Articles

https://doi.org/10.1038/s41893-018-0023-2

Nanoporous polyethylene microfibres for

large-scale radiative cooling fabric
Yucan Peng 1,5, Jun Chen 1,5, Alex Y. Song 2,5, Peter B. Catrysse2,5, Po-Chun Hsu3, Lili Cai1, Bofei Liu1,
Yangying Zhu1, Guangmin Zhou1, David S. Wu1, Hye Ryoung Lee1, Shanhui Fan2 and Yi Cui1,4*

Global warming and energy crises severely limit the ability of human civilization to develop along a sustainable path. Increasing
renewable energy sources and decreasing energy consumption are fundamental steps to achieve sustainability. Technological
innovations that allow energy-saving behaviour can support sustainable development pathways. Energy-saving fabrics with a
superior cooling effect and satisfactory wearability properties provide a novel way of saving the energy used by indoor cool-
ing systems. Here, we report the large-scale extrusion of uniform and continuous nanoporous polyethylene (nanoPE) microfi-
bres with cotton-like softness for industrial fabric production. The nanopores embedded in the fibre effectively scatter visible
light to make it opaque without compromising the mid-infrared transparency. Moreover, using industrial machines, the nanoPE
microfibres are utilized to mass produce fabrics. Compared with commercial cotton fabric of the same thickness, the nanoPE
fabric exhibits a great cooling power, lowering the human skin temperature by 2.3 °C, which corresponds to a greater than 20%
saving on indoor cooling energy. Besides the superior cooling effect, the nanoPE fabric also displays impressive wearability and
durability. As a result, nanoPE microfibres represent basic building blocks to revolutionize fabrics for human body cooling and
pave an innovative way to sustainable energy.

N
owadays, problems associated with energy crises and climate quadratically proportional to the pore diameter15, the large pores
change are becoming increasingly serious stumbling blocks resulting from the knitted/woven configuration would dominate
for human sustainable development1–4. Fifteen per cent of the air permeability, rendering it as air-permeable as a cotton fab-
electricity consumption and ten per cent of greenhouse gas emis- ric. Meanwhile, the large pores can also benefit convective cooling,
sions are ascribed to cooling systems globally5. Therefore, devel- because heat exchange via air convection can easily happen between
oping new technologies to reduce the energy demand for spacing the skin and the ambient environment. Second, the knitted/woven
heating and cooling is significant for energy saving. Notably, an structure enables the fabric to be more stretchable. Fibres with inter-
increase in the cooling set-point of 2 °C can save over 20% of energy laced structure can undergo much more strain than a straight one
for indoor temperature regulation6. of the same material16–18. Third, knitted/woven fabrics tend to show
Without wasting excess energy on the entire building7–9, personal better mechanical strength. When force is applied to a fabric, the
thermal management emphasizing heating and cooling of only the ordered interlaced structure can effectively transfer the applied load
human body and its local environment is more energy-efficient and around with uniform distribution on the fabric (Supplementary
cost-effective10–12. In a typical indoor environment, the human body Fig. 3). Besides, the knitted/woven fabric can greatly decrease the
with a normal skin temperature of about 33.5 °C emits mid-infrared electrification and irreversible wrinkling compared to the nanoPE
thermal radiation in the wavelength range of 7 to 14 µ​m, account- thin film (Supplementary Videos 1 and 2), which is also very critical
ing for over 50% of the total body heat loss13 (see Supplementary toward comfortable wearing.
Note 1 and Supplementary Figs. 1 and 2 for more details). Recently, Accordingly, to realize practical radiative human body cooling
we demonstrated effective radiative human body cooling using a and renovate the human cloth toward a sustainable civilization,
nanoporous polyethylene (nanoPE) non-woven thin film due to developing a new process for massive production of nanoPE micro-
its superior mid-infrared transparency and high opacity to visible fibres as the building blocks for large-scale fabric manufacture is a
light11. However, polymer thin film is by nature unsuitable for normal must. Although polyethylene is currently obtained from fossil fuels,
clothing, because it lacks essential wearable properties, such as air further innovation of plastic production will amplify the sustain-
permeability and touch comfort. For instance, the polyethylene thin ability potential of such cooling fabric. Moreover, nanoPE fabric
film is very electronegative while human skin is electropositive; thus, can be easily recycled by melting and re-extrusion for new clothing
a strong contact electrification would be induced during wearing14, manufacture.
which would result in great discomfort. These concerns can be well
resolved by developing knitted/woven fabric using nanoPE microfi- NanoPE microfibre
bres with cotton-like softness as the building blocks. To bring the real impact into human society, large-scale pro-
The knitted/woven structure in fabric holds several advantages duction of high-quality nanoPE microfibres is the cornerstone
for wearing. First, it contains hierarchical pores, ranging from hun- and first step. Normal polyethylene (PE) fibres, such as gel-spun
dreds of micrometres to submicrometres. As air permeability is ultrahigh-molecular-weight PE fibre and melt-spun high-density

Department of Materials Science and Engineering, Stanford University, Stanford, CA, USA. 2E. L. Ginzton Laboratory, Department of Electrical Engineering,
1

Stanford University, Stanford, CA, USA. 3Department of Mechanical Engineering, Stanford University, Stanford, CA, USA. 4Stanford Institute for Materials
and Energy Sciences, SLAC National Accelerator Laboratory, Menlo Park, CA, USA. 5These authors contributed equally: Yucan Peng, Jun Chen, Alex Y. Song
and Peter B. Catrysse. *e-mail: [email protected]

Nature Sustainability | VOL 1 | FEBRUARY 2018 | 105–112 | www.nature.com/natsustain 105

© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Articles Nature Sustainability

polyethylene (HDPE) fibre, are widely used as ropes, cables and with a smooth surface. As for nanoPE microfibres, the pore size and
ballistic-resistant composites due to their high mechanical strength porosity increase as the oil ratio increases. When the paraffin oil
and chemical resistance19–21. However, normal PE fibres cannot be ratio is low, the droplets of oil can distribute in the solid PE matrix
used for next-to-skin fabric due to their mechanical stiffness and as tiny separate domains. As the paraffin oil ratio increases, tiny
visible transparency. To obtain a cooling fabric with satisfactory droplets of paraffin oil tend to distribute as contiguous nanoscale oil
wearability, here we propose to introduce nanoscale porosities into domains, and eventually form an interconnected matrix within the
PE microfibres via large-scale fabrication techniques. This crucial solid PE system, as demonstrated in Supplementary Figs. 7 and 8.
innovation generates two functions simultaneously: an optical func- The ratio of PE and paraffin oil is also critical for the mechani-
tion, where the nanoscale pores (100 nm to 1,000 nm) scatter visible cal strength and softness of the nanoPE microfibres. As shown in
light (400–700 nm in wavelength) strongly while permitting mid- Supplementary Fig. 9, the tensile strength of the nanoPE microfi-
infrared light (7 to 14 µ​m) to transmit, creating the necessary opti- bres depends on the ratio of paraffin oil to PE. Generally, with a
cal effect of visible opacity and mid-infrared transparency; and a higher ratio of paraffin oil to PE, the nanopore volume fraction in
mechanical function, where the nanoscale pores reduce the mass the nanoPE microfibres increases. This would greatly improve the
occupation of PE inside the microfibres and create a lot of localized fibre softness, but reduce its mechanical strength. Conversely, with
empty space to accommodate the bending motion of the microfi- a lower ratio of paraffin oil to PE, fewer nanoscale cavities would be
bres. The mechanical hardness of solid microfibres can be reduced formed in the obtained microfibre, which gives it a higher mechani-
significantly when modified into nanoporous microfibres, making cal strength. However, the low porosity would lead to visible trans-
PE materials soft to wear. lucence and large stiffness, which can greatly reduce the feasibility
However, no nanoPE fibres for fabric have been reported to date. and comfort for wearing. Namely, a trade-off exists between the
PE is a very chemically inert material and is not soluble in most sol- fibre softness and mechanical strength. Experimentally, a ratio of
vents. To create nanoPE, we invent a new process to manufacture the 1:3.5 (PE weight/oil volume) is tested as the optimized recipe to
nanoPE fibres. Paraffin oil is selected as the solvent for PE because deliver a satisfactory mechanical strength together with cotton-like
it has chemical properties similar to PE, which consists of nonpo- softness.
lar, saturated, high-molecular-weight hydrocarbons22. The proper-
ties in common make them compatible with each other at a suitable Industrial-scale textile fabrication
temperature (180 °C) and they form a highly viscous homogeneous Tens of pounds of nanoPE microfibres can be rapidly and con-
solution, which can be continuously extruded by an industrial tinuously produced via the fibre extrusion process. Consequently,
extruder machine to form fibres with microscale diameters. Once nanoPE fabrics are efficiently realized via industrial knitting/weav-
extruded, the oily PE fibre cools down fast, and then solid-phase ing equipment. As illustrated in Fig. 2a, the nanoPE fabric not only
PE is separated out from the liquid-phase paraffin oil. The phase maintains high mid-infrared transmittance and visible opacity,
separation process creates nanoscale domains of oil intermixed but also greatly improves the wearability. The achievement mainly
together with PE. Subsequently, methylene chloride, which can originates from two aspects: fabrication of nanoPE microfibres and
selectively resolve paraffin oil, is used to extract the oil phase from adoption of a knitted/woven structure. Figure 2b exhibits a piece
the oily PE fibres. The removal of the oil phase leaves a nanoporous of woven nanoPE fabric with a dimension of 1.8 m by 0.5 m. An
space embedded into the PE solid matrix, generating nanoPE micro- enlarged view of the fabric, to show its woven structure, is dem-
fibres. The whole procedure is illustrated in Fig. 1a. Supplementary onstrated in Fig. 2c. Besides woven fabric, the nanoPE microfibres
Fig. 4 shows a photograph of the extruder producing nanoPE are also demonstrated to construct a fabric via knitting (Fig. 2d). In
microfibres, which can be continuously produced with satisfactory such knitted nanoPE fabric, the nanoPE fibre loops are interlocked
uniformity and mechanical strength at a speed of about 50 metres with each other, forming an ordered knitted structure (Fig. 2e).
per minute (Supplementary Video 3). As detailed in the Methods, In contrast, the appearance of the nanoPE thin film is similar
the production of the nanoPE microfibres is straightforward and to a piece of plastic film with randomly distributed disordered
compatible with large-scale manufacturing. nanoscale branches, as the sketch, photograph and SEM image show
Spools of nanoPE microfibres are mass produced by large-scale in Supplementary Fig. 10.
extrusion not only in the laboratory (Supplementary Fig. 5), but
also in industry (Fig. 1b). The inset of Fig. 1b shows an array of Optical and thermal properties
nanoPE microfibres in milky white colour with cotton-like soft- The visible opacity of the nanoPE fabric, using the knitted one as
ness (Supplementary Video 4). Figure 1c compares the appearance an example, is measured with an ultraviolet–visible spectrometer.
of nanoPE fibres and normal PE fibres in the same image using Opacity is defined as (1 −​ specular transmittance). As demonstrated
an optical microscope. The nanopores in nanoPE fibres can scat- in Fig. 3a, the nanoPE fabric shows visible opacity of nearly 90%.
ter visible light strongly and render the nanoPE fibre milky white Although this number is a little lower than that of the nanoPE thin
in colour, whereas the normal PE fibre looks visibly transparent. film, cotton fabric and Tyvek, it is reasonably good for wearing. In
Scanning electron microscope (SEM) images of the surface (Fig. addition, as shown in the inset, ascribed to the light scattering by
1d) and a cross-section view (Fig. 1e) of a nanoPE microfibre show the nanopores, the nanoPE fabric is much more visibly opaque than
that the nanoscale cavities are contiguous with each other from the fabric made of normal PE fibres. However, the latter still displays
the inner surface to the outer fibre surface. The inset of Fig. 1e is some opacity, because the interlaced structure contributes to addi-
a lower-magnification SEM image showing the well-preserved tional opacity compared to a flat normal PE film. According to the
cross-section of the microfibre. The sizes of these nanoscale cavi- simulation via Mie theory (Supplementary Fig. 11), the opacity of
ties (100 nm–1,000 nm) are too small to scatter mid-infrared light knitted fabric can reach up to 54% even for those made of transpar-
(7–14 µ​m), but are perfect for scattering visible light (400–700 nm) ent normal PE fibres, indicating the extra contribution of the knit-
to realize the opacity. Moreover, the nanoscale pores embedded in ted structure to visible opacity.
the microfibre can easily alter their shapes to adapt to the bending The total infrared light transmittance is investigated with a
of the microfibre, as illustrated in Fig. 1f. This is the reason why the Fourier transform infrared (FTIR) spectrometer equipped with a
nanoPE microfibre has cotton-like softness, whereas the normal PE diffuse gold integrating sphere, to characterize mid-infrared trans-
microfibre is much more rigid (Supplementary Fig. 6). parency of the nanoPE fabric. The experimentally measured mid-
To control the pore size and fibre porosity, the ratio of PE and infrared transmittance is shown in Fig. 3b. The nanoPE thin film
paraffin oil plays a key role. Normal PE microfibre is transparent exhibits the highest infrared transmittance owing to it having the

106 Nature Sustainability | VOL 1 | FEBRUARY 2018 | 105–112 | www.nature.com/natsustain

© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Nature Sustainability Articles
a b c
HDPE and paraffin oil
Mould

Screw rod
Heating cover Blow head

Oily PE
Nanoporous Oil
extraction Pulley
PE microfibre

d e f

Bending
Nanopores

NanoPE
microfibre

Fig. 1 | NanoPE microfibre. a, A schematic diagram of the manufacturing process for the nanoPE microfibre. b, A photograph of three spools of nanoPE
microfibres produced by industrial extrusion. Scale bar, 9 cm. The inset shows that the fibres are milky white with cotton-like softness. Scale bar, 1.5 cm.
c, An optical micrograph of a nanoPE microfibre (left) and a normal PE fibre (right). The white colour of the nanoPE microfibre is mainly attributed to the
strong visible light scattering by the embedded nanoscale cavities, while normal PE looks transparent under visible light. Scale bar, 200 μ​m. d, SEM image
of the surface of a nanoPE microfibre. Scale bar, 2 μ​m. e, SEM image of the cross-section view of a nanoPE microfibre. Scale bar, 2 μ​m. The inset shows a
lower-magnification SEM image of the well-preserved cross-section of the microfibre. Scale bar, 50 μ​m. f, An illustration showing the nanoscale porosities
adapting to a bent deformation, thereby improving the softness for wearing.

a b
Mid-infrared light

Visible light

Water vapour Air convection

c d e

Fig. 2 | NanoPE fabrics. a, A schematic diagram of the nanoPE fabric. The nanoPE fabric exhibits high mid-infrared transparency, visible opacity and
good wearability. b, A photograph of a large woven nanoPE fabric. Scale bar, 0.35 m. c, An enlarged view of the woven nanoPE fabric showing the woven
structure. Scale bar, 200 μ​m. d, A photograph of a knitted nanoPE fabric. Scale bar, 3 cm. e, An enlarged view of the knitted nanoPE fabric showing
the knitted structure. Scale bar, 1.5 mm.

smallest thickness of only 12 μ​m. The nanoPE fabric with a thickness non-woven PE textile made by Dupont, shows a little lower mid-
of 450 μ​m, about 37 times thicker than the nanoPE thin film, still infrared transparency than that of the nanoPE fabric. Notably, for
delivers an infrared transmittance over 70%. Tyvek, a 178-μ​m-thick a common cotton fabric with a similar thickness to the nanoPE

Nature Sustainability | VOL 1 | FEBRUARY 2018 | 105–112 | www.nature.com/natsustain 107

© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Articles Nature Sustainability

a NanoPE film (12 µm) Cotton (457 µm) b

Tyvek (178 µm) NanoPE fabric (450 µm) NanoPE film (12 µm) NanoPE fabric (450 µm)
Normal PE fabric (410 µm) Tyvek (178 µm) Cotton (457 µm)

100 100 12 µm
450 µm
450 µm
80 80

Transmittance (%)
410 µm
178 µm
Opacity (%)

60 60

40
40

20
20
No textile Normal PE NanoPE 457 µm
fabric fabric 0
0
400 500 600 700 800 4 6 8 10 12 14 16 18
Wavelength (µm) Wavelength (µm)

c 225 µm 450 µm 675 µm 900 µm d 225 µm 450 µm 675 µm 900 µm

1,125 µm 1,350 µm 1,575 µm 1,125 µm 1,350 µm 1,575 µm
100 100

80
Transmittance (%) 80
Transmittance (%)

60 60

40 40

20 20
NanoPE
NanoPE fabric thin film
0 0
4 6 8 10 12 14 16 18 4 6 8 10 12 14 16 18
Wavelength (µm) Wavelength (µm)

Fig. 3 | Optical properties of the nanoPE fabric and other textiles. a, Visible opacity measurement of nanoPE fabric (thickness: 450 μ​m), normal PE fabric
(thickness: 410 μ​m), nanoPE thin film, Tyvek and cotton. The inset shows that the nanoPE fabric has a better opacity than the normal PE fabric.
b, Measured total FTIR transmittance of nanoPE fabric, nanoPE thin film, Tyvek and cotton, with a thickness, respectively, of 450 μ​m, 12 μ​m, 178 μ​m
and 457 μ​m. c,d, The dependence of the mid-infrared transmittance on the textile thickness using rigorous coupled-wave analysis for nanoPE fabrics (c)
and nanoPE thin films (d).

fabric, almost no infrared transmittance is observed due to its infrared camera receives the least amount of infrared radiation.
strong infrared absorption. Furthermore, the nanoPE fabric exhib- Therefore, the thermal imaging of cotton appears the coldest among
its better infrared transmittance than the nanoPE thin film of an the samples. The nanoPE thin film and the nanoPE fabric allow most
identical thickness, which can be justified by the infrared trans- of the infrared radiation to transmit, which can greatly improve the
mittance simulation of the nanoPE fabric and nanoPE thin film heat loss from the human body. Moreover, the cooling effect of the
with varied thickness using rigorous coupled-wave analysis23. The nanoPE fabric is experimentally demonstrated with an artificial skin
dependence of mid-infrared transmittance on the textile thick- temperature measurement (Supplementary Fig. 12). With a similar
ness is respectively shown in Fig. 3c for the nanoPE fabric and thickness to the cotton fabric, the nanoPE fabric merely increases
Fig. 3d for the nanoPE thin film. As the thickness increases, the the skin temperature by 1.3 °C, while the cotton fabric increases the
mid-infrared transmittance of the nanoPE fabric and that of the skin temperature by 3.6 °C (Fig. 4b). The set-point increase of indoor
nanoPE thin film both decreases, but the transmittance of the temperature corresponds to the decrease of simulated skin tempera-
nanoPE fabric decreases less rapidly than that of the nanoPE thin ture (see Supplementary Note 2 for a more detailed explanation).
film. With a thickness of 900 μ​m, the infrared transmittance of the This difference indicates that people who wear the nanoPE fabric can
nanoPE thin film drops to less than 30%, while the nanoPE fabric increase the set-point by 2.3 °C but still feel as thermally comfortable
still shows an infrared transmittance of over 60%. This difference (that is, having the same skin temperature) as the ones who wear cot-
clearly demonstrates the advantage in the cooling capacity of the ton clothes. Despite the much larger thickness, the cooling effect of
nanoPE fabric over the nanoPE thin film. the nanoPE fabric is still comparable to that of the nanoPE thin film.
To visualize the infrared transmission through various textiles, a This is potentially due to the desirable convective cooling effect aris-
thermal camera is employed to detect the infrared radiation trans- ing from the interlaced structure. Perhaps for the same reason, Tyvek
mitting from a simulated human skin. As demonstrated in Fig. 4a, does not show as effective a cooling effect as the nanoPE fabric, even
cotton fabric is non-transparent to infrared radiation, and thus the though they display similar mid-infrared transmittances.

108 Nature Sustainability | VOL 1 | FEBRUARY 2018 | 105–112 | www.nature.com/natsustain

© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Nature Sustainability Articles
a
38 °C

32 °C
NanoPE film NanoPE fabric Tyvek Cotton
b 500
c 50
38
Cotton NanoPE fabric

Cooling energy saving (%)

40
400

Textile thickness (µm)

Skin temperature (°C)

36
30
300
34
20
200

32 10
100
0
30 0
20 21 22 23 24 25
in

n
ric

k
to

ve
sk

fil

ot
fa

Ty
E

Set-point (°C)
re

C
oP

E
Ba

oP
an

an
N

Fig. 4 | Thermal measurement of the nanoPE fabric and other textiles. a, Infrared images of the nanoPE fabric, nanoPE film, Tyvek and cotton on a
simulated human skin. b, Artificial skin temperature measurement of bare skin, the nanoPE fabric, nanoPE thin film, cotton and Tyvek. The 45-µ​m-thick
nanoPE fabric has a comparable cooling effect to the 12-µ​m-thick nanoPE thin film. Both are superior to Tyvek and cotton. c, Cooling energy saving
calculation. As the indoor set-point increases, the cooling energy can be reduced significantly.

The significance of reducing cooling energy consumption via the nanopores greatly facilitate the water vapour transmission. Air
nanoPE fabric could be quantified (see Supplementary Note 3 for permeability, which is a measure of how well a fabric allows the
more details). With an increase of indoor set-point, the input cool- passage of air, is also essential to the performance of wearable
ing energy can be reduced sharply (Fig. 4c). To be more specific, the fabrics. Experimentally, the pressure drop as a function of air
2.3 °C set-point difference between the nanoPE fabric and cotton flow rate of a given fabric area is studied. Cotton, a renowned
can help save about 20.1% of the cooling energy, which is consistent breathable fabric, transmits air with almost zero pressure drop.
with results obtained via applying other models6, 24. This decrease of The nanoPE fabric presents an equal air permeability to cotton.
energy consumption not only means energy saving, but also sug- However, both the nanoPE thin film and Tyvek display poor air
gests CO2 emission reduction, which is significant in global sustain- permeability (Fig. 5b). Furthermore, the water wicking rate is
able development with a wide range of applications. critical for wearable textile, especially for next-to-skin clothes,
The nanoPE fabric is primarily designed for an air-conditioned which would be wetted by body perspiration. A higher water
indoor environment for energy saving. The room temperature is wicking rate can expedite the sweat transportation and evapo-
usually lower than human body skin temperature in order to realize ration from human skin. Even though perspiration is not very
thermal comfort for the human body. The mid-infrared transpar- common in an indoor environment, the water-wicking rate of the
ency of the nanoPE fabric provides an additional radiative cooling nanoPE fabric can be enhanced through facile modification even
route to dissipate the majority of heat, besides convection and vapor- though PE is intrinsically hydrophobic. We chemically altered
ization, compared with conventional textiles such as cotton. This its surface property through short-time treatment with a benign
superior radiative cooling effect can help save a tremendous amount hydrophilic and bio-compatible agent, polydopamine. The polar
of energy on indoor cooling systems. In addition, the nanoPE fabric groups –OH and –NH– make the nanoPE fabric hydrophilic26–28,
may also serve as a potential textile for outdoor cooling. Its superior and additional sparse cotton mesh can further elevate the capil-
mid-infrared transparency can help it to harvest coldness from the lary force for a further improvement of the fabric water-wicking
universe even in a hot environment7, 8,25. rate. The modified nanoPE fabric demonstrates a wicking rate up
to 3 cm per 30 s, approaching that of cotton at around 4 cm per
Wearability properties 30 s (Fig. 5c), In addition, adequate mechanical strength of a tex-
Besides the superior cooling effect, a fabric should also possess tile is necessary. Figure 5d shows that the 1.5-cm-wide nanoPE
suitable wearability properties usually associated with traditional fabric can bear a tensile force of about 15 N, which is satisfactory
clothes, such as high water vapour transmission rate, air perme- for human cloth fabrication and almost twice that of the nanoPE
ability, water wicking rate and mechanical strength. First, people thin film11. Moreover, the ultimate load of the nanoPE fabric is
prefer fabrics with a high water vapour transmission rate, which increased to over 40 N via the addition of a cotton mesh, which
allows the moisture on the skin to evaporate during perspiration is stronger than that of Tyvek and cotton with ultimate loads of
to alleviate discomfort. As shown in Fig. 5a, the nanoPE fabric has approximately 35 N.
the highest water vapour transmission rate of 0.023 g cm−2 h−1), The durability and washability of the nanoPE fabric is also
which is higher than that of cotton and the nanoPE thin film tested via a commercial washing and drying machine. After a regu-
with a water vapour transmission rate of 0.015 g cm−2 h−1). This lar duration of continuous washing and drying (Supplementary
advantage is attributed to the structure design of the nanoPE Fig. 13a,b), the nanoPE fabric stays intact (Supplementary Video 5
fabric and the nanoscale cavities inside the nanoPE microfibre. and Supplementary Fig. 13c,d). The nanoPE fabric is durable
The space among the fibres and the ideal channel network of for multiple cycles of washing and drying, maintaining high

Nature Sustainability | VOL 1 | FEBRUARY 2018 | 105–112 | www.nature.com/natsustain 109

© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Articles Nature Sustainability

a b
1.4 NanoPE fabric 14,000 Tyvek

Wator vapour transmission (g cm–2)

NanoPE film NanoPE film
1.2 12,000
Tyvek Cotton
Cotton

Pressure drop (Pa)

1.0 10,000 NanoPE fabric

0.8 8,000

0.6 6,000
0.4 4,000
0.2 2,000
0.0 0
0 10 20 30 40 50 60 0 20 40 60 80 100 120 140 160 180
Time (h) Air flow linear velocity (cm s–1)

c d NanoPE fabric + mesh Cotton

Tyvek NanoPE fabric NanoPE film

4 40
Wicking distance (cm)

3 30

Load (N)
2 20

10
1
Hydrophobic

0
0
0 50 100 150 200 250
m
n

ic
h tto ic

k
an PD E
to

ve
br
es o r

fil
+ oP
oP A
n
m + c fab
ot

Elongation (%)
fa

Ty
E
ic n
C

oP
br a

E
A E

fa N
D oP

an
+P an

N
N
N

Fig. 5 | Wearability tests for the nanoPE fabric and other textiles. a, A water vapour transmission rate test showing the speed of perspiration transmission
through the textile from human skin. b, An air permeability test measuring how well the textile allows the passage of air. c, A water-wicking test showing
the textile capability for perspiration absorbance, transportation and evaporation. d, A mechanical strength test showing the relationship between tensile
load and elongation.

mid-infrared transmittance (Supplementary Fig. 13e). For the (mould zone 1), 230 °C (barrel zone 3), 230 °C (barrel zone 2) and 230 °C (barrel
zone 1). Fibres of various recipes were all produced through the same process. For
purposes of clothing fashion, the white nanoPE fabric could be example, for the nanoPE fabric sample for characterization, we adopted the ‘1:3.5’
coloured by mixing the PE with a very small amount of biocompat- recipe (1 g PE with 3.5 ml paraffin oil) to produce fibres. Thirty-five millilitres of
ible pigment without compromising its mid-infrared transmittance. paraffin oil (Fisher Scientific), in which 0.078 g antioxidant (Irganox 1010, BASF)
Further study on the coloration is looking for solutions. was dissolved, and 10 g HDPE (pellets, melt index: 2.2 g per 10 min, Sigma Aldrich)
were fed in at the inlet as raw materials. After maintaining a suitable rotation
speed for a while, the roller was turned on to collect the fibre when the mixture of
Summary HDPE and paraffin oil was extruded out from the outlet of the mould. The fibre
In summary, we report an energy-saving fabric constructed from was quenched by air. The diameter of the fibre was adjustable by controlling the
rotation speed of the main motor and winding machine. All of the fibres were
nanoPE microfibres using economically viable materials and
transferred to bobbins, sent to Kimbrell Campus and Textile Technology Center
industrial fabrication technologies. The embedding of nanocavi- at Gaston College (North Carolina, USA), and knitted/woven there by industrial
ties into the PE microfibre not only ensures the visible opacity, equipment. The paraffin oil was extracted by methylene chloride (Fisher Scientific)
but also achieves cotton-like softness. Compared with the com- after the fabric was finished. The other way to produce nanoPE microfibres was
mercial cotton fabric, the nanoPE fabric’s superior cooling effect using multi-filament extrusion at Hills Inc. (North Carolina, USA). The other
textile samples are all commercially available: nanoPE thin film (Teklon, 12 µ​m,
brings about a 2.3 °C drop of the indoor set-point, which cor- Entek International LLC), cotton fabric (single jersey cotton, 130 g m−2) and
responds to a 20% energy saving. Besides the enticing cooling Tyvek (7 mil, Dupont).
effect, the nanoPE fabric also exhibits a collection of compelling
wearability features. With large-scale production, it can be imme- Sample characterization. The photograph of the nanoPE fibres and normal
diately and extensively adopted as a sustainable energy-saving PE fibres was taken with an optical microscope (Olympus). The SEM images
solution to improve our way of living. We expect that the nanoPE were taken by an FEI XL30 Sirion SEM (5 kV). The infrared transmittance
fabric will not only revolutionize textile for radiative cooling, but was measured by an FTIR spectrometer (Model 6700, Thermo Scientific)
accompanied with a diffuse gold integrating sphere (PIKE Technologies).
also achieve a breakthrough in reducing energy consumption for
The visible opacity was measured by an ultraviolet–visible spectrometer
global sustainability. (Agilent, Cary 6000i).

Methods Visible-light scattering simulation. When a beam of light hits an object, it will
Textile preparation. There were two methods for nanoPE microfibre production. be scattered in all directions (Supplementary Fig. 11a). The opacity is defined as
The nanoPE microfibre used in the nanoPE fabric was produced by an extruder (1 −​  specular transmittance)11. Both diffusive transmission and back reflection
(SJ25/25 Plastic Lab Extruder). Before feeding in materials, the extruder was help prevent the object from being recognized; thus, both contribute to opacity.
preheated for 1 h until the temperatures of four heating zones were stable at 180 °C Using Mie theory for cylinders29, the differential scattering cross-section versus

110 Nature Sustainability | VOL 1 | FEBRUARY 2018 | 105–112 | www.nature.com/natsustain

© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Nature Sustainability Articles
different scattering angles can be obtained (Supplementary Fig. 11b). As an initial prepared by attaching a cotton mesh with an opening of 1 ×​ 1 cm2 to the nanoPE
estimation, for a human observer with an eye aperture of 1 cm standing at a regular fabric sample. For the mechanical test for the fibres of various recipes, one sample
distance of 1 m from another human wearing the fabric, the specular angle can contained ten 5-cm-long fibres aligned carefully, and the gauge distance was 3 cm
be defined as being within the scattering angle of 0.01 rad, with the rest being long. The displacement rate was 20 mm min−1.
diffusive. The calculated differential scattering cross-section as a function
of diffraction angle is shown in Supplementary Fig. 11b. Integrating over the angle, Bending test. The bending test was conducted by TA instrument Q800 dynamic
the resulting ‘specular’ and ‘diffuse’ cross-sections defined above versus wavelength mechanical analysis in dual-cantilever mode. Normal PE microfibres and nanoPE
in the visible-light range are shown in Supplementary Fig. 11c, respectively. Finally, microfibres are all 170 μ​m ±​  20 μ​m in diameter. The tested samples were fibre
the opacity was calculated as: opacity =1− (Ttot−Tdif ) = 1− ((1−mσb)̄ −mσf,dif ̄ ). arrays with 30 fibres, and were 7 cm in length and 1.5 cm in width. The frequency
Here m is the density of the fibres, σb̄ is the total backscattering cross-section and sweeping range was 1 Hz–10 Hz.
̄ is the total diffusive forward-scattering cross-section. Inserting numbers as
σf,dif
m = 0.008 μ​m−1, σb= ̄ 14.65 μm, σf,dif
̄ = 52.76 μm, the opacity of the normal Porosity measurement. The porosity measurement was based on Archimedes’
principle. An object with higher density than water was sunk into distilled water.
PE fabric is 54%.
The increased volume (V1) was measured, which shall equal the volume of the
Simulation of thickness-dependent mid-infrared transmittance. We simulated object. The fibres were weighted and then immerged together with the object into
the infrared transmittance of the nanoPE fabric and the nanoPE film with the water. The increased volume (V2) was again measured, and the volume of the
varied thickness using rigorous coupled-wave analysis. Each fibre is modelled fibres was expressed as V2 −​  V1. The bulk density of the nanoPE fibres (ρnano) can
as a round cylinder, and the fabric is formed by a periodic array of fibres. As therefore be calculated. Note, the PE fibres are hydrophobic, so the water cannot
the infrared wavelength is much larger than the size of the nanopores, we used enter the nanopores. The density of normal, non-porous PE fibres (ρnormal) was also
effective-medium material parameters in simulating nanoPE textiles; that is measured with the same method. The porosity (ϕ) of nanoPE fibres is calculated
εnanoPE = (1−η) εPE + ηεair, where η is the porosity. as ϕ =​ 1 – ρnano/ρnormal.

Thermal measurement. We used a silicone-rubber/fibreglass-insulated flexible Data availability. The data that support the findings of this study are available
heater (Omega, 72 cm2) that was connected to a power supply (Keithley 2400). from the corresponding author upon request.
A ribbon-type hot-junction thermocouple (0.3 mm in diameter, K-type, Omega) was
in contact with the top surface of the simulated skin to measure the skin temperature. Received: 5 October 2017; Accepted: 16 January 2018;
Another heater the same as the simulated skin was placed below the simulated skin Published online: 9 February 2018
to act as a guard heater. The temperature of the guard heater was always set the
same as the skin heater, to ensure the heat flux of simulated skin is only upwards.
A thermocouple was in contact with the bottom surface of the guard heater. All of
References
1. King, D. A. Climate change science: adapt, mitigate, or ignore? Science 303,
the device was enclosed in a chamber, and the air temperature (ambient temperature)
176–177 (2004).
in the chamber was monitored by a third thermal couple and controlled at 23.5 °C
2. Chu, S. & Majumdar, A. Opportunities and challenges for a sustainable
by a circulating water heater/chiller. The heating power of the skin heater was set
energy future. Nature 488, 294–303 (2012).
to be 95 W m−2, which rendered the skin temperature of 33.5 °C when the ambient
3. Roemmich, D. et al. Unabated planetary warming and its ocean structure
temperature was 23.5 °C. During measurements, the simulated skin was covered
since 2006. Nat. Clim. Change 5, 240–245 (2015).
with textile samples (5.5 ×​ 4.5 cm2). The textile sample blocked the heat transfer from
4. Lee, R. The outlook for population growth. Science 333, 569–573 (2011).
the skin heater to the ambient environment, so the skin temperature increased. We
5. Goldstein, E. A., Raman, A. P. & Fan, S. Sub-ambient non-evaporative fluid
set the temperature of the guard heater to be the same as the temperature of the
cooling with the sky. Nat. Energy 2, 17143 (2017).
skin heater. The temperature regulation of the guard heater was applied with much
6. Hoyt, T., Arens, E. & Zhang, H. Extending air temperature setpoints:
caution to ensure the temperature of the guard heater was never higher than that of
simulated energy savings and design considerations for new and retrofit
the skin heater. When their temperatures were the same and stable for 20 min, we
buildings. Build. Environ. 88, 89–96 (2015).
read the displayed temperature as the temperature data. The ambient temperature in
7. Raman, A. P., Anoma, M. A., Zhu, L., Rephaeli, E. & Fan, S. Passive radiative
the measurements was stabilized at 23.5 °C ±​ 0.1 °C. The thermal images were taken
cooling below ambient air temperature under direct sunlight. Nature 515,
by a calibrated thermal camera (MikroSHOT, Mikron).
540–544 (2014).
Water vapour transmission rate test. The testing procedure was based on ASTM 8. Kou, J., Jurado, Z., Chen, Z., Fan, S. & Minnich, A. J. Daytime radiative
E96 with modification. Medium bottles (100 ml; Fisher Scientific) were filled with cooling using near-black infrared emitters. ACS Photonics 4, 626–630 (2017).
80 ml of distilled water, and sealed with the textile samples using open-top caps and 9. Zhai, Y. et al. Scalable-manufactured randomized glass-polymer hybrid
silicone gaskets (Corning). The exposed area of the textile was 3 cm in diameter. The metamaterial for daytime radiative cooling. Science 355, 1062–1066 (2017).
sealed bottles were placed into an environmental chamber in which the temperature 10. Hsu, P.-C. et al. Personal thermal management by metallic nanowire-coated
was held at 35 °C. The mass of the bottles and the samples was measured textile. Nano. Lett. 15, 365–371 (2014).
periodically. By dividing the reduced mass of the water by the exposed area of the 11. Hsu, P.-C. et al. Radiative human body cooling by nanoporous polyethylene
bottle (3 cm in diameter), the water vapour transmission rate was calculated. textile. Science 353, 1019–1023 (2016).
12. Tong, J. K. et al. Infrared-transparent visible-opaque fabrics for wearable
Air permeability test. The testing procedure was based on ASTM D737 with personal thermal management. ACS Photonics 2, 769–778 (2015).
modification. The textile sample was fixed between two pipes using double-side 13. Hardy, J. D. & Dubois, E. F. Regulation of heat loss from the human body.
tape with a 0.5 cm2 exposed area, and the pipe joint was sealed well to ensure its Proc. Natl Acad. Sci. USA 23, 624–631 (1937).
airtightness. One pipe was connected to a T-connector at the short/straight leg and 14. Chen, J. & Wang, Z. L. Reviving vibration energy harvesting and self-powered
then connected to the compressed air source. A flowmeter (Dwyer, maximum flow sensing by a triboelectric nanogenerator. Joule 1, 480–521 (2017).
rate: 5 l min−1) was placed between the compressed air source and the T-connector. 15. Viklund, C., Svec, F. & Frechet, J. M. J. Monolithic, “molded”, porous
The other pipe was also connected to a T-connector at the short/straight leg and materials with high flow characteristics for separations, catalysis, or
then connected to open air. A differential pressure gauge (UEi Test Instruments) solid-phase chemistry: control of porous properties during polymerization.
was connected to both long/branch legs to measure the pressure drops across the Chem. Mater. 8, 744–750 (1996).
textile sample at different air flow rates. The linear velocity of the air flow was 16. Rogers, J. A., Someya, T. & Huang, Y. Materials and mechanics for stretchable
calculated using the air flow rate divided by the 0.5 cm2 exposed area. electronics. Science 327, 1603–1607 (2010).
17. Khang, D.-Y., Jiang, H., Huang, Y. & Rogers, J. A. A stretchable form of
Wicking test. The testing procedure was based on AATCC TM 197 with single-crystal silicon for high-performance electronics on rubber substrates.
modification. The textile samples were cut into 1-cm-wide strips and dipped into Science 311, 208–212 (2006).
distilled water. Several drops of ink were added to the water for the convenience 18. Ahn, B. Y. et al. Omnidirectional printing of flexible, stretchable, and
of observation. We measured the climbing distance of the water in the duration of spanning silver microelectrodes. Science 323, 1590–1593 (2009).
30 s. As for the modification of the nanoPE fabric sample, a polydopamine coating 19. Fang, X., Wyatt, T., Hong, Y. & Yao, D. Gel spinning of UHMWPE fibres with
solution was prepared firstly. The polydopamine coating solution was made by polybutene as a new spin solvent. Polym. Eng. Sci. 56, 697–706 (2016).
dissolving dopamine hydrochloride (2 g l−1, Sigma-Aldrich) in Tris buffer solution 20. Ruan, S., Gao, P. & Yu, T.-X. Ultra-strong gel-spun UHMWPE fibers
(10 mM, pH 8.5, Teknova). Then the nanoPE fabric sample was soaked in the coating reinforced using multiwalled carbon nanotubes. Polymer 47,
solution for 1 h to make the surface hydrophilic. A cotton mesh with an opening of 1604–1611 (2006).
1 ×​ 1 cm2 was attached on the above sample to enhance the capillary force. 21. Samon, J. M., Schultz, J. M. & Hsiao, B. S. Structure development in the early
stages of crystallization during melt spinning. Polymer 43, 1873–1875 (2002).
Mechanical test. The tensile strength test was measured by an Instron 5565. For 22. Ihm, D., Noh, J. & Kim, J. Effect of polymer blending and drawing conditions
the mechanical test for the textile samples, the sample size was 1.5 cm wide and on properties of polyethylene separator prepared for Li-ion secondary battery.
3 cm long, and the gauge distance was 1 cm long. The ‘nanoPE fabric +​  mesh’ was J. Power Sources 109, 388–393 (2002).

Nature Sustainability | VOL 1 | FEBRUARY 2018 | 105–112 | www.nature.com/natsustain 111

© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
Articles Nature Sustainability
23. Liu, V. & Fan, S. S4: A free electromagnetic solver for layered periodic Author contributions
structures. Comput. Phys. Commun. 183, 2233–2244 (2012). Y.P., J.C. and Y.C. conceived the idea, planned the study, designed the experiment,
24. Ghahramani, A., Zhang, K., Dutta, K., Yang, Z. & Gerber, B. Energy savings analysed the data and composed the manuscript. Y.P. and J.C. performed all of the
from temperature setpoints and deadband: Quantifying the influence of experiments with the assistance of P.-C.H., L.C., B.L., G.Z., D.S.W. and H.R.L. Y.P. and
building and system properties on savings. Appl. Energy 165, 930 (2016). J.C. addressed all of the reviewers’ concerns together. A.Y.S. performed the optical
25. Chen, Z., Zhu, L., Raman, A. & Fan, S. Radiative cooling to deep sub-freezing simulation. Y.Z. performed the energy saving calculation. P.B.C. coordinated the project.
temperatures through a 24-h day–night cycle. Nat. Commun. 7, 13729 (2016). Y.C. and S.F. supervised the project. All of the authors reviewed and commented on
26. Lee, H., Dellatore, S. M., Miller, W. M. & Messersmith, P. B. Mussel-inspired the manuscript.
surface chemistry for multifunctional coatings. Science 318, 426–430 (2007).
27. Ryu, J., Ku, S.-H., Lee, H. & Park, C. B. Mussel‐inspired polydopamine
coating as a universal route to hydroxyapatite crystallization. Adv. Funct. Competing interests
Mater. 20, 2132–2139 (2010). Y.C., S.F., Y.P., J.C., A.Y.S., P.B.C. and P.-C.H. have a US patent application No. 62/399,974
28. Ryou, M.‐H., Lee, Y. M., Park, J.‐K. & Choi, J.W. Mussel‐inspired related to this work.
polydopamine‐treated polyethylene separators for high‐power Li‐ion
batteries. Adv. Mater. 23, 3066–3070 (2011).
29. Kerker, M. & Matijević, E. Scattering of electromagnetic waves from Additional information
concentric infinite cylinders. J. Opt. Soc. Am. 51, 506–508 (1961). Supplementary information accompanies this paper at https://doi.org/10.1038/s41893-
018-0023-2.
Acknowledgements Reprints and permissions information is available at www.nature.com/reprints.
This work was sponsored by the Advanced Research Projects Agency–Energy (ARPA-E),
US Department of Energy, under award DE-AR0000533. The authors thank H. Dai for Correspondence and requests for materials should be addressed to Y.C.
lending them the thermal camera, and thank J. Lopez and V. Feig for helping with sample Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in
measurements. published maps and institutional affiliations.

112 Nature Sustainability | VOL 1 | FEBRUARY 2018 | 105–112 | www.nature.com/natsustain

© 2018 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.