surfactant
`monolayer with the hydrophobic portion of the
`surfactant at the oil/water interface and the hydro(cid:173)
`philic portion sticking outward into the water phase
`would seem to be preferred if there are additional
`single molecules available. While most of the
`additional surfactant in excess of the CMC aggre(cid:173)
`gates in the form of micelles, the monomer (single
`molecule) concentration can also increase (however
`small). The resultant increase in total surfactant
`available for adsorption would stabilize more and
`smaller particles during manufacture of the emul(cid:173)
`sion. Even without an increase in monomer con(cid:173)
`centration, greater micelle concentration can con(cid:173)
`ceivably cause a change in the dissolved monomer(cid:173)
`adsorbed monomer equilibrium without invoking
`
`10
`8
`6
`
`~4
`~2
`z
`0
`()
`0:: 1.0
`LLI 0.8
`~ 0.6
`en 5 0.4
`
`2
`LLI
`
`0.2
`
`0.1 .__ _ __L_ _ __L__~ __ .l..__ _ _.__~ _
`0
`2
`4
`6
`8
`10
`12
`SURFACE AVERAGE DIAMETER, f.L
`
`__ .
`
`14
`
`Fig. 4.-Surface average diameter of mineral oil/
`water emulsions as a function of emulsifier concen(cid:173)
`tration (data from Reference 14).
`
`(a.) Monom·eric
`
`(b.) Micellar adsorption
`Fig. 5.-Schematic cross-sectional view of sur(cid:173)
`factant adsorption at oil/water interface. Key:
`(a) oil particle stabilized by condensed monolayer
`of surfactant (symbolized by circle for polar or
`ionic head and straight tail for hydrocarbon chain);
`(b) oil particle stabilized partly b'y adsorbed single
`ions and partly by micellar aggregates.
`
`In either of the latter two
`micellar adsorption.
`possibilities, the activity of the surfactants may in(cid:173)
`crease with increasing concentration, perhaps as a
`log function of concentration,8 which would mean
`that the median diameter or total surface area of
`particles in an emulsion could be a simple function of
`the activity of the surfactant.
`
`REFERENCES
`
`(1) Knoechel, E. L., and Wurster, D., THIS JouRNAL,
`48, 1 (1959).
`(2) Langevin, L. M., ibid., 22, 728(1933).
`(3) Higuchi, W. I., Okada, R., and Lemberger, A. P.,
`ibid., 51, 683(1962).
`(4) Wachtel, R. E., and LaMer, V. K., J. Colloid Sci.,
`17, 531(1962).
`(5) Bulletin A-2, Coulter Industrial Sales, Elmhurst,
`Ill.
`
`(6) Cockbain, E. G., Trans. Faraday Soc., 48, 185(1952).
`(7). Void, R. D., and Groot, R. C., J. Soc. Cosmetic
`Chemists, 14, 233(1963).
`(8) Rehfeld, S. ]., J. Phys. Chem., 66, 168(1962).
`(9) Void, R. D., and Groot, R. C., ibid., 66, 171(1962).
`(10) Fischer, E. K., and Harkins, W. D., ibid., 36, 98
`(1932).
`(11) van den Tempel, Proc. 2nd Intern. Congr. Surface
`Activity, 1, 440(1957).
`(12) Adamson, A. W., "Physical Chemistry of Surfaces,"
`Interscience Publishers, Inc., New York, N. Y., 1960, p. 70.
`(13) Bulletin on Surface Active Agents, published by
`Atlas Chemical Industries, Wilmington, Del., 1950.
`(14) Riegelman, S., and Pichon, G., American Perfumer,
`77, 31(1962).
`(15) Fuerstenau, D. W., J. Phys. Chem., 60, 981(1956).
`(16) Ottewill, R. H., and Watanabe, A., Kolloid Z., 170,
`132(1960).
`. (17) Harned, H. S., and Owen, B. B., "Physical Chemistry
`of Electrolytic Solutions," 3rd ed., Reinhold Publishing
`Corp., New York, N. Y., 1958, p. 732.
`(18) Shinoda, K., et al., "Colloidal Surfactants,'' Academic
`Press Inc., New York, N. Y., 1963, p. 29.
`
`s The dependence of SDS or polysorbate 80 activity on con(cid:173)
`centration at higher concentrations is not known, but there
`are some data on smaller analogous compounds, e.g., a log(cid:173)
`arithmic plot of sodium heptylate concentration against
`activity is linear (data from Reference 17) above the calcu(cid:173)
`lated CMC (using equation 1.46 in Reference 18, pp. 42, 43).
`
`APOTEX 1032, pg. 6
`
`