ARTICLES

PUBLISHED ONLINE: 9 MAY 2016 | DOI: 10.1038/NMAT4635

An elastic second skin

Betty Yu1, Soo-Young Kang1, Ariya Akthakul2, Nithin Ramadurai2, Morgan Pilkenton1, Alpesh Patel1,
Amir Nashat2, Daniel G. Anderson3,4,5,6, Fernanda H. Sakamoto7,8, Barbara A. Gilchrest8,
R. Rox Anderson4,7,8 and Robert Langer3,4,5,6*

We report the synthesis and application of an elastic, wearable crosslinked polymer layer (XPL) that mimics the properties of
normal, youthful skin. XPL is made of a tunable polysiloxane-based material that can be engineered with specific elasticity,
contractility, adhesion, tensile strength and occlusivity. XPL can be topically applied, rapidly curing at the skin interface without
the need for heat- or light-mediated activation. In a pilot human study, we examined the performance of a prototype XPL that
has a tensile modulus matching normal skin responses at low strain (<40%), and that withstands elongations exceeding
250%, elastically recoiling with minimal strain-energy loss on repeated deformation. The application of XPL to the herniated
lower eyelid fat pads of 12 subjects resulted in an average 2-grade decrease in herniation appearance in a 5-point severity
scale. The XPL platform may offer advanced solutions to compromised skin barrier function, pharmaceutical delivery and
wound dressings.

T
he skin is a versatile organ with multiple physiological Existing solutions fall short of addressing both the mechanical
functions, performed in the context of a fluctuating and physiological functions of normal healthy skin. Individuals with
environment. Importantly, the skin is a visible organ and, compromised skin barrier function, for example, still rely on the
thus, makes a psychosocial statement about the health of the use of cumbersome occlusive dressings, often in combination with
individual. The primary function of skin is to provide a protective topical ointments that often fail owing to poor patient compliance3 .
barrier that mitigates exposure to external factors such as extreme Although existing flexible and skin-adherent pre-formed films
temperatures, toxins, microorganisms, radiation, and mechanical interface with the skin15,1921 , these materials are not designed to
insult, while maintaining the internal milieu. serve as a second skin that can be worn invisibly to restore normal
The important role of skin is inevitably compromised by skin recoil and aesthetics.
ageing, environmental insult, and certain disease states. Conditions The design and adoption of such a second skin with respect to
characterized by skin barrier dysfunction remain prevalent in a wearable, skin-conformal material that is topically applied pose
the population14 . Cutaneous water loss is clinically observed as several fundamental challenges. First, the material must be safe on
chronic inflammation and pruritus that are, in turn, responsible skin without irritation and sensitization. Second, the material must
for decreased quality of life, deranged sleep patterns, and mood be incorporated in a topical formulation that is readily spreadable on
disturbances that include anxiety and depression5 . Skin ageing, skin and form a safe second skin in situ. Third, the material must
whether chronological or accelerated by sun exposure, creates adhere to skin, while providing a breathable, but protective, interface
persistent problems that will be faced by all members of society6 . to the environment. Fourth, the material must possess mechanical
As the skin ages, it loses biomechanical integrity, while also properties that accommodate normal skin mechanical responses to
becoming more susceptible to environmental injury6 and impaired motion while reinforcing inherent skin tension and elastic recoil.
wound healing7 . Fifth, the material must mimic, or at least not interfere with, the
The ageing-related changes in skin mechanical integrity result appearance of healthy normal skin, for a wide range of individual
primarily in loss of elastic recoil and baseline laxity. This loss of skin types.
elastic recoil is not reversible by existing therapies8 and consequently On the basis of the above design challenges, we selected
remains the universal fate of our largest and most visible organ9 . the XPL chemistry for siloxane polymers and their associated
Recent progress in materials science and engineering has yielded crosslinked networks in part as a result of the numerous safety
flexible skin interfacing technologies1012 that are functionalized in and biocompatibility data existing in the literature22,23 . In addi-
some cases for pharmaceutical delivery13 , rapid analyte detection14 , tion to the established safety profile, silicone crosslinked networks
continuous physiological monitoring of medical conditions15 , and offer readily tunable architectures that enable the development
wound healing1618 . Unfortunately, these advances have not focused of materials that are optimizable for a myriad target mechanical
on restoring the mechanical and aesthetic properties of the skin to and transport properties, such as flexibility, elongation, elastic-
match its original, youthful state. ity, toughness, adhesion and moisture/oxygen permeability. The

1 Living Proof, Inc., Cambridge, Massachusetts 02142, USA. 2 Olivo Laboratories, LLC, Cambridge, Massachusetts 02142, USA. 3 Institute for Medical

Engineering and Science, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA. 4 Harvard-MIT Division of Health Science and
Technology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA. 5 David H. Koch Institute for Integrative Cancer Research,
Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA. 6 Department of Chemical Engineering, Massachusetts Institute of
Technology, Cambridge, Massachusetts 02139, USA. 7 The Wellman Center for Photomedicine, Massachusetts General Hospital and Harvard Medical
School, Boston, Massachusetts 02114, USA. 8 Department of Dermatology, Massachusetts General Hospital and Harvard Medical School, Boston,
Massachusetts 02114, USA. *e-mail: [email protected]

NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials 1

2016 Macmillan Publishers Limited. All rights reserved
ARTICLES NATURE MATERIALS DOI: 10.1038/NMAT4635

a 16 b 1.0

(i) Baseline (ii) XPL, 2 min (iii) XPL, 3 h

Normalized storage modulus (Pa Pa1)
XPL 80%
14 normalized
Cica-Care
0.8 modulus, Modulus
12 Tegaderm plateau
t = 175 s
Stress (MPa)

10 0.6
8 30%
normalized
6 0.4
modulus,
t = 120 s
4
0.2 Activation
2
0 0.0
0 1 2 3 4 5 6 100 0 100 200 300 400 500
Strain (mm mm1) Reaction time (s)

Figure 1 | Mechanical properties of XPL and its crosslinking kinetics. a, Representative tensile stressstrain plots for XPL and two commercial pre-formed
wound dressings: Cica-Care silicone sheets and Tegaderm polyurethane sheets with acrylic adhesive backing. Topically applied XPL and pre-formed
Cica-Care silicone sheets exhibit good elastic recoil until break. In contrast, pre-formed Tegaderm polyurethane sheets exhibit plastic deformation profiles
that yield after 15% strain. b, XPL crosslinking kinetics. Crosslinking kinetics of the topical Step 1 formulation containing active prepolymers are represented
by the normalized XPL storage modulus as a function of reaction time. The S-shape curve was obtained using the rheometer test for crosslinking kinetics
(see Methods). Moduli are normalized by the storage modulus measured at the last time point. In the method, three distinct regions of the reaction are
observed with respect to the rate of increase in storage modulus (the slope of the reaction curve). The first region captures the catalyst-stabilizer effect,
where the reaction is suppressed by chemically stabilizing the platinum catalyst to allow sufficient sample preparation time before the rheometer records
the beginning of the actual cure kinetics. The steep increase in the storage modulus after 100 s signals the rapid crosslinking from hydrosilylation over the
next 100 s. After 200 s, the storage modulus increases asymptotically as the crosslinking reaction approaches completion. Photographs illustrating the
impact of modulus build on compressing the herniated fat pad are shown. (i) Prominent fat pad protrusion (lower lid bag severity grade 4 is defined by the
concavity at the lower tear trough at baseline). XPL was applied to the region demarcated by the dashed line. (ii) Two minutes following XPL application,
the skin texture was smoother and the skin laxity seen by the mild skin wrinkling at the medial and central fat pads was tightened. (iii) At 3 h following XPL
application, the peak compression shows a 2-grade improvement in bag severity using a standardized scoring system.

room-temperature, platinum-catalysed hydrosilylation chemistry Synthesis of network architecture that optimizes elastic response.
we selected does not yield reaction by-products that are typical Uniaxial tensile testing that investigated skin anisotropy with
of condensation reactions, nor is it dependent on additional en- respect to the Langer lines (topological lines corresponding
ergy sources, such as heat or ultraviolet, required to initiate the to structural orientation of collagen fibres in the dermis) at
topical crosslinking reaction in situ. Therefore, the XPL can be different body locations and orientations reported a linear elastic
safely and easily deposited on skin in situ using a two-step topical deformation zone at low strains (up to 1040%) and corresponding
delivery system, where the first step comprises the application of elastic moduli ranging from 0.5 to 1.95 MPa, with the resulting
a flowable, skin-conformal, reactive polysiloxane component that fracture strain measured at 140180%24 . On the basis of these
is then crosslinked on exposure to the platinum catalyst that is established skin mechanical properties, the target XPL modulus
introduced in the second step. The two-step system also enables space was identified at 0.5 to 1.95 MPa, and the elastic strain region
optical modulation of the second skin by depositing light scattering was specified to be greater than 180%.
particles at the XPL surface from the second step delivery vehicle. To identify polymer compositions exhibiting the target design
The complement of the surface optical particles and the underlying criteria, a material library was generated by systematically varying
crosslinked polysiloxane network generates a specular and diffuse the following parameters: relative functional (vinyl- and hydride-)
reflectance approximating the appearance of youthful skin. polysiloxane chain lengths; the crosslink density as a function of
In essence, the XPL presented in this study is a skin-adherent, the corresponding functional polysiloxane concentrations; and the
three-dimensional (3D) polymer network that is chemically concentration of reinforcing fumed silica. A plot of the fracture
crosslinked in situ on skin, following a two-step topical application strain as a function of the tensile modulus illustrates the mechanical
process. Each of the materials selected in the formulations is design space spanned by the materials synthesized from the screen
either on the US Food and Drug Administrations list of Generally (see Supplementary Fig. 7), where more than 100 reactive polymer
Regarded as Safe substances, or has individual safety documents that blend (RPB) compositions yielded tensile moduli ranging from 0.1
demonstrate safety for leave-on skin applications. Here, we report to 2.5 MPa and fracture strains as high as 800%. On the basis of an
the design and development of a biomimetic skin-conforming analysis of reported values, we restricted our analysis of the library
crosslinked polymer material that can be formulated for easy to formulations with tensile moduli between 0.4 and 0.8 MPa. A lead
topical application to form an invisible elastic film in situ. The RPB candidate, with tensile modulus of 0.48 MPa, fracture strain of
film can be safely worn on the skin, restores normal skin aesthetics, 826%, and adhesive strength of 78 N mm1 , that also satisfied the
and exerts stresses in the plane of the skin to change its form. The elasticity criteria (plastic strain of 1.3% and hysteresis strain-energy
tunability of the XPL chemistry and its corresponding ingredients in density loss of 0.66 kJ m3 following cyclic loading at 15% strain) was
the topical formulation render this second skin platform potentially selected on the basis of an optimization of the material performance
useful for a wide range of cosmetic and medical applications. attributes desired for a biomimetic second skin (see Supplementary
Table 1). This lead RPB was later incorporated with other excipients
Results into the final two-step delivery system to form the XPL in situ.
The development of a wearable second skin began with the Although the fumed silica imparts mechanical toughness to
design and optimization of XPL materials to mimic the mechanical the XPL films, it also exponentially increased the viscosity of the
properties of natural skin, with respect to elastic recoil, flexibility RPB. As a result, the lead RPB (containing 27% w/w fumed silica)
and elongation, to achieve optimal skin-conformal performance. demonstrated poor topical spreading properties, with a viscosity of

2 NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials

2016 Macmillan Publishers Limited. All rights reserved

NATURE MATERIALS DOI: 10.1038/NMAT4635 ARTICLES
XPL No treatment
a
With XPL application No treatment

Baseline
b

Figure 2 | The visual impact of a 2-grade improvement (from grade 3 to

grade 1) after applying XPL to the under-eye area. This image is an
Immediately after release
illustrative example of a 2-grade improvement, which was consistently d
observed following repeated split-face application of XPL (n > 10).

600 Pa s1 (600,000 cP) measured at 0.5 s1 . At such a high viscosity, 3 s after release
the RPB spreadability on skin approximated that of highly viscous
resins such as tar pitch. The detailed formulation strategies of this Figure 3 | Time-lapse photos extracted from video footage of skin
material are discussed in the Methods. retraction following a dermatological tenting test. XPL was applied to the
right side. The left side served as the untreated control. ad, The four
Mechanical properties of the XPL. Three key mechanical design frames extracted show the left and the right side before tenting (a), during
criteria in the development of the skin mimetic XPL required that the tenting test (gentle pinch) (b), immediately after the skin is
the elastic recoil, flexibility and elongation approximated the values released (c), and 3 s after the skin is released (d). The side wearing the XPL
of natural skin subjected to an initial elastic large deformation demonstrates minimal plastic strain and retracts to the original position,
zone (strain under 40% and tensile modulus of 0.51.95 MPa)25,26 . whereas the untreated side demonstrates poor elastic recoil and sustained
In Fig. 1a, we report that the XPL maintained elastic recoil the plastic strain resulting from the imposed stress. These results were
until break (fracture strain: 250 82%) and the tensile modulus consistent with repeated tenting studies (n > 10).
(0.51 0.01 MPa) measured was comparable to that of natural skin.
The measure of elastic recoil (elasticity) was based on the energy personal care products such as make-up, sunscreens and barrier
loss during the loadunload cycles of the material, quantifiable creams. Polyurethane film cast from Avalure UR 450 polymer
from the area under the hysteresis loop of the stressstrain profile demonstrated a tensile modulus that was greater than 30 times
during tensile testing (hysteresis strain-energy density loss). The that of natural skin (37.6 MPa) and that yielded after 15% strain.
measured XPL elasticity was 16.8 0.8 kJ m3 of hysteresis strain- Supplementary Fig. 9 compares the stressstrain hysteresis loops
energy density loss, following 25 loadunload cycles at 100% strain. for polyurethane film cast from Avalure UR 450 polymer and the
We tested elasticity at 100% strain to capture a conservative safety XPL, where polyurethane Avalure UR 450 yielded with a 12% loss
factor of material performance that stretches beyond the reported in tensile modulus and plastically deformed with 2% plastic strain,
elastic region for normal skin (less than 40% strain). after only 15 loadunload cycles at 15% strain.
With the above mechanical properties to match those of In light of the mechanical characteristics exhibited by the
natural skin, we compared the XPL as a wearable skin-conforming commercial benchmarks, these results further reinforced the
material with example commercial benchmarks of film wear on uniqueness of the XPL composition as a topically applied,
skin, either through pre-formed wound dressings or topical film skin-adherent, elastomeric network with mechanical properties
former formulations. optimized to match those of natural skin in motion. Moreover, the
First, the XPL was compared with two commercial benchmarks XPL demonstrated superior skin cosmesis when compared with
of pre-formed wound dressings: Cica-Care silicone gel sheets (Smith the commercial benchmarks. Pre-formed Cica-Care silicone sheets
and Nephew) and Tegaderm polyurethane sheets coated with acrylic are 20 times thicker (1,300 m) and visibly prominent on the skin,
adhesive (3M). As shown in Fig. 1a, similar to the XPL, pre- resulting from the glossy, light reflective optics that also delineate the
formed Cica-Care silicone elastomer sheets exhibited robust elastic film boundary. The 40-m-thick films (Tegaderm and polyurethane
recoil until break (fracture strain: 553 82%), with an order film cast from Avalure UR 450 polymer), which approximated the
of magnitude lower tensile modulus (0.08 0.04 MPa) compared XPL thickness, were also highly visible on skin as a result of the
with that of natural skin (0.51.95 MPa). These results confirmed mismatched film optics, which were further accentuated by the
that the topically deposited XPL provides a 3D silicone network generation of film wrinkling patterns on skin in motion.
comparable to pre-formed commercial silicone elastomer sheets,
with mechanical properties much closer to those of natural skin. In vivo application and use. This section describes four in vivo
In contrast to the Cica-Care, pre-formed Tegaderm polyurethane studies evaluating the XPL utility as a natural-looking, elastic sec-
sheets exhibited plastic deformation profiles that yield after 15% ond skin. Pilot human studies were performed to measure changes
strain, highlighting an elastic strain region much less than that of in the skin mechanical properties and barrier function after wearing
natural skin. Moreover, the measured Tegaderm tensile modulus XPL on the lower lids where under-eye fat pads may be prominent
of 8.85 2.20 MPa was approximately 7 times greater than natural (Study A) and on normal volar forearm skin (Study B). To further
skin. For comparison, Tegaderm exhibited substantially greater investigate the XPL as a second skin for conditions of compromised
plastic deformation than XPL following 25 loadunload cycles at barrier function, transepidermal water loss (TEWL) was measured
100% strain, as seen by the order of magnitude increases in the in clinically dry leg skin (Study C). Finally, the clinical performance
hysteresis strain-energy density loss and plastic strain measured at of the XPL was evaluated in a double-blind, randomized, placebo-
237.0 25.5 kJ m3 and 13.56 1.25%, respectively. controlled study to demonstrate the translation of the above de-
Poor elasticity was also measured for the film deposited from the sign criteria (safe, topically formed in situ, skin-adherent, skin-
topical film-forming waterborne polyurethane dispersion Avalure conforming barrier protection, and natural-looking) to the skin of
UR 450 polymer (Lubrizol), a raw material ingredient used in topical the lower lid application sites during normal wear (Study D).

NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials 3

2016 Macmillan Publishers Limited. All rights reserved
ARTICLES NATURE MATERIALS DOI: 10.1038/NMAT4635
a 700 b 700
Baseline Viscoelastic region Baseline 24 h with XPL
600 Elastic linear region Hysteresis 600 35 min after XPL removal

Suction pressure (kPa)

Suction pressure (kPa)

500 500

400 400

300 300

200 200

100 100

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Time (s) Time (s)

c Baseline 24 h 35 min after XPL removal d Baseline 24 h 35 min after XPL removal e XPL versus UC p < 0.001
1,400 500 P versus UC p < 0.05
XPL versus P p < 0.05
450 400
1,200

Change in conductance (S)

Suction pressure (kPa)
400
Retraction time (ms)

1,000
350 300
300
800
250
600 200
200
400 150
100 100
200
50
0 0 0
XPL Twice No XPL Twice No 1 2 3 4 5 6 7
moisturizer treatment moisturizer treatment Time (h)

Figure 4 | Pilot study evaluating XPL-induced changes in skin elasticity (Study B) when worn over 24 h on normal volar forearm skin. a, At baseline, a
representative plot of the negative pressure required to extend the skin to a fixed height using the suction cup method is shown. The two distinct regions
highlighted in red and blue correspond to the viscoelastic and the elastic skin response regions, respectively. The hysteresis experienced following each
skin distension cycle is shown by the dashed blue line. b, Following 24 h of XPL wear, the treated skin sites demonstrated increased elastic behaviour,
showing a linear elongation response with time (blue squares) for the same representative skin sites. Thirty-five minutes following XPL removal, after 24 h
of wear (filled diamonds), the skin approached the baseline behaviour. c, The average skin retraction time was measured for six different individuals. The
three treatments were applied to each arm, yielding duplicate treatment sites assigned using a predetermined rotational scheme for each subject. The sites
exposed to XPL demonstrated a marked decrease in the skin retraction time (from 838 164 ms to 533 57 ms, p < 0.005, paired Students t-test)
compared with the baseline values, unlike the 12% lactic acid moisturizer applied twice, initially and after 12 h (879 174 ms to 851 164 ms, not significant
(NS)) and the untreated control sites (893 140 ms to 795 104 ms, NS). Thirty-five minutes after XPL removal, the skin recoil time and the corresponding
elongation pressures returned to the baseline values. The negative pressure recorded at 12% distension similarly decreased 29% (from 382 42 kPa to
273 22 kPa). d, In the same six subjects, p < 0.05 compared with the baseline values, unlike the twice-treated moisturizer sites (418 27 kPa to
396 23 s, NS) and the untreated control sites (432 39 kPa to 423 39 kPa, NS). e, Changes in skin hydration were evaluated in 24 subjects following
application of XPL and petrolatum with post hoc analysis (Tukey method, = 0.05). XPL (116.65 S at 1 h) demonstrated significant increases in skin
hydration compared with petrolatum (P) (42.65 S at 1 h; p < 0.05), and untreated control sites (UC) (21.95 S at 1 h; p < 0.001). The error bars represent
the standard errors of the mean. Data sets marked with asterisks ( ) represent statistically significant differences using a paired Students t-test.

Study A examined mechanical reshaping of the lower lid region. XPL reaction proceeds. To study the crosslinking kinetics of the
Study A explored the feasibility of reshaping of the lower lid region XPL after the two-step topical delivery system on skin, the relative
over time following the application of the XPL and the subsequent storage modulus of the XPL was measured in vitro as a function
progression of the crosslinking reaction in situ. In older individuals, of time, following the introduction of the platinum catalyst, also
there is often a protrusion of the fat pad underlying the skin in shown in Fig. 1b. The start of the first inflection region of the curve
the lower lid area. In this study, a 5-point 04 photonumeric lower corresponds to the initiation of the crosslinking reaction, which is
lid herniation scale was introduced to quantify the changes in the analogous to the exposure of the Step 1 formulation on skin to the
extent of fat pad herniation following XPL treatment. We reasoned Step 2 formulation comprising 200 ppm platinum catalyst. The steep
that the protrusion of the fat pad underlying the skin in the lower increase in the normalized storage modulus over approximately
lid area could potentially be corrected by a steady, compressive 100 s provides an indication of how quickly the XPL is generated
force imparted to the skin by the XPL. In our two-step topical in situ. The final region of the S-shape curve represents a slower rate
delivery system, the compressive force results from the collapse of change in the modulus as the reaction asymptotically approaches
of the polymer coil network during solvent loss, rather than by completion. The delay in the onset of crosslinking indicated at times
the chemical bond formation during hydrosilylation. Therefore, t < 0 was the effect of the catalytic stabilizer introduced in the
in the selection of the lead XPL studied here, the solvents and sample preparation protocol to allow for sufficient working time
their concentrations in the formulations were optimally selected to conduct experimental measurement. At t = 0 s, the crosslinking
to provide sufficient compression-based shrinkage of the skin with reaction was initiated. The storage modulus approached 80% of
minimal discomfort. the plateau value at t = 175 s and then levelled off, corresponding
As seen in Fig. 1b, perceivable reshaping of the lower lid region to the approach of reaction completion. The photos of the subject
for a subject with severe bags (severity grade of 4) occurs as the shown at baseline, 2 min after XPL application, and 3 h following

4 NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials

2016 Macmillan Publishers Limited. All rights reserved

NATURE MATERIALS DOI: 10.1038/NMAT4635 ARTICLES
8
Baseline 2h 24 h the skin demonstrates linear elastic behaviour, corresponding to
the initial linear increase in negative pressure with time; at higher
Mean transepidermal water loss (g m2 h1)

7
p < 0.001 p < 0.05
strain, the skin exhibits viscoelastic behaviour with a corresponding
6 nonlinear negative pressure dependence on time29 . Skin hysteresis
p < 0.01 was also observed over the five strain cycles, as characterized by the
5 successive decrease in negative pressure required to induce the fixed
strain (Fig. 4a).
4 The skin retraction properties corresponding to XPL-treated
skin sites following 24 h of wear demonstrated a consistent, linear
3
response (Fig. 4b). Compared with the untreated skin sites at
2 baseline, the consistent linear slope resulting from sequential strain
cycles with minimal hysteresis demonstrates the linear elastic
1 behaviour exhibited by the bulk skin when the XPL is worn. With
the XPL, faster retraction times were also measured along with lower
0 negative pressures required to achieve the target distension height.
XPL Untreated control Petrolatum Commercial
moisturizer Thirty-five minutes following XPL removal, the skin distension
curve (filled diamonds) approached the baseline value, while not
Figure 5 | Clinical data showing enhanced skin barrier function and fully exhibiting the baseline viscoelastic behaviour, suggesting a
sustained skin hydration for subjects (n = 22) with dry skin (Study C). transient skin memory resulting from the prior 24 h of XPL wear.
At two hours, the sites treated with petrolatum and XPL demonstrated In summary, the XPL-coated skin exhibited greater elasticity than
significantly reduced TEWL values compared with the untreated control normal, uncoated skin.
values ( p < 0.05 and p < 0.001). A reduction in the TEWL was sustained For the six subjects, significant decreases in the average skin
with the XPL over the course of 24 h ( p < 0.01). At 24 h, the differences retraction time and the average vacuum pressure corresponding
measured for P and for commercial moisturizer were not statistically to the 12% strain were also measured (36% and 28%, respectively
significant when compared to the UC. The error bars represent one in Fig. 4c,d), compared with the corresponding baseline values
standard deviation from the mean. The p values were calculated on the (Fig. 4b). Both measurements returned to the baseline values at
basis of one-way ANOVA analyses. 35 min following XPL removal.
Skin hydration resulting from XPL wear was evaluated in 24
XPL application depict the immediate impact of the XPL on flaccid volunteers, aged 2255, using a skin surface hygrometer to measure
skin, followed by the impact on bag compression that occurs over a skin conductance. The change in skin conductance from baseline
longer time frame (3 h to peak). This immediate effect reflects the values was calculated for sites treated with the XPL and with
rapid XPL crosslinking kinetics depicted in Fig. 1b, where close to the petrolatum at baseline, 1, 2, 4 and 6 h following test article
80% of the final elastic modulus value was achieved over the course application (see Fig. 4e). Significant increases in skin conductance
of 2 min. Notably, the fat pad herniation demonstrated a 2-grade were measured for the petrolatum- and the XPL-treated sites
improvement, shown additionally in Fig. 2. Such a result has been (p < 0.05 and p < 0.001, respectively). Moreover, the XPL-treated
previously achieved only by a lower lid blepharoplasty, an invasive sites demonstrated significantly higher increases in conductivity
surgical procedure. Sequential frames extracted from a time-lapse over the six hours of test article wear compared with the petrolatum-
video (Fig. 3) demonstrated recovery of skin elasticity after XPL treated sites (p < 0.05). The skin hydration properties of the XPL
application to lax skin (Fig. 3a). The skin was tented (pinched) for 3 s were further complemented by the measurement of TEWL in
with equal pressure on the XPL-treated (right) and XPL-untreated subjects with clinically dry skin, described below.
(left) lower lids and then released (Fig. 3b). Immediately after the
skin was released (Fig. 3c), and 3 s (Fig. 3d) following release, there Study C examined improved TEWL for dry skin. Study C explored
was a clear difference in the skin recoil between both sides. Whereas the XPL aspect as a second skin barrier against TEWL. The
the right lower lid (XPL applied) retracted to its original shape, the XPL barrier effect was measured on the legs of 22 volunteers
untreated lid remained tented as a result of the remaining plastic with moderately to severely dry skin at the test sites, based on a
strain from the imposed stress. 9-point visual dryness scale. In this study, the XPL was compared
with petrolatum, the most effective skin occlusive reported in
Study B examined improved elastic recoil of volar forearm skin. the literature30 , and a cosmetically elegant, high end commercial
Study B explored the mechanical impact of the XPL layer on skin moisturizer. An untreated control site served as the fourth test site.
over a 24-h wear period and after XPL removal, to evaluate the TEWL values were measured at baseline, 2 h, and at 24 h following
XPL-induced modulation of skin elasticity during wear and the test article application (Fig. 5). Two hours following test article
potential residual effects on the underlying internal skin stresses application, the petrolatum- and the XPL-treated skin sites showed
following XPL removal. The clinical changes in skin mechanical significantly decreased TEWL values relative to the untreated
properties for six subjects were evaluated at baseline, following 24 h control sites (p < 0.05 and p < 0.001, respectively), whereas the
of XPL wear, and 35 min following XPL removal. Using a suction cosmetically elegant moisturizer did not. At 24 h following test
cup device, the volar arm skin was distended to a fixed position article application, a 23% decrease in the TEWL values from the
of 12% strain and then released. This test procedure enables the baseline (0.95 0.41 g m2 h, p < 0.01) persisted at the XPL sites,
measurement of bulk skin retraction properties, a clinical indicator whereas there was no decrease in TEWL compared to baseline at
of skin mechanical function akin to the dermatologic skin tenting the other sites. The results confirmed the XPL as a second skin
test. The negative pressure detected in the suction chamber scales barrier that protects the skin from excessive moisture loss to the
with the skin distensibility, and the retraction times provide an environment.
indication of skin elasticity27,28 . Figure 4a shows a representative skin
distensibility plot over five consecutive cycles of vacuum application Study D examined second skin performance with normal wear.
to a representative untreated skin site at baseline. For the first cycle, Study D explored the clinical performance of the XPL during
the elastic and the viscoelastic regions are highlighted by the solid normal wear in reshaping the lower lid. Notably, the elastic and
blue and red lines, respectively. As expected, under low strains, adhesive properties of the XPL detailed above are important, as

NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials 5

2016 Macmillan Publishers Limited. All rights reserved
ARTICLES NATURE MATERIALS DOI: 10.1038/NMAT4635
a Baseline With XPL d 4
Baseline p < 0.0005
1h 4h p < 0.0005
# p < 0.001
3

Lower lid severity score

#

2
SC #
E XPL
D
E SC
D
1
b Baseline With XPL

0
Bag Bag Wrinkle Wrinkle
(XPL) (vehicle) (XPL) (vehicle)

e 8
XPL Vehicle
7

Surface irregularity (102)

6
c Baseline With XPL 5

0
0 1 2 3 4
Time (h)

Figure 6 | Second skin performance with normal wear. ac, Representative 3D photography (upper panels) and non-invasive imaging of the underlying
epidermal layers using OCT (lower panels) for three subjects at baseline and 4 h after XPL application. The blue and red lines illustrate the two lengths
used to calculate the surface irregularity. The three vertical pairs of metallic markers that were placed along the lower lid delineate the three different OCT
measurement sites that were evaluated at each side. The two arrows in the OCT image correspond to the placement of one pair of markers. After 4 h, it is
possible to visualize the XPL film over the stratum corneum. The stratum corneum (SC), epidermis (E), dermis (D) and XPL regions are identified in a. Pixel
intensity analysis of the three OCT sections acquired at each application side at each time point for the twelve subjects suggested a greater level of
hydration resulting from the XPL application compared with the vehicle at 4 h (25,888 3,207 versus 28,774 3,535, respectively, p < 0.0001, paired
Students t-test), where lower pixel intensities represent a greater fraction of dark pixels, indicative of high water content. At baseline, the average pixel
intensities measured were 31,477 3,708 (XPL) and 30,878 3,821 (vehicle), translating to decreases of 18% and 7% in the pixel intensities, respectively.
d, Double-blind, randomized, placebo-controlled pilot clinical trial to evaluate XPL contractile function at the lower lid using 04 photonumeric scales for
quantifying extent of skin response. At 1 h (2.38 0.51) and at 4 h (1.23 0.60) following test article application, the XPL demonstrated statistically
significant (p < 0.0005) 1-grade and 2-grade improvements compared with the corresponding vehicle grades of 3.23 0.44 and 2.31 0.63, respectively.
The baseline average values for the XPL and the vehicle sites were 3.31 0.48 and 3.23 0.44, respectively. The lower lid wrinkle values for the vehicle at
baseline, 1 h and 4 h are 2.92 0.64, 2.23 0.60, and 1.92 0.64, respectively. For the XPL treatment sites, the wrinkle grades at the corresponding time
points are 2.85 0.55, 1.77 0.73 and 1.0 0.58, respectively. At four hours, a significant improvement in wrinkles was observed (p < 0.001) for the XPL
side compared with the vehicle. The error bars represent one standard deviation of the mean. The symbols ( , , #) identify data sets demonstrating
statistical differences using a paired Students t-test. e, The surface irregularity values measured from the triplicate OCT images obtained at each
treatment side show significantly decreased skin surface roughness at four hours following XPL application compared with the vehicle control
(2.18 102 1.00 102 versus 3.89 102 3.27 102 , respectively, p < 0.01). Zero surface irregularity values correspond to smooth surfaces with no
measurable roughness. Baseline surface irregularity values showed no significant differences between the two groups (4.30 102 2.34 102 versus
4.08 102 1.96 102 , respectively), validating the randomization scheme used. At one hour, the difference measured between the two groups trended
towards significance (2.13 102 1.16 102 versus 2.59 102 1.38 102 , respectively, p = 0.18). The error bars denote one standard deviation of
the mean. The p values were calculated using paired Students t-tests.

facial expressions invoke repeated stress and strain to the under-eye The placebo control (the vehicle-treated group) comprised a
area, creating a challenge to the XPL second skin durability during similar two-step formulation from which the catalyst was removed
normal wear. to prevent network crosslinking. The blinded and trained graders
To begin characterizing the XPL-induced changes in skin found significant improvements in bag severity at 1 and 4 h
mechanics and cosmesis, two 04 photonumeric scales were after XPL application (Fig. 6d). The modest and statistically
developed to enable quantification of XPL-induced skin shape nonsignificant improvements from baseline in the vehicle-treated
changes by trained observers. The potential mechanisms responsible sites are likely to be due to the skin hydration properties of
for the observed effects were further studied using optical coherence XPL and/or to the light scattering ingredients incorporated in the
tomography (OCT) to non-invasively examine the underlying vehicle delivery of our two-step system. At one hour following test
changes in the hydration and the mechanics of the skin that article application, both the XPL and the vehicle groups showed
interfaces the XPL second skin. improvements in their average wrinkle severity grade, which did not

6 NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials

2016 Macmillan Publishers Limited. All rights reserved

NATURE MATERIALS DOI: 10.1038/NMAT4635 ARTICLES
differentiate the effect of the XPL and the vehicle (not significant). At the catalyst for crosslinking (vehicle-treated control), and hence, not
four hours, the continued improvement at the XPL-treated sites was attributable to the product moisturizing properties alone. A signif-
observed to exceed the effects perceived at the vehicle-treated sites icant increase in the surface smoothness was also measured for the
(p < 0.0001). Examples correlating the improvement of the lower skin sites exposed to the XPL compared with the vehicle. Together,
lid following XPL application to the underlying changes in skin the double-blinded studies distinguish the contractile, mechanical
hydration are shown by pairing photographs with the corresponding effects achieved by the XPL crosslinked polysiloxane network on
OCT images (Fig. 6). the cosmetic appearance of the skin from the vehicle effects that are
Over the course of four hours, the XPL-treated sites were likely to arise from skin hydration and optical manipulation of the
significantly smoother than the vehicle control sites (Fig. 6e, skin surface. The visible flattening of the lower lid fat pad following
2.18 102 versus 3.89 102 , respectively, p < 0.01). The surface 3 h of XPL application would also suggest that internal stress, or the
irregularity calculations also suggested a transient improvement XPL shrinkage, resulting from the volatiles loss is another important
in surface smoothness for the vehicle-treated sites compared with design parameter deserving additional characterization for effecting
the baseline at 1 h (p < 0.001), which was likely to be an effect changes in skin shape.
of hydration. By 4 h, however, the vehicle-treated site surface The XPL also significantly decreased moisture loss from the
irregularity approached the corresponding baseline value, whereas skin, as evidenced by OCT imaging that quantified the increased
the XPL-treated sites sustained their surface smoothness relative state of epidermal hydration at sites exposed to the XPL. The
to baseline (p < 0.0001). This result is consistent with the visual OCT measurements were further supported by the quantitative
assessment of a smoother skin surface with fewer irregularities at measures of TEWL. Of note, the XPL performed better than a
4 h after XPL application, compared with the vehicle. leading commercial moisturizer and a conventional highly occlusive
The OCT images were further analysed to quantify the change in agent, petrolatum, over the course of 24 h. Indeed, the elasticity and
pixel intensity resulting from the XPL and the vehicle treatments. durability of XPL film on skin should also contribute to consistent
Lucent pixels depict regions of low water content, whereas dark barrier properties of the film, during wear. By comparison, the
pixels indicate regions of increased hydration (Fig. 6). Within the topical creams such as petrolatum or inelastic film formers such
viable epidermis, there was a significant decrease (p < 0.00001, see as Avalure UR 450 polyurethane exhibit plastic deformation that
Supplementary Table 2) in the OCT pixel intensity during the 4 h potentially may decrease the coating uniformity during wear. These
following the XPL application compared with the baseline values localized variations in the film integrity may produce weak spots
(Fig. 6ac), which we suggest was due to the enhanced skin barrier that compromise barrier protection.
imposed by XPL wear, as suggested by the TEWL measurements. Overall, the observed elastic skin memory is reasoned to result
At 4 h, the average intensity for the XPL sections analysed was 10% from the XPL-induced increases in skin hydration and/or the
lower than for the vehicle sections (p < 0.0001, see Supplementary transient changes to the underlying skin tensile properties while
Table 2), further supporting our interpretation that the XPL wear interfacing the tangential XPL contractile stresses imposed at the
increases skin hydration. skin surface. The potential of the XPL to impose a mechanical skin
In another study evaluating the XPL durability, 23 out of 25 memory that translates to a transient change in the underlying
subjects demonstrated excellent film integrity at the end of a 16-h skin mechanical behaviour following removal of the XPL will be
wear period. Two of the 25 subjects showed a visible film boundary the subject of future studies examining the long-term benefits to
at the lower lid application site, although the XPL remained intact skin following daily, repeated XPL wear. The association between
and adherent to the skin. Repeated daily wear resulted in no report anisotropic skin mechanical behaviour and the Langer lines is
of irritation or other adverse events resulting from the XPL use. well documented in the literature26 . Skin anisotropy and the loss
Furthermore, the XPL remained intact following activities such as of skin mechanical integrity with age have been attributed to
swimming, running and exposure to rain. changes in the elastin and collagen architecture that constitute
the dermis. The application of a thin, contractile layer to the
Discussion stratum corneum leads to isotropic, tangential stresses exerted by
We report here that a new, two-step, flowable polysiloxane- the XPL in the plane of the skin. The role of these tangential
based formulation, applied topically and crosslinked in situ to stresses on modulating the anisotropic skin stress tensor deserves
form an invisible thin film on human skin, can provide a additional investigation. New devices that measure the in-plane
durable, skin-conforming elastic second skin. This second skin is stresses inherent to skin have further reinforced the anisotropic
wearable, moisturizing, safe, well tolerated, and provides enhanced tension underlying skin mechanics31 and may be leveraged to
mechanical integrity to the underlying skin. characterize the in-plane mechanical changes induced by the XPL.
The XPL material design criteria for tensile modulus and elastic Moreover, coupling these new techniques with additional detailed
recoil were selected to mimic youthful skin properties, and the analyses of skin adhesion and the histology associated with the skin
extent of elongation was selected to provide comfort through the biomechanical responses utilizing human skin explant models may
range of motion the skin experiences. The material described here provide additional insight into the XPL-induced changes on the
demonstrated a tensile modulus of 0.5 MPa, which was designed to skin architecture. The elucidation of the mechanisms underlying the
cover the range of elastic response in normal healthy skin reported in observed changes in the bulk skin mechanical behaviour resulting
the literature24,25 . Such elastic recoil behaviour, and the 250% elastic from the introduction of a thin skin-conforming layer at the surface,
strain region until fracture demonstrated by this material, are likely along with ongoing efforts to delineate the associated spatial changes
to be design parameters that promote the seamless integration of in skin geometry, may offer additional insights into the material
the XPL at the skin interface, thereby allowing for even extreme design criteria required for target indications addressed by the next-
ranges of motion while providing the instantaneous recoil that is generation XPLs.
characteristic of youthful, intact skin. A wearable material layer that restores the normal skin
The proof-of-concept human studies showed that XPL contrac- mechanics and appearance is unprecedented and provides a
tion reliably reshapes the skin surface, as well as mechanically flat- platform to solve other cutaneous problems. Although the focus of
tening herniated fat pads that cause under-eye bagging. A double- this work has been to elucidate the impact of the elasto-mechanical
blind, randomized, placebo-controlled study documented a marked properties of the XPL on skin, further directions of research
decrease in under-eye bag severity and wrinkling that was signifi- and improvement to create better products for barrier protection
cantly greater than that produced by the same composition without and cosmetics present unique opportunities to enhance patient

NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials 7

2016 Macmillan Publishers Limited. All rights reserved
ARTICLES NATURE MATERIALS DOI: 10.1038/NMAT4635

compliance and quality of life. One such improvement might include 18. Davis, S. C. et al. The healing effect of over-the-counter (OTC) wound healing
extending the XPL wear beyond the current once-daily application agents applied under semi-occlusive film dressing. Br. J. Dermatol. 172,
specification. In addition, skin-adherent and skin-conforming 544546 (2015).
19. Borde, A. et al. Increased water transport in PDMS silicone films by addition of
polymers utilizing a similar concept could be developed as durable excipients. Acta Biomater. 8, 579588 (2012).
ultraviolet protection or as concealers of vascular anomalies such 20. Hammock, M. L., Chortos, A., Tee, B. C., Tok, J. B. & Bao, Z. 25th anniversary
as port-wine stains. Devices that deliver pharmaceutical actives article: the evolution of electronic skin (e-skin): a brief history, design
or assay biomarkers at the skin surface are additional potential considerations, and recent progress. Adv. Mater. 25, 59976038 (2013).
applications. In summary, the XPL second skin concept described 21. Webb, R. C. et al. Ultrathin conformal devices for precise and continuous
here could potentially yield a next generation of functionalized thermal characterization of human skin. Nature Mater. 12, 938944 (2013).
22. Linear Polydimethylsiloxanes CAS No. 63148-62-9 (European Centre for
second skins with broad medical benefits while allowing skin to look Ecotoxicology and Toxicology of Chemicals, 2011).
and behave naturally. 23. Lewis, L. N., Stein, J., Gao, Y., Colborn, R. E. & Hutchins, G. Platinum catalysts
used in the silicones industry. Their synthesis and activity in hydrosilylation.
Methods Platinum Met. Rev. 41, 6675 (1997).
Methods and any associated references are available in the online 24. Annaidh, A. N., Bruyere, K., Destrade, M., Gilchrist, M. D. & Ottenio, M.
Characterization of the anisotropic mechanical properties of excised human
version of the paper. skin. J. Mech. Behav. Biomed. Mater. 5, 139148 (2012).
25. Agache, P. G., Monneur, C., Leveque, J. L. & Rigal, J. D. Mechanical properties
Received 15 May 2015; accepted 31 March 2016; and Youngs modulus of human skin in vivo. Arch. Dermatol. Res. 269,
published online 9 May 2016 221232 (1980).
26. Hendriks, F. M. et al. A numerical-experimental method to characterize the
References non-linear mechanical behaviour of human skin. Skin Res. Technol. 9,
274283 (2003).
1. Irvine, A. D., McLean, W. H. & Leung, D. Y. Filaggrin mutations associated
27. Manschot, J. F. & Brakkee, A. J. The measurement and modelling of the
with skin and allergic diseases. N. Engl. J. Med. 365, 13151327 (2011).
mechanical properties of human skin in-vivoII. The model. J. Biomech. 19,
2. Schon, M. P., Boehncke, W. H. & Brocker, E. B. Psoriasis: clinical
517521 (1986).
manifestations, pathogenesis and therapeutic perspectives. Discov. Med. 5,
28. Diridollou, S. et al. In vivo model of the mechanical properties of the human
253258 (2005).
skin under suction. Skin Res. Technol. 6, 214221 (2000).
3. Schon, M. P. & Boehncke, W. H. Psoriasis. N. Engl. J. Med. 352,
29. Cua, A. B., Wilhelm, K. P. & Maibach, H. I. Elastic properties of human skin:
18991912 (2005).
relation to age, sex, and anatomical region. Arch. Dermatol. Res. 282,
4. Bieber, T. Atopic dermatitis. N. Engl. J. Med. 358, 14831494 (2008).
283288 (1990).
5. Yosipovitch, G. & Bernhard, J. D. Clinical practice. Chronic pruritus. N. Engl. J.
30. Kraft, J. N. & Lynde, C. W. Moisturizers: what they are and a practical approach
Med. 368, 16251634 (2013).
to product selection. Skin Ther. Lett. 10, 18 (2005).
6. Gilchrest, B. A. Skin aging and photoaging: an overview. J. Am. Acad.
31. Boyer, G. et al. Assessment of the in-plane biomechanical properties of human
Dermatol. 21, 610613 (1989).
skin using a finite element model updating approach combined with an optical
7. Balin, A. K. & Pratt, L. A. Physiological consequences of human skin aging.
full-field measurement on a new tensile device. J. Mech. Behav. Biomed. Mater.
Cutis 43, 431436 (1989).
27, 273282 (2013).
8. Anderson, R. R. Lasers for dermatology and skin biology. J. Invest. Dermatol.
133, E21E23 (2013).
9. Rushmer, R. F., Buettner, K. J., Short, J. M. & Odland, G. F. The skin. Science Acknowledgements
154, 343348 (1966). The authors thank G. Grove and Z. D. Draelos for their helpful discussions and M. Su for
10. Kim, D. H., Xiao, J., Song, J., Huang, Y. & Rogers, J. A. Stretchable, curvilinear her assistance with the in vivo use studies.
electronics based on inorganic materials. Adv. Mater. 22, 21082124 (2010).
11. Rogers, J. A., Someya, T. & Huang, Y. Materials and mechanics for stretchable Author contributions
electronics. Science 327, 16031607 (2010). B.Y., S.-Y.K., F.H.S., B.A.G. and R.R.A. contributed to the design and analysis of the
12. Jang, K.-I. et al. Soft network composite materials with deterministic and in vivo use studies. A.A., N.R. and M.P. contributed to the design and development of
bio-inspired designs. Nature Commun. 6, 6566 (2015). material synthesis and topical formulation. A.A., N.R., M.P., A.N. and D.G.A. contributed
13. Mitragotri, S., Blankschtein, D. & Langer, R. Ultrasound-mediated transdermal to the characterization and analysis of the in vitro mechanical and rheological data.
protein delivery. Science 269, 850853 (1995). S.-Y.K. supervised the execution of the in vivo use studies. A.P. conducted in vivo Study B.
14. Kost, J., Mitragotri, S., Gabbay, R. A., Pishko, M. & Langer, R. Transdermal B.Y. and R.L. managed the research efforts. B.Y., A.A., A.N., B.A.G., R.R.A. and R.L.
wrote the manuscript with the help of the co-authors.
monitoring of glucose and other analytes using ultrasound. Nature Med. 6,
347350 (2000).
15. Xu, S. et al. Soft microfluidic assemblies of sensors, circuits, and radios for the Additional information
skin. Science 344, 7074 (2014). Supplementary information is available in the online version of the paper. Reprints and
16. Clark, R. A., Ghosh, K. & Tonnesen, M. G. Tissue engineering for cutaneous permissions information is available online at www.nature.com/reprints.
wounds. J. Invest. Dermatol. 127, 10181029 (2007). Correspondence and requests for materials should be addressed to R.L.
17. Meddahi-Pell, A. et al. Organ repair, hemostasis, and in-vivo bonding of
medical devices by aqueous solutions of nanoparticles. Angew. Chem. Int. Ed. Competing financial interests
53, 63696373 (2014). All of the authors hold a financial interest in Living Proof and/or Olivo Labs.

8 NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials

2016 Macmillan Publishers Limited. All rights reserved

NATURE MATERIALS DOI: 10.1038/NMAT4635 ARTICLES
Methods Smith & Nephew (Cica-Care Silicone Gel Sheet). The bi-phasic remover was
XPL two-step delivery system formulation development. To coat the skin with a provided by Living Proof.
thin and uniform layer of the lead RPB, formulation design efforts focused on the
development of a new emulsion system comprising an aqueous-compatible internal Formulation procedures. Lead RPB emulsion or Step 1. In the main mixer, the
phase and a siloxane-compatible external phase. The formulation development lead RPB and emulsifiers were combined with the silicone phase diluents until a
efforts required the incorporation of an internal phase thickener that enabled the homogeneous mixture was visible. In a side mixer, the aqueous phase containing
appropriate shear-thinning effects to facilitate the deposition of the siloxane phase. the polyacrylic acid hydrogel and preservatives were mixed until the aqueous
When the siloxane phase was deposited too rapidly onto the skin, the uniformity of dispersion was homogeneous. The aqueous phase was slowly introduced to the
the XPL was compromised, yet a prolonged application playtime interfered with main mixer containing the silicone phase to form a white creamy emulsion. Finally,
the mechanical integrity of the crosslinked layer. Moreover, we constrained the the emulsion was neutralized to target pH of 5.5 to 6.5.
total amount of non-volatile silicone-compatible ingredients that could act as
plasticizers in the final XPL film to 2% (w/w) or less. Optimization of the Platinum catalyst delivery system or Step 2. In the main mixer, the polysiloxane
formulation flow properties and the siloxane deposition time was achieved by phase components and emulsifiers were mixed until the mixture was uniform. In a
adjusting the emulsifier and thickener chemistries, as well as the overall secondary mixer, the aqueous phase containing glycols, water and preservatives
ingredient compositions, to provide a new delivery system that maintained were mixed until the phase was homogeneous. The aqueous phase was slowly
the above RPB mechanicaladhesive attributes for the in situ crosslinked introduced to the silicone phase dropwise to form a clear emulsion. Nylon 10-I2
XPL film. was slowly introduced to the main mixer and the emulsion was then homogenized.
After multiple rounds of formulation optimization, a two-step delivery system Finally, the platinum catalyst solution was added to the emulsion and mixed until it
comprising the Step 1 and Step 2 formulations was developed. Water-in-silicone was uniformly dispersed.
emulsion systems were developed to enable the delivery of the lead RPB (at 32%
w/w) of Step 1 and the platinum catalyst (at 200 ppm w/w) of Step 2 through the Mechanical testing. Uniaxial tensile tests were performed with an Instron 3342
continuous silicone phase. With this approach, Step 1 facilitated a consistent (Norwood) based on ASTM D5083 methods. Briefly, a premix of the cure specimen
deposition of the lead RPB coating on skin that was subsequently crosslinked in situ (tetravinyltetramethylcyclotetrasiloxane/Karstedt platinum catalyst solution/RPB
to form the XPL film when exposed to the platinum catalyst in Step 2. The authors or the RPB emulsion; in a ratio of 0.5:0.5:99.0; w/w/w) was transferred into a dog
are not aware of other topical emulsion-based formulations that successfully bone mould (flat sheet type 0.50 wide) and cured for 24 h. The cure specimen was
incorporate such a high concentration of silica-reinforced polysiloxanes. then subjected to either 15 loadunload cycles at 15% tensile strain or 25
The Step 1 formulation containing the lead RPB demonstrated improved loadunload cycles at 100% tensile strain at room temperature to obtain a
spreading properties on lax skin compared with the lead RPB alone. Plots of the stressstrain plot or a hysteresis loop. Then, the cure specimen was subjected to a
instantaneous viscosity as a function of the shear rate illustrated a stronger full tensile load until break where the tensile modulus and the fracture strain were
shear-thinning effect of the Step 1 formulation containing the lead RPB, compared evaluated from the stressstrain response.
with the lead RPB alone, with a two- to threefold decrease in the instantaneous Adhesion-in-peel tests were conducted using an Instron 3342 based on ASTM
viscosity at shear rates above 5 s1 (Supplementary Fig. 8). Note that maximum C794 methods. Briefly, 0.75 g of a cure specimen premix
shear rates from 104 to 105 s1 have been reported for the application of topical (tetravinyltetramethylcyclotetrasiloxane/Karstedt platinum catalyst solution/RPB
cream products on skin28 . The rheological behaviour of Step 1 allows the emulsion or the RPB emulsion; in a ratio of 0.5:0.5:99.0; w/w/w) was sandwiched in between
to exhibit lower viscosities at the high shear rates associated with topical two leather strips with a contact surface area of 1 inch by 2.5 inches and cured for
formulation application, despite the comparable viscosities of 1,500 Pa s1 (Step 1 24 h. The peel test procedure subjected one side of the cured leather strip to a
containing the lead RPB) and 1,800 Pa s1 (the lead RPB alone) measured at low 10 mm s1 extension until the leather strips completely separated. The adhesive
shear rate of 0.15 s1 . In addition, the high formulation viscosities measured at low force was averaged over the entire length of the cured location and normalized to
shear rates support prolonged shelf-life stability by lowering the rates of the width of the sample.
gravity-driven sedimentation. Rheometer tests developed for measuring the crosslinking kinetics were
Finally, the XPL optical properties were adjusted to complement the light conducted using the Bohlin CVO Rheometer (Malvern) to implement a dynamic
scattering properties of natural skin. The XPL optics were modulated by mechanical test based on ASTM D7750 standards. Briefly, a premix of the cure
introducing light scattering particles into the Step 2 formulation such that these specimen (tetravinyltetramethylcyclotetrasiloxane/Karstedt platinum catalyst
particles were deposited at the XPL surface. Surface-treated nylon particles with a solution/RPB or the RPB emulsion; in a ratio of 0.67:0.33:99.0; w/w/w) was
particle diameter approximating 8 m and a refractive index of 1.54 (Nylon 10-I2, transferred to a 25-mm flat plate with a 250-m gap size on the rheometer. The
KOBO Products) provided the best optical match between the XPL and natural experiment was conducted at 70 C under the oscillatory mode with a 1-Hz
skin, following repeated clinical evaluation of several commercially available light frequency to obtain a plot of the storage modulus (G0 ) over time.
scattering particles. Clinical examples of the optical effects are shown in Figs 1b, 2,
3 and 6. Pilot human studies. Normal volar forearm skin recoil study (Study B). The study
protocol was approved by the Allendale IRB. Briefly, 6 women, ages 40 and older,
Materials. Vinyl dimethicone, hydrogen dimethicone, and Karstedt platinum were enrolled in the study. Duplicate volar forearm skin sites of 3 cm by 3 cm in
catalyst were purchased from AB Specialty Silicones. PMX-1184 (dimethicone and area per panelist were assigned to three different treatments, XPL, AmLactin
trisiloxane), FZ-2233 (bis-isobutyl PEG/PPG-10/7/dimethicone copolymer), and Moisturizing Body Cream (Upsher Smith Laboratories), and an untreated control
DC-9045 (decamethylcyclopentasiloxane and dimethicone crosspolymer) were site, following a predetermined randomized scheme. Measurements were taken at
purchased from Dow Corning Corporation. KSG-240 (dimethicone/PEG-10/15 baseline, 24 h after test article application and 35 min following test article removal.
crosspolymer and decamethylcyclopentasiloxane) and KF-995 A 0.08 ml volume of the lead RPB emulsion (Step 1) and 0.12 ml of the platinum
(decamethylcyclopentasiloxane) were purchased from Shin-Etsu Chemical catalyst delivery system (Step 2) were applied to each designated site to create the
Corporation. Pemulen TR-2 (acrylates/C10-30 alkyl acrylate crosspolymer), XPL. AmLactin (0.05 ml) was applied twice daily to the corresponding sites. The
Ultrez 20 (acrylates/C10-30 alkyl acrylate crosspolymer) and waterborne remover (1.5 ml) was applied to each of the eight sites evaluated after the 24-h
polyurethane dispersion Avalure UR 450 polymer (PPG-17/IPDI/DMPA measurement and left in contact with the XPL for at least 30 s to saturate and swell
copolymer) were purchased from Lubrizol Corporation. Jeecide CAP-4 the film. The swollen film was then rolled off the skin using a cotton pad. Residual
(phenoxyethanol and caprylyl glycol) and Jeechem BUGL (butylene glycol) were test article was removed with a dry tissue.
purchased from JEEN International. Glycerin and propylene glycol were purchased All suction cup measurements were taken following a 2530 min acclimation
from Ruger Chemical. Sodium chloride and sodium hydroxide (10N) were period in a controlled environment with the relative humidity maintained at less
purchased from Spectrum Chemical MFG. Nylon 10-I2 (Nylon 12 and isopropyl than 50% and temperature maintained at 1922 C. A DermaLab USB (Cortex
titanium triisostearate) was purchased from Kobo Products. Technology) with a suction cup was used to evaluate skin elasticity following
The lead RPB emulsion or Step 1 comprised 32% RPB 6 (see Supplementary methods described in the literature32,33 . Statistical analyses were conducted using a
Table 1), 12% PMX-1184, 4% KSG-240, 0.1% FZ-2233, 50.32% water, 0.5% paired Students t-test.
Pemulen TR-2, 0.5% Ultrez 20, 0.33% Jeecide CAP-4, and 0.25% 10 N sodium Skin conductance was measured in 24 volunteers (aged 25 to 55 years old)
hydroxide solution. using a skin surface hygrometer (USB Dermalab with a flat-faced hydration probe,
The platinum catalyst delivery system or Step 2 comprised 10% DC9045, 4% Cortex Technology) for two test articles: petrolatum (Vaseline, Unilever) and XPL.
KSG-240, 0.2% FZ-2233, 15.79% decamethylcyclopentasiloxane, 4.5% Nylon 10-I2, To enable probe contact with the skin during XPL wear, an annulus approximating
29.50% water, 0.5% sodium chloride, 0.5% Jeecide CAP-4, 20% propylene glycol, the probe tip dimensions (1.4 cm2 area) was created at the centre of the 5 cm by
10% butylene glycol, 4% glycerin and 1.01% Karstedt platinum catalyst solution 5 cm application site, where bare skin remained readily accessible to take the skin
(2% platinum). The polyurethane and silicone wound dressing sheets were conductance measurements. All measurements were taken following a 2530 min
purchased from 3M (Tegaderm Transparent Film Dressing 1626W) and acclimation period in a controlled environment with the relative humidity

NATURE MATERIALS | www.nature.com/naturematerials

2016 Macmillan Publishers Limited. All rights reserved

ARTICLES NATURE MATERIALS DOI: 10.1038/NMAT4635

maintained at less than 50% and temperature maintained at 1922 C. visualized by the smooth facial contour observed from the lower lid to the cheek.
Measurements were taken at baseline and following test article application A score of 4 described severe protrusion of the lower lid fat pad, where the lower lid
at 1 h, 2 h, 4 h and 6 h. Statistical analyses were conducted using Tukeys fat pad extended beyond the midpoint of the distance defined by the lower lid to
post hoc test. the cheek. For severe bags, the smooth facial contour is disrupted by the concavity
of the boundary highlighting the bag. Twelve women were enrolled in the study.
Dry skin study evaluating TEWL (Study C). The study protocol was approved by the Briefly, the XPL and a randomized placebo control were applied split face to the
Allendale IRB. A 9-point dryness grading scale was used to characterize the extent lower lid application sites following a predetermined randomization scheme. The
of skin dryness, where a score of 0 refers to no dryness and a score of 8 refers to placebo comprised the two steps of the XPL, or the vehicle, where the catalyst for
severe dryness with marked roughness. Healthy female volunteers, ages 40 and with the crosslinking reaction was absent from the platinum catalyst delivery system
visual dryness scores of 3, corresponding to moderately dry skin, or greater at each (Step 2). A Canfield Vectra 3D imaging system (Canfield Scientific) was used to
test site were included in the study. capture the changes observed in the shape of the skin contours at baseline, 1 h, and
Subjects were instructed to stop the use of all moisturizing products (soaps, 4 h following XPL application. OCT imaging and live trained grader evaluations
lotions, sunscreens, insect repellents and so on) on their legs during a 3-day were also conducted with the evaluator blinded to the treatments at the three
pre-conditioning period before testing. Shaving was allowed up to 3 days before the time points.
start of the study and after the completion of the 24-h measurements. Exercise or Metallic skin markers were applied to the under-eye area of each subject to
drinking of hot or caffeinated beverages within 2 h before each instrument visit was enable consistent placement of the OCT probe at the same skin sites for each time
discouraged as this could affect the measurements. Intensive, heavy exercise point. For each under-eye area, three sites were consistently imaged at each of the
promoting excessive sweating was not permitted. The subjects were also instructed three time points. These markers are identified by the arrows in the corresponding
to wear shorts or pants that could be rolled above the knees to each visit. OCT images (Fig. 6). These OCT images were analysed to quantitatively assess the
Exclusion criteria included women with a skin condition other than dry skin at skin surface shaping effect of the XPL, using ImageJ (National Institutes of Health).
the test sites (such as psoriasis, eczema and atopic dermatitis); subjects with marks The distance between the two epidermal markers was measured using a straight
(such as tattoos, and scars) at the test sites that might interfere with the visual line (SLD). The epidermal surface length (ESL) was measured by tracing along the
grading, were also excluded from the study. Twenty-two subjects completed the uneven epidermal surface in between the two markers. The surface irregularity
entire study. values, described by the difference between the ESL and SLD (normalized by the
For each subject, two of the four test products were applied in duplicate to four ESL), were calculated at baseline, one hour, and four hours after XPL application.
of the five sites and the remaining site served as the non-treated control site. Test Values of surface irregularity that approach zero represent the ideal case of no
articles assigned to each site of the five outer calf test sites (three sites on one leg surface roughness, as the ESL approaches the SLD.
and two sites on the other leg) were based on a predetermined rotation scheme. The three sets of OCT image sections recorded for each skin site at
A total of 0.15 ml of each Step 1 and Step 2 formulation was applied to cover the baseline and at 4 h for the vehicle and for the XPL were further analysed to quantify
entire application site. Approximately 0.05 ml of petrolatum (Vaseline, Unilever) the changes in the distribution of the pixel intensity. Sections of each image,
and 0.05 ml of the commercial product (Crme De La Mer, Estee Lauder comprising approximately 400 pixels in width and 40 pixels in thickness,
Companies) was applied to the corresponding designated test sites using a finger immediately below the region identified as the stratum corneum, were selected by a
cot according to the randomization schedule. Expert graders were blinded to the blinded analyst. Owing to area constraints placed by the sample geometries for a
application process. limited number of images, the area analysed was reduced to remain contained
Before baseline acclimation, subjects test areas were cleansed with wet within the epidermal layer. These image sections were then processed with ImageJ
Kimwipes (Kimberly-Clark Corporation) and then patted dry with dry Kimwipes. to calculate the average pixel intensity for each section selected. The ImageJ
The cleansing was as minimal as wiping twice with wet Kimwipes and subsequently intensity scale is a 065,535 scale, corresponding to the dark and the lucent
patting dry by stroking twice with dry Kimwipes. Five 5 cm by 5 cm skin test sites extremes of pixel intensity, respectively. After the average intensity was calculated
were then marked. Following the established methods described above, for each site, the samples were un-blinded, and Students t-test was implemented to
transepidermal water loss (TEWL) measurements were taken at baseline, two hours determine the significance of the differences detected between the XPL and the
and twenty-four hours following test product application using a CyberDERM RG1 vehicle groups.
Evaporimeter (CyberDERM). Subjects reported for their 24-h visit and the
investigator determined the film integrity of the XPL-treated sites. If the XPL was Photography. Time-lapse photographs were taken using three digital single lens
not minimally an 80% continuous film, the subjects participation in the study reflex cameras (two Nikon D80 and one Nikon D300), equipped with 1770 mm
would end and day 2 measurements would not be taken. For the TEWL values, a macro lenses and circular polarizers, positioned to simultaneously photograph the
one-way ANOVA was used to compare the net change from baseline at each time subject at three angles: 45 angles from the left and the right, and one at the centre
point. For all analyses, a two-tailed p < 0.05 was taken as the level of significance. position. The subject was lit using three 1-ft2 -LED studio panel lights, each
providing 1,400 LUX of 5,600 K lighting. Additional photographs were taken using
Double-blind, randomized, placebo-controlled, split-face evaluation of XPL a digital single lens reflex camera (Canon 5D Mark II) with an 85 mm lens (Sigma)
performance (Study D). Optical coherence tomography (OCT, Michelson Vivosight and a strobe light unit (Alien Bees B800, Paul C. Buff) with an umbrella (Westcott)
OCT, Michelson Diagnostics) was used for non-invasive, in vivo, real-time skin angled at approximately 45 towards the subject.
microscopy imaging to elucidate the skin hydration and topical mechanical shaping
mechanism of the XPL technology. The clinical study was reviewed and approved References
by the Western Institutional Review Board (WIRB protocol number 20121506). 32. Pedersen, L. & Jemec, G. B. Mechanical properties and barrier function of
Subjects between the ages of 4075 with baseline lower lid severity scores of 3 or 4, healthy human skin. Acta Derm. Venereol. 86, 308311 (2006).
based on a 04 photonumeric lower lid severity scale, were the primary inclusion 33. Gniadecka, M. & Serup, J. in Handbook of Noninvasive Methods and the Skin
criteria. A score of 0 corresponded to the absence of lower lid fat herniation, as (eds Serup, J. & Jemec, G. B. E.) 329335 (CRC Press, 2006).

NATURE MATERIALS | www.nature.com/naturematerials

2016 Macmillan Publishers Limited. All rights reserved

 SUPPLEMENTARY INFORMATION
DOI: 10.1038/NMAT4635

An elastic second skin

An Elastic Second Skin

Betty Yu1, Soo-Young Kang1, Ariya Akthakul2, Nithin Ramadurai2, Morgan Pilkenton1, Alpesh
Patel1, Amir Nashat2, Daniel G. Anderson3,4,5,6, Fernanda H. Sakamoto7, Barbara A. Gilchrest8,
R. Rox Anderson7 and Robert Langer3,4,5,6*
1 Living Proof, Inc., Cambridge, MA 02142, USA
2 Olivo Laboratories, LLC., Cambridge, MA 02142 USA
3 Institute for Medical Engineering and Science, Massachusetts Institute of Technology,
Cambridge, MA, 02139, USA
4 Harvard-MIT Division of Health Science and Technology, Massachusetts Institute of Technology,
Cambridge, MA, 02139, USA
5 David H. Koch Institute for Integrative Cancer Research, Massachusetts Institute of Technology,
Cambridge, MA, 02139, USA
6 Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA
02139, USA
7 The Wellman Center for Photomedicine, Massachusetts General Hospital and Harvard Medical
School, Boston, MA 02114, USA
8 Department of Dermatology, Massachusetts General Hospital and Harvard Medical School,
Boston, MA 02114, USA

*To whom correspondence should be addressed.

Supplementary Information

Movie file: Figure3_ElasticityDemonstration.mov

NATURE MATERIALS | www.nature.com/naturematerials 1

2016 Macmillan Publishers Limited. All rights reserved.
SUPPLEMENTARY INFORMATION DOI: 10.1038/NMAT4635

3.00
RPB series A
2.50
RPB series B
Tensile Modulus (MPa)

2.00

1.50

1.00

0.50

0.00
0% 100% 200% 300% 400% 500% 600% 700% 800% 900%
Fracture Strain (%)
Figure 7| Broad utility of XPL materials platform space is depicted by the range of tensile
moduli and the corresponding fracture strains achieved by modifying three materials design
parameters: (1) the cross-linking density, (2) the molecular weight of functional polysiloxanes
(vinyls and hydrides), and (3) the reinforcement particle loading. Subsets of lead RPB
compositions are highlighted by the red diamonds and the open diamonds. The detailed
mechanical properties of these compositions are described in Table 1.

2 NATURE MATERIALS | www.nature.com/naturematerials

2016 Macmillan Publishers Limited. All rights reserved.
DOI: 10.1038/NMAT4635
SUPPLEMENTARY INFORMATION
10000
RPB
Step1 formulation containing RPB

1000
Viscosity (Pa.s)

100

10

1
0.1 1.0 10.0 100.0
Shear rate (1/s)

Figure 8| Shear thinning behavior of the Step 1 formulation containing RPB (open circle) versus
that of the RPB alone (filled square).

NATURE MATERIALS | www.nature.com/naturematerials 3

2016 Macmillan Publishers Limited. All rights reserved.
SUPPLEMENTARY INFORMATION DOI: 10.1038/NMAT4635

Figure 9| Elasticity of XPL in comparison with a commercial benchmark. Side-by-side

comparison of stress-strain hysteresis loops evaluating at 15% deformation over 15 load-unload
cycles for the XPL (left figure) and a film formed by a commercial polyurethane Avalure UR450
dispersion (right figure). The hysteresis loop at the 15th cycle that drifts away from the hysteresis
loop at the 1st cycle in the polyurethane UR450, in comparison to the XPL, indicated plastic
deformation behaviour at 15% strain.

4 NATURE MATERIALS | www.nature.com/naturematerials

2016 Macmillan Publishers Limited. All rights reserved.
DOI: 10.1038/NMAT4635
SUPPLEMENTARY INFORMATION
Reactive Plastic
Hysteresis
Polymer Fumed Leather Fracture Tensile strain after
Polysiloxane loss energy
Blend (RPB) Silica Adhesion Strain Modulus tensile
composition density
composition (wt%) (N/mm) (%) (MPa) hysteresis
(kJ/m3)
No. (%)
4.71 57 0.43 0 0.035
1 A 0%
(5.89) (28) (0.03) (0) (0.007)
15.2 197 0.64 0.18 0.090
2 A 15%
(3.69) (14) (0.02) (0.15) (0.017)
58.0 318 0.96 0.45 0.443
3 A 27%
(29.7) (16) (0.02) (0.01) (0.012)
11.0 157 0.25 0.43 0.105
4 B 0%
(1.53) (11) (0.02) (0.01) (0.007)
70.0 286 0.31 0.76 0.240
5 B 15%
(12.1) (128) (0.01) (0.06) (0.017)
78.0 826 0.48 1.3 0.657
6 B 27%
(2.74) (120) (0.03) (0.26) (0.017)

Table 1| RPB compositions and their associated material properties as a function of increasing
fumed silica loading for two different polysiloxane compositions. Polysiloxane composition A
contains 12.66% VS165K, 63.28% VS10K, and 24.06% XL-11. Polysiloxane composition B contains
13.57% VS165K, 58.04% VS10K, and 28.39% XL-11. Values are reported as the average of three
samples. The standard deviations of the measurements are noted in parentheses. Six RPB
compositions arising from two different lead polysiloxane compositions, RPB series A and B, each
loaded at three different fumed silica concentrations (0%, 15%, and 27%) are highlighted by the
red and the open diamonds in Figure 7. RPB (series B at 27% fumed silica concentration) is the
lead composition evaluated. Polysiloxane composition A comprises a higher percentage of the
short chained vinyl-terminated polysiloxane at 63.28% (versus 58.04% in RPB series B) and a
lower cross-linker percentage at 24.06% (versus 28.30% for RPB series B). Close to a two-fold
increase in modulus was measured for the RPB series A compared to RPB series B at each fumed
silica loading (0.43, 0.64, and 0.96 MPa compared to 0.25, 0.31, and 0.48 MPa, for 0%, 15% and
27% fumed silica loading, for RPB series A and for RPB series B, respectively). Moreover,
incremental increases in the fumed silica loading increased the leather adhesion, fracture strain,
and tensile modulus within each polysiloxane blend composition. For each of these compositions,
the plastic strain after cyclic hysteresis testing remained close to only 1% and the dissipative
(hysteresis loss) energy density calculated from the area delineated by the hysteresis loop was
less than 0.7 kJ/m3. Lead RPB series B (No.6) at high silica loading was selected as the material
basis for the formulation development because of the superior leather adhesive strength (78
N/mm) and elongation properties (fracture strain over 800%), compared to the other RPB
compositions listed.

NATURE MATERIALS | www.nature.com/naturematerials 5

2016 Macmillan Publishers Limited. All rights reserved.
SUPPLEMENTARY INFORMATION DOI: 10.1038/NMAT4635

Time point XPL Vehicle

31477* 30878**
baseline
(3708) (3821)
25888*,# 28774**,#
4 hours
(3207) (3535)

Table 2| Skin hydration measurements based on calculations of average epidermal pixel

intensity of OCT image sections for XPL and vehicle exposed skin sites. The average pixel
intensities for the XPL and the vehicle treated sites are shown for the total of 36 OCT image
sections that were analyzed per treatment group. Dark pixels, or lower intensity values, reflect
regions of greater hydration, based on the OCT imaging technique. The differences measured for
the XPL group and for the vehicle group at four hours compared to their baseline intensities were
significant (*p<0.00001, **p<0.05, respectively), translating to average pixel intensity decreases
of 18% and 7%. The XPL sections showed a significantly lower average pixel intensity compared
to the vehicle (#p<0.001). The standard deviation of the mean is reported in parentheses.
Statistical significance determined using paired Students t-tests.

6 NATURE MATERIALS | www.nature.com/naturematerials

2016 Macmillan Publishers Limited. All rights reserved.