Am J Clin Dermatol 2010; 11 (6): 413-421

ORIGINAL RESEARCH ARTICLE 1175-0561/10/0006-0413/$49.95/0

ª 2010 Adis Data Information BV. All rights reserved.

Characterization of the UVA Protection Provided by

Avobenzone, Zinc Oxide, and Titanium Dioxide in
Broad-Spectrum Sunscreen Products
Donathan G. Beasley and Thomas A. Meyer
Merck Consumer Care, Memphis, Tennessee, USA

Abstract Background: Solar UV radiation (UVR) is composed of UVB (290–320 nm) and UVA (320–400 nm)
wavelengths. Only two sunscreen active ingredients approved in the US, avobenzone (butylmethoxy-
dibenzoylmethane) and zinc oxide (ZnO), provide true broad-spectrum protection against UVA wave-
lengths >360 nm. Although effective against shorter UVR wavelengths <360 nm, titanium dioxide (TiO2) is
also often believed to confer broad-spectrum protection and is substituted for ZnO or avobenzone. To sus-
tain its absorption capacity within a sunscreen film during UVR exposure, avobenzone needs to be for-
mulated into sunscreen products using sound formulation strategies.
Objectives: To characterize the efficacy of avobenzone, ZnO, and TiO2 in terms of their abilities to provide
broad UVA protection and to demonstrate the effectiveness of the different formulation strategies used
today to maintain the efficacy of avobenzone even during prolonged exposures to UVR.
Methods: UVA efficacy was assessed by measuring absorbance profiles in vitro using Vitro Skin (IMS Inc.,
Orange, CT, USA) as an inert substrate and by determining UVA protection factors (PFA) on human skin.
The impact of avobenzone loss on sun protection factor (SPF) and PFA values was evaluated by serially
reducing avobenzone concentrations in an otherwise photostable product. The photostabilizing influence of
specific formulation ingredients was monitored by measuring the extent to which they prevented UVR-
induced degradation of avobenzone, whereas photostability of commercial sunscreen products was quan-
tified by measuring the percentage change in absorbance within the UVB and UVA spectral regions fol-
lowing irradiation of thin product films on inert substrates.
Results: Model formulations containing 3% avobenzone or 5% ZnO provided superior attenuation of UVA
wavelengths >360 nm compared with formulas containing 5% TiO2. Additionally, sunscreen products of similar
SPF containing avobenzone or ZnO exhibited significantly higher PFA values than those containing TiO2. The
addition of photostabilized avobenzone or ZnO increased PFA values nearly 3-fold, whereas the addition of
TiO2 increased PFA values only modestly. Judicious selection of sunscreen actives alone or in combination with
extra stabilizing agents maintained the photostability of avobenzone in formulations to deliver sustained broad-
spectrum absorbance during 4 hours of exposure to UVR. Small losses (<20%) of avobenzone did not signi-
ficantly reduce a product’s protective effects as measured by SPF and PFA values on human skin.
Conclusions: TiO2 provided neither the same level of UVA attenuation nor the same degree of UVA protection
on human skin as did products containing photostabilized avobenzone or ZnO. Hence, TiO2 cannot be
considered a substitute for avobenzone or ZnO in providing high levels of UVA protection to human skin. Use
of proper formulation strategies can ensure that avobenzone losses are minimized to the extent that they have no
impact on a product’s ability to deliver sustained protection, even over periods of prolonged exposure to UVR.

Solar UV radiation (UVR) is comprised of UVB (290– associated with erythema,[1] immunosuppression,[2] photoaging,[3]
320 nm) and UVA (320–400 nm) wavelengths, with the latter and skin cancer.[4] Although UVB is more potent in eliciting
accounting for about 94% of the UVR reaching the Earth’s sur- skin responses, such as erythema and edema, mounting scien-
face. Overexposure of unprotected skin to UVR induces damage tific evidence also suggests that UVA contributes significantly
414
to skin damage, including photodermatoses,[5] photoaging,[6-8]
and skin cancer.[9-11] The growing substantiation of the harmful
effects of UVA radiation provides a compelling reason for
sunscreen products to protect skin from both UVB and UVA in
a balanced manner. Sunscreen products with values of sun
protection factor (SPF) >15 must attenuate significant levels of
shorter UVA wavelengths (320–360 nm) in order to meet their
labeled SPF claim.[12] However, with the increasing recognition
of the need for greater levels of UVA protection, sunscreen
products in the US today typically contain either avobenzone
(butylmethoxydibenzoylmethane) or zinc oxide (ZnO) for ex-
tended UVA coverage beyond 360 nm.
Avobenzone and ZnO are the only UVA sunscreen actives
approved for use in the US capable of conferring protection
against UVA wavelengths >360 nm. Although effective against
UVR wavelengths <360 nm, titanium dioxide (TiO2) is some-
times considered to be an alternative option to avobenzone
or ZnO against wavelengths up to and beyond 360 nm. Avo-
benzone is an organic molecule and attenuates UVR by ab-
sorption, whereas ZnO and TiO2 are inorganic molecules that
attenuate UVR through absorption and scattering mechanisms.
US FDA guidelines currently exclude combinations of avoben-
zone with either of these inorganic actives in sunscreen prod-
ucts;[13] as a consequence, products usually contain only one of
these ingredients or possibly a combination of TiO2 with ZnO.
In a previous comparative study, ZnO was shown to provide
more effective UVA absorbance than TiO2, especially at the
higher UVA wavelengths >340 nm.[14] However, this study did
not compare the effectiveness of these ingredients against avo-
benzone, nor did it provide comparative measurements of UVA
protection factors on human skin. In order to sustain its absorp-
tion capacity within a sunscreen film during UVR exposure,
avobenzone needs to be formulated into sunscreen products
using sound formulation strategies. The objectives of this study
were to characterize the effectiveness of avobenzone, ZnO, and
TiO2 in conferring UVA protection and to demystify the main
formulation strategies used today to sustain the capacity of
avobenzone to absorb UVA even after prolonged exposure to
UVR.
Methods
Measurement of Absorbance Profiles
The absorbance profiles of sunscreen products were meas-
ured using Vitro Skin (IMS Inc., Orange, CT, USA) as the
substrate. Test lotions were drawn up into a syringe and applied
to 50 cm2 of substrate at an application density of 2 mL/cm2 by
ª 2010 Adis Data Information BV. All rights reserved.
Beasley & Meyer
regularly dotting the lotion across its surface and then spread-
ing the lotion evenly across the substrate using a finger covered
with a finger cot. After application, substrates were mounted in
6 · 7 cm photographic slide mounts and air-dried for 20 min-
utes. An untreated, hydrated Vitro Skin film served as a ref-
erence and was also slide-mounted and air-dried for 20 minutes.
An average absorbance profile (290–400 nm) for each product
film was determined using an SPF 290S Analyzer (Optometrics
LLC, Ayer, MA, USA) equipped with a computer-controlled
X-Y sampling stage that recorded 12 absorbance profiles from
different, nonoverlapping locations on each Vitro Skin film
mount. The variability in area under UVB and UVA spectral
regions between locations on each substrate was <4%, in-
dicating that products were spread uniformly across the sub-
strate’s surface and that the method yielded highly reproducible
results.
Preparation of Model Formulations
All model formulations were water-in-oil emulsions pre-
pared using 5% PEG-12 dimethicone cross-polymer as the
emulsifier and containing either 3% avobenzone (Neo Helio-
pan 357; Symrise, Inc., Teterboro, NJ, USA), 5% ZnO
(Z-Cote HP1; BASF Corp., Florham Park, NJ, USA) or 5%
TiO2 (TNP50M170; Kobo Products, Inc., South Plainfield, NJ,
USA) as single sunscreen active ingredients. The oil phase ad-
ditionally comprised 12% cyclopentasiloxane, 4% bis-hydroxyeth-
oxypropyl dimethicone, and 7.0–31.5% di-isobutyl adipate. The
water phase contained 5% propylene glycol and 0.5% sodium
chloride, which was added to the oil phase under constant mixing
to give final compositions.
Human UVA Testing
UVA protection factors (PFA) were determined using
methodology that employed delayed erythema as the biologic
endpoint,[15] as previously recommended by the FDA.[16] All
calculated PFA values represent the mean of five to ten separate
measurements. Studies have shown that in vivo methods based
on either delayed erythema or persistent pigment darkening as
endpoints to measure the extent to which sunscreen products
provide protection against UVA radiation yield comparable
results and may be used interchangeably.[17]
Photostability Assessment
Photostability of avobenzone in product vehicles was eval-
uated using two different methods. In one method, the per-
Am J Clin Dermatol 2010; 11 (6)
UVA Protection of Avobenzone, Zinc Oxide, and Titanium Dioxide 415

centage of the original avobenzone remaining after UVR ex- Results

posure was tracked to assess the impact of various formulation
ingredients on their ability to maintain the intact chemical
structure of avobenzone during irradiation. In this method,
known weights (14–18 mg) of avobenzone-containing sun-
screen formulations were spread onto glass microscope slides
using a finger cot-covered finger at an application density of
approximately 2 mg/cm2. The glass slides were then air-dried in
the dark for at least 20 minutes prior to irradiation. Non-
irradiated, product-treated glass slides served as controls and
were stored in the dark until extraction. Each product-treated
glass slide was positioned on a turntable and irradiated at a
distance of 120 cm from the UVR source, using a 1000 watt
xenon arc solar simulator (Spectral Energy, Washingtonville,
NY, USA) filtered with a WG320 filter (Light Company, Inc.,
Glenside, PA, USA). This configuration prevents heating is-
sues, as measured with a thermocouple on the substrate’s sur-
face, so that all substrates were maintained at ambient
temperatures during irradiation. Each product was tested in
triplicate and irradiated with doses of 20, 40, 60, and 80 joules/cm2.
The output of the source was measured using an OL 754
Spectroradiometer (Optronics Laboratories, Inc., Orlando,
FL, USA) prior to irradiation. Based on previously published
measurements[18] of outdoor sunlight, each 20-joule increment
(3.6 minimal erythemal doses) corresponds to an exposure of
about 1 hour of midday June sun in Memphis, TN, USA. After
irradiation, product-treated slides were extracted with isopro-
panol and UV spectra were recorded using a Lambda 40
spectrophotometer (PerkinElmer, Wellesley, MA, USA).
Avobenzone loss was monitored by following changes in ab-
sorbance at 360 nm, which was corrected for background
contributions from other sunscreen actives. Calculated values
of avobenzone represent the mean from three separate deter-
minations with standard deviations typically <2%.
Attenuation Profiles for Single Ingredients in a Model
Formula
The attenuation profiles of films of model sunscreen formulas
containing either avobenzone, ZnO, or TiO2 as single ingre-
dients at common usage concentrations are shown in figure 1.
Avobenzone at a 3% concentration provided greater attenu-
ation of radiation across the entire UVA spectrum than 5%
TiO2. In contrast, ZnO at a concentration of 5% appeared to
offer greater attenuation only at UVA wavelengths >360 nm
relative to 5% TiO2, although the magnitude of difference ob-
served with ZnO was lower than that observed with avobenzone.
Formulation Strategies to Preserve the Chemical Integrity
of Avobenzone
There are three main formulation strategies that can be used
to preserve the intact chemical structure of avobenzone during
UVR exposure, including (i) avoiding use of photochemically
incompatible ingredients; (ii) leveraging the stabilizing power of
sunscreen actives; and (iii) using additional photostabilizers
when sunscreen actives themselves are insufficient. The first
strategy is illustrated by the use of octinoxate as an incompat-
ible ingredient. When a thin film containing octinoxate and
5% TiO2
3% Avobenzone
5% ZnO
2.5
2.0
Absorbance units

The second method measured the photostability of com- 1.5

mercial sunscreen products over the full UV spectrum. In this
method, thin product films were prepared on Vitro Skin and
1.0
irradiated with increments of 20 joules/cm2 up to and including
100 joules/cm2 using the same calibrated solar simulator
and procedures as described above. Prior to and after each 0.5
successive increment of UVR exposure, an average absorbance
profile was computed from 12 individual scans from non-
0
overlapping spots on each substrate using the SPF 290S Ana- 300 320 340 360 380
lyzer and was used to calculate areas under the UVB UVB UVA
(290–320 nm) and UVA (320–400 nm) spectral regions. Photo- Wavelength (nm)
stability was then monitored by following changes in area
Fig. 1. Comparison of UV attenuation profiles for avobenzone, zinc oxide
under the UVB and UVA spectral regions as a function of UVR (ZnO), and titanium dioxide (TiO2) as single ingredients in a model for-
dose. mulation.

ª 2010 Adis Data Information BV. All rights reserved. Am J Clin Dermatol 2010; 11 (6)
416
Loss of octinoxate
2.5
2.0
Absorbance units
1.5 Loss of avobenzone
1.0
0.5
0
300 320 340 360
UVB
Wavelength (nm)
Fig. 2. Loss of absorbance in UVB and UVA spectral regions as octinoxate
and avobenzone undergo a destructive photochemical reaction in an emul-
sion film irradiated with UV radiation over 4 hours.
avobenzone was irradiated (figure 2), octinoxate reacted with
avobenzone photochemically in an irreversible manner, which
destroyed the UVR absorption properties of both molecules.
The incompatibility of octinoxate is believed to stem from the
presence of an exocyclic double bond, which undergoes an al-
lowed photochemical (2 + 2) reaction with avobenzone to form
primary photoproducts that subsequently collapse to form a
range of degradation products.[19-21]
The second strategy is exemplified in figure 3, which shows
that other sunscreen actives commonly used in the US can en-
hance the photostability of avobenzone. As a single sunscreen
active present in an emulsion, avobenzone degraded chemically
by about 60% after 1 hour and 90% after 4 hours of UVR
exposure. The addition of individual sunscreen actives to the
same emulsion substantially boosted the photostability of
avobenzone. Octocrylene (10%) maintained the chemical struc-
ture of avobenzone virtually intact, whereas oxybenzone (6%)
maintained about 70% of the original concentration of avo-
benzone after 4 hours of UVR exposure. Octisalate (5%) and
homosalate (12%) were much less effective but still provided
modest benefits compared with avobenzone alone. These re-
sults demonstrate the power of individual sunscreen actives to
photostabilize avobenzone when thin films of emulsions are
exposed to UVR over extended durations. When used in combi-
nation at lower concentrations (7% octocrylene, 4% oxy-
benzone, 5% octisalate, 13% homosalate), sunscreen actives
themselves provided complete stabilization of avobenzone
during UVR exposure (figure 4).
ª 2010 Adis Data Information BV. All rights reserved.
Beasley & Meyer
Before UV exposure
1-hour UV exposure
When the types and concentrations of sunscreen actives were
2-hour UV exposure insufficient (i.e. no octocrylene), addition of the special photo-
3-hour UV exposure
4-hour UV exposure
stabilizers diethylhexyl-2,6-naphthalate (DEHN) 5% or dieth-
ylhexylsyringylidene malonate (DESM) 2% augmented the
photostability of avobenzone by about 30% relative to the base
formula after 4 hours of UVR exposure (figure 5).
Comparison of Photostability Profiles for Commercial
Products
Three commercial sunscreen products sold in the US were
assessed for their ability to maintain their absorbance within
380 400
UVB and UVA spectral regions as the thin product films were
UVA exposed to varying doses of UVR. The types and concen-
trations of sunscreen actives contained in each product were
listed on their back labels, which included:
 Product M: SPF 45, containing 2% avobenzone, 7.5%
octinoxate, 6% oxybenzone, 12% homosalate, and 5%
octisalate;
 Product K: SPF 50, containing 3% avobenzone, 7%
octocrylene, 4% oxybenzone, 13% homosalate, and 5%
octisalate;
 Product G: SPF 45, containing 3% avobenzone, 2.4%
octocrylene, 6% oxybenzone, 15% homosalate, 5% octisa-
late, plus an unspecified amount of the stabilizer DEHN.
Figure 6 illustrates the extent to which the photostability of
sunscreen actives affected the absorbance properties of the
three commercial products. As expected, the nonphotostable
product (M) containing octinoxate lost significant absorption
With 10% octocrylene
With 6% oxybenzone
With 5% octisalate and 12% homosalate
Avobenzone alone
100
80
Avobenzone remaining (%)
60
40
20
0
0 1 2 3 4
Duration of UVR exposure (h)
Fig. 3. Individual sunscreen actives can improve the photostability of avo-
benzone in emulsion films irradiated over 4 hours with UV radiation (UVR).
Am J Clin Dermatol 2010; 11 (6)
UVA Protection of Avobenzone, Zinc Oxide, and Titanium Dioxide
Emulsion with 3% avobenzone, 5% octisalate,
13% homosalate, 4% oxybenzone, 7% octocrylene
Emulsion with 3% avobenzone alone
100
Avobenzone remaining (%)
80
60
40
20
0
0 1 2 3
Duration of UVR exposure (h)
Fig. 4. Mixtures of the right types and concentrations of sunscreen actives
as the only stabilizers can photostabilize avobenzone over 4 hours of UV
radiation (UVR) exposure compared with the same emulsion that contained
only avobenzone.
capacity within UVB and UVA regions, by 20% and 42% re-
spectively. In contrast, the photostable products (K, G), which
contained stabilizing levels of sunscreen actives and/or DEHN
as an extra stabilizer, sustained their absorption capacity with
only minor losses (1–5%) across both UVB and UVA regions.
Impact of Avobenzone Loss on Protection Factors
To illustrate how the protective qualities of a non-
photostable product can be adversely affected during UVR
exposure, we determined SPF and PFA values on human skin
for a series of photostable formulations, in which the concen-
tration of avobenzone (3%) in an original SPF 50 formula was
serially reduced by 20%, 33%, 66%, and 100% (i.e. removed
completely). Apart from the reduction in avobenzone, the
formulas were otherwise identical in the makeup of the other
sunscreen actives (7% octocrylene, 4% oxybenzone, 5% octi-
salate, 13% homosalate) and inactive ingredients. Since the
other sunscreen actives were unchanged in this experimental
approach, all the formulas maintained absorbance within the
UVB region to about the same extent while reducing their levels
of UVA absorbance. This enabled the impact of avobenzone
loss on PFA to be examined in a way that mimicked what could
occur on human skin if avobenzone degraded during periods of
UVR exposure.
The impact on PFA and SPF of reducing avobenzone con-
centrations in an SPF 50 sunscreen product appears in figure 7.
Reductions in avobenzone concentrations resulted in a linear
loss of protection, as measured by PFA and SPF. The loss of
ª 2010 Adis Data Information BV. All rights reserved.
417
protection observed as avobenzone was reduced became sig-
nificant when the PFA and SPF values were reduced by ‡33%
and ‡66%, respectively, corresponding to avobenzone reduc-
tions of 1% and 2%. Small losses in avobenzone (<20%) did not
appear to significantly reduce the product’s protective qualities.
Effectiveness of UVA Protection on Human Skin
As a final measure of efficacy, we tested the ability of several
commercially available products of comparable SPF (each with
avobenzone, TiO2, or ZnO as the main UVA sunscreen active
ingredient) to confer broad-spectrum protection by measuring
4 PFA values on human skin. Results appear in table I and
highlight several salient points. Products with 3% avobenzone
(product L) or 5% ZnO (product H) exhibited as much as twice
the PFA as those with varying concentrations of TiO2 (prod-
ucts B, D, and E). Product J, which contained only oxybenzone
(at the maximum allowed concentration) as the UVA sunscreen
active, had the lowest PFA value. The addition of photo-
stabilized avobenzone or ZnO to oxybenzone (products L and
H, respectively) greatly increased PFA values, compared with
the addition of TiO2 (product D).
Discussion
To our knowledge, this is the first study to compare the
UVA protection of avobenzone, ZnO, and TiO2 in a manner
that reflects real-world sunscreen product usage. Commercial
Emulsion with sunscreens + extra stabilizers
Emulsion with sunscreens (3% avobenzone,
5% octisalate, 15% homosalate, 6% oxybenzone)
3% avobenzone alone
100
Avobenzone remaining (%)
80
60
40
20
0
0 1 2 3 4
Duration of UVR exposure (h)
Fig. 5. Addition of diethylhexyl-2,6-naphthalate (5%) or diethylhexylsy-
ringylidene malonate (2%) as extra stabilizers to an emulsion that lacked the
right mixture of sunscreen actives (i.e. no octocrylene) can augment the
photostability of avobenzone over 4 hours of UV radiation (UVR) exposure.
Am J Clin Dermatol 2010; 11 (6)
418 Beasley & Meyer

sunscreen products typically contain avobenzone concen-
trations of 2–3%, ZnO concentrations of up to 5%, or TiO2
concentrations of 2–9%. UV attenuation profiles of model
formulas containing each sunscreen active at typical concen-
trations were measured, and the UVA absorbance efficiencies
of avobenzone, ZnO, and TiO2 were compared. The results
showed that both avobenzone and ZnO were more efficient
than TiO2 at attenuating UVA radiation, particularly at
wavelengths >360 nm (figure 1).
The attenuation profiles of TiO2 and ZnO shown here are
representative of the raw material grades used as sunscreen
actives and agree well with diffuse reflectance spectral com-
parisons of TiO2 and ZnO recorded from human skin in vivo.[14]
As inorganic particulates, TiO2 and ZnO attenuate UVR and
visible radiation by scattering and absorption, both of which
are affected by the size distribution of the particulates.[22,23]
Scattering of visible radiation promotes whitening effects on
human skin. By optimizing particle size, whitening can be
minimized while efficiency of UVR attenuation can be maxi-
mized. In the case of TiO2, optimization of particle size leads to
special grades for use as a sunscreen active that attenuate UVB
and short UVA wavelengths (<360 nm) more efficiently than
long UVA wavelengths (>360 nm). In contrast, the attenuation
profile of ZnO is less affected by particle size and, owing to its
lower refractive index (2.0 for ZnO vs 2.6 for TiO2), it is also less
whitening than TiO2.[23] Hence, by comparison, avobenzone
and ZnO attenuate longer UVA wavelengths more effectively
than attenuation grades of TiO2.
Although avobenzone is a highly efficient absorber of UVA,
its ability to sustain its absorption capacity during exposure to
UVR is highly dependent upon its chemical environment.[19-26]
Studies have established that the photostability of avobenzone
can be significantly improved by controlling solvent polarity in
solution,[19,21] by optimizing the dielectric constant of the oil
phase in emulsions,[27] and by including energy transfer agents
that quench the singlet or triplet excited states of avoben-
zone.[20,28] Since most photochemical reactions for organic com-
pounds occur from the longer-lived triplet state,[29] quenching
agents that either reduce the formation of (i.e. singlet quench-
ers) or directly quench the triplet state of avobenzone represent
highly effective mechanisms to preserve the absorption capacity
of avobenzone as a sunscreen active during exposure to UVR.
Thus, by following logical formulation strategies and by se-
lecting formulation ingredients judiciously, the photostability
of avobenzone can be preserved during exposures to UVR.
By far the most important formulation strategy to maintain
an intact chemical structure of avobenzone is to avoid use of
incompatible ingredients that react photochemically with avo-
benzone to destroy its absorption properties. This includes the
popular UVB sunscreen active octinoxate as well as the less

K
G
M

a b
110 110
UVB absorbance remaining (%)

UVA absorbance remaining (%)

100 100

90 90

80 80

70 70

60 60

50 50
0 1 2 3 4 0 1 2 3 4
Duration of UVR exposure (h)

Fig. 6. Percentage changes in integrated absorbances over (a) UVB and (b) UVA spectral regions as films of commercial products were irradiated with UV
radiation (UVR) over 4 consecutive hours. Owing to the presence of octinoxate, product M lost significant absorption in both UVB and UVA regions of the
spectrum. Conversely, products K and G sustained absorbance within UVB and UVA spectral regions over the entire time course of UVR exposure. Product K
leverages sunscreen actives, whereas product G relies on sunscreen actives combined with the stabilizer diethylhexyl-2,6-naphthalate to maintain the
photostability of avobenzone. Product M: SPF 45, containing 2% avobenzone, 7.5% octinoxate, 6% oxybenzone, 12% homosalate, and 5% octisalate; Product
K: SPF 50, containing 3% avobenzone, 7% octocrylene, 4% oxybenzone, 13% homosalate, and 5% octisalate; Product G: SPF 45, containing 3% avobenzone,
2.4% octocrylene, 6% oxybenzone, 15% homosalate, 5% octisalate, plus an unspecified amount of diethylhexyl-2,6-naphthalate.

ª 2010 Adis Data Information BV. All rights reserved. Am J Clin Dermatol 2010; 11 (6)
UVA Protection of Avobenzone, Zinc Oxide, and Titanium Dioxide 419

a b
20 54
18
16 50

PFA value

SPF value
14 * *
46
12 *
10 42
8 R2 = 0.9991 * R2 = 0.9995
*
6 38
0 20 40 60 80 100 0 20 40 60 80 100
Reduction in avobenzone concentration (%)

Fig. 7. Reductions in (a) UVA protection factor (PFA) and (b) sun protection factor (SPF) values as avobenzone concentration in an SPF 50 emulsion is serially
reduced from 3% to 0%. * p < 0.05 vs the original value with avobenzone 3%.

frequently used UVB sunscreen active 2-ethylhexyl-4-dimethyl-
aminobenzoate (padimate O).[26,30,31] In general, while they can
partially offset the loss of avobenzone, addition of extra sta-
bilizing agents typically cannot completely prevent avobenzone
from reacting photochemically with incompatible ingredients.
Hence, ingredients that are known or suspected of being photo-
chemically incompatible with avobenzone should either be
avoided or tested before adding them to a formulation.
Once incompatible ingredients are avoided, avobenzone in a
formula can be photochemically stabilized either by (i) the pre-
sence of the other sunscreen actives; or (ii) the inclusion of extra
quenching agents, such as DEHN,[28] DESM,[32] and polyester-8[20]
along with the sunscreen actives. The sunscreen actives
oxybenzone and octocrylene, believed to be efficient singlet and
triplet quenchers,[20] respectively, are especially effective pho-
tostabilizers for avobenzone. The results shown here (figure 6)
emphasize that the stabilizing power of the right mixture of
sunscreen filters themselves is as effective as the use of addi-
tional stabilizing agents such as DEHN to preserve the absorp-
tion properties of avobenzone in sunscreen products.
Several studies have documented that loss of absorbance due
to degradation of avobenzone during UVR exposure decreased
the protective capacity of sunscreen films. One study reported
significant reductions in protection toward several biologic
endpoints within reconstructed skin in vitro,[33] while other
studies observed significant reductions in SPF following irra-
diation on inert substrates.[21,26,30] We report here that reducing
avobenzone by as much as 20% in a photostable SPF 50
sunscreen product does not significantly change either SPF or
exposure do not compromise the protective qualities of a sun-
screen film on human skin, while larger losses clearly com-
promise both UVB and UVA protection. While we understand
that the results presented here for this specific high SPF product
may not translate directly to other SPF products with different
sunscreen actives, they do clearly demonstrate that through
careful formulation the photostability of avobenzone can be
maintained at a level where minor fluctuations have no impact,
for all practical purposes, on a product’s ability to deliver
sustained protection even over long periods of UVR exposure.
As a final comparison of efficacy, we confirmed that com-
mercial sunscreen products containing TiO2 did not confer the
same degree of UVA protection to human skin as did those
containing either ZnO or photostabilized avobenzone. The
results for different commercial sunscreen products of similar
SPF (table I) demonstrated convincingly that addition of TiO2
yielded PFA values on human skin that fell markedly below
products that contained either avobenzone or ZnO. Indeed,
products containing TiO2 provided PFA values that were more
similar to products containing oxybenzone, a sunscreen active
that, like TiO2, effectively attenuates UVA wavelengths up to
Table I. Concentrations of UVA actives present in commercial sunscreen
products accompanied by their labeled sun protection factor (SPF) claim and
determined UVA protection factor (PFA)
Product Oxy (%) Avo (%) TiO2 (%) ZnO (%) SPF PFA
J 6.0 45 6.0
B 2.4 50 10.5

PFA levels from their original values on human skin (figure 7). D 6.0 5.9 50 9.5

Reductions in avobenzone affected PFA more than SPF values, E 9.1 30 8.4
as expected, because avobenzone predominantly absorbs long- L 4.0 3.0 50 18.2
er UVA radiation whereas erythema is caused mainly by UVB H 6.0 5.0 50 16.0
plus short UVA wavelengths. Our results show for the first time Avo = avobenzone; Oxy = oxybenzone; TiO2 = titanium dioxide; ZnO = zinc
that small losses in avobenzone that may occur during UVR oxide.

ª 2010 Adis Data Information BV. All rights reserved. Am J Clin Dermatol 2010; 11 (6)
420
360 nm. Even a product containing 9.1% TiO2 did not provide
PFA values comparable to products containing avobenzone or
ZnO. These studies demonstrate that TiO2 should not be con-
sidered an equivalent substitute for either avobenzone or
ZnO in its ability to protect skin from the damaging effects
of UVA radiation or to imbue products with full-spectrum
protection.
Conclusions
As a practical matter in developing sunscreen products over
a range of SPF values, the photostability of avobenzone can be
easily maintained using sound formulation strategies to provide
sustained UVA absorbance, even over prolonged exposures to
UVR. Avobenzone and ZnO are the only UVA sunscreen ac-
tives currently approved by the FDA that provide true broad-
spectrum protection beyond 360 nm. TiO2 provided neither the
same level of UVA attenuation nor the same degree of UVA
protection on human skin as products containing photo-
stabilized avobenzone or ZnO. For these reasons, TiO2 cannot
be substituted for avobenzone or ZnO in sunscreen products to
confer the same level of broad-spectrum protection.
Acknowledgments
Donathan G. Beasley and Thomas A. Meyer are employees of Merck
Consumer Care. This study was funded by Merck Consumer Care, which
was responsible for the design and conduct of the study; collection, man-
agement, analysis, and interpretation of the data; and preparation, review,
and approval of the manuscript. Editorial assistance in the preparation of
this article was provided by BioScience Communications, New York, NY,
USA.
References
1. Urbach F. The negative effect of solar radiation: a clinical overview. In:
Giacomoni PU, editor. Sun protection in man, ESP comprehensive series
in photosciences. Vol 3. Amsterdam: Elsevier Sciences, 2001: 41-67
2. Kaija K, Hanneman DO, Cooper KD, et al. Ultraviolet immunosuppression:
mechanisms and consequences. Dermatol Clin 2006; 24: 19-26
3. Fisher GJ. The pathophysiology of photoaging of the skin. Cutis 2005; 75: 5-9
4. De Gruijl FR. Photocarcinogenesis: UVA vs. UVB radiation. Skin Pharmacol
Appl Skin Physiol 2002; 15: 316-20
5. DeLeo V. Sunscreen use in photodermatoses. Dermatol Clin 2006; 24: 27-33
6. Sander CS, Chang H, Salzman S, et al. Photoaging is associated with protein
oxidation in human skin in vivo. J Invest Dermatol 2002; 118: 618-25
7. Vielhaber G, Grether-Beck S, Koch O, et al. Sunscreens with an absorption
maximum of ‡360 nm provide optimal protection against UVA1-induced
expression of matrix metalloproteinase-1, interleukin-1 and interleukin-6 in
human dermal fibroblast. Photochem Photobiol Sci 2006; 5: 275-82
ª 2010 Adis Data Information BV. All rights reserved.
Beasley & Meyer
8. Schroeder P, Gremmel T, Berneburg M, et al. Partial depletion of mitochon-
drial DNA from human skin fibroblasts induces a gene expression profile
reminiscent of photoaged skin. J Invest Dermatol 2008; 128: 2297-303
9. Haliday GM, Agar NS, Barnetson RSC, et al. UVA fingerprint mutations in
human skin cancer. Photochem Photobiol 2005; 81: 3-8
10. Runger TM. C-T transition mutations are not solely UVB-signature muta-
tions, because they are also generated by UVA. J Invest Dermatol 2008; 128:
2138-40
11. Sander SS, Hong C, Hamm F, et al. Role of oxidative stress and the antioxidant
network in cutaneous carcinogenesis. Int J Dermatol 2004; 43: 326-35
12. Sayre RM, Dowdy JC, Lott DL, et al. Commentary on ‘UVB-SPF’: the SPF
labels of sunscreen products convey more than just UVB protection. Photo-
dermatol Photoimmunol Photomed 2008; 24: 218-20
13. Food and Drug Administration. Sunscreen drug products for over-the-counter
use: final monograph. Fed Reg 1999 May 21; 64 (98): 27666-93
14. Pinnell SR, Fairhurst D, Gillies R, et al. Microfine zinc oxide is a superior
sunscreen ingredient to microfine titanium dioxide. Dermatol Surg 2000;
26: 309-13
15. Cole C. Multicenter evaluation of sunscreen UVA protectiveness with the
protection factor test method. J Am Acad Dermatol 1994; 5 (1): 729-36
16. Food and Drug Administration. Sunscreen drug products for over-the-counter
use: amendment to the tentative final monograph, enforcement policy. Fed
Regist 1998 Oct 22; 63 (204): 56587
17. Stanfield J, Edmonds S, Agin P. An evaluation of methods for measuring
sunscreen ultraviolet A protection factors. In: Lowe NJ, Shaath NA,
Pathak MA, editors. Sunscreens: development, evaluation and regulatory
aspects. New York: Marcel Dekker, 1997: 537-58
18. Beasley DG, Beard J, Stanfield JW, et al. Evaluation of an economical sunlamp
that emits a near solar UV power spectrum for conducting photo-
immunological and sunscreen immune protection studies. Photochem Pho-
tobiol 1996; 64 (2): 303-9
19. Schwack W, Rudolph T. Photochemistry of dibenzoyl methane UVA filters:
part I. J Photochem Photobiol B: Biol 1995; 28: 229-34
20. Bonda CA. Research pathways to photostable sunscreens. Cosmet Toil 2008;
123 (2): 49-60
21. Huong SP, Rocher E, Fourneron JD, et al. Photoreactivity of the sunscreen
butylmethoxydibenzoylmethane (DBM) under various experimental con-
ditions. J Photochem Photobiol A: Chem 2008; 196: 106-12
22. Stamatakis P, Palmer BR, Bohren CF, et al. Optimum particle size of titanium
dioxide and zinc oxide for attenuation of ultraviolet radiation. J Paint Tech
1990; 62 (789): 95-8
23. Schlossman D, Shao Y. Inorganic ultraviolet filters. In: Shaath NA, editor.
Sunscreens: regulation and commercial development. 3rd ed. Boca Raton
(FL): Taylor & Francis, 2005: 239-79
24. Tarras-Wahlberg N, Stennhagen G, Larko O, et al. Changes in ultraviolet
absorption of sunscreens after ultraviolet irradiation. J Invest Dermatol 1999;
113: 547-53
25. Deflandre A, Lang G. Photostability assessment of sunscreens: benzyl-
idene camphor and dibenzoylmethane derivatives. Int J Cosmet Sci 1998; 10:
53-62
26. Diffey B, Stokes RP, Forestier S, et al. Suncare product photostability: a key
parameter for a more realistic in vitro efficacy evaluation. Eur J Dermatol
1997; 7: 226-8
27. Bonda CA. The photostability of organic sunscreen actives: a review.
In: Shaath NA, editor. Sunscreens: regulation and commercial development.
3rd ed. Boca Raton (FL): Taylor & Francis, 2005: 321-52
28. Bonda CA, Steinberg DC. A new photostabilizer for full spectrum sunscreens.
Cosmet Toil 2000; 115 (6): 37-45
29. Turro NJ. Modern molecular photochemistry. Menlo Park (CA): Benjamin/
Cummings Publishing Co., 1978: 329
Am J Clin Dermatol 2010; 11 (6)
UVA Protection of Avobenzone, Zinc Oxide, and Titanium Dioxide
30. Maier H, Schauberger B, Brunnhofer K, et al. Change of ultraviolet absorb-
ance of sunscreens by exposure to solar-simulated radiation. J Invest Derm
2001; 117: 256-62
31. Sayre RM, Dowdy JC. Photostability testing of avobenzone. Cosmet Toil
1999; 114 (5): 85-91
32. Chaudhuri RK, Lascu Z, Puccetti G, et al. Design of a photostabilizer having
built-in antioxidant functionality and its utility in obtaining broad-spectrum
sunscreen formulations. Photochem Photobiol 2006; 82: 823-8
ª 2010 Adis Data Information BV. All rights reserved.
421
33. Marrot L, Belaidi JP, Lejeune F, et al. Photostability of sunscreen products
influences the efficiency of protection with regard to UVR-induced genotoxic
or photoaging-related endpoints. Br J Dermatol 2004; 151: 1234-44
Correspondence: Dr Thomas A. Meyer, SkinCare R&D, Merck Consumer
Care, 3030 Jackson Avenue, Memphis, TN 38151, USA.
E-mail: tom.meyer@merck.com
Am J Clin Dermatol 2010; 11 (6)