P1: KEE

Annals of Biomedical Engineering [AMBE]

pp1282-ambe-490790

July 28, 2004

17:49

Style file version 14 Oct, 2003

Annals of Biomedical Engineering, Vol. 32, No. 8, August 2004 (2004) pp. 11311140

Effects of Droplet Velocity, Diameter, and Film Height on Heat

Removal during Cryogen Spray Cooling
BRIAN M. PIKKULA,1 JAMES W. TUNNELL,1,2 DAVID W. CHANG,3 and BAHMAN ANVARI1
1

Department of Bioengineering, Rice University, Houston, TX; 2 Current address: G.R. Harrison Spectroscopy Laboratory,
Massachusetts Institute of Technology, Cambridge, MA; and 3 Department of Plastic Surgery, University of Texas
M.D. Anderson Cancer Center, Houston, TX
(Received 16 May 2003; accepted 31 March 2004)

result in nonspecific heating of the epidermis. Furthermore,

the fluence at which the threshold for epidermal damage occurs is lower in individuals with greater epidermal melanin
content,5,12 resulting in decreased laser treatment efficacy.
Protecting the epidermis from nonspecific heating can
be accomplished by precooling the skin prior to laser
irradiation.9,19,25,37,41 Immediately upon laser exposure,
the precooled epidermis ideally would be heated to a
temperature below the threshold for thermal injury. One
method of selectively cooling the epidermis is to spray a
short (on the order of tens of milliseconds) cryogen spurt
onto the skin surface.4 Cryogen spray cooling (CSC) has
proven to protect the epidermis from nonspecific thermal
injury.9,19,24,25,36,37 Although CSC can protect patients with
fair to medium pigmentation levels from nonspecific thermal injury,19,25,36 heavily pigmented patients such as individuals of Asian and African descent5,12,17 are at an increased risk for epidermal injury while undergoing cutaneous laser therapy due to increased energy absorption by
the higher concentration of melanin within the epidermis.
Increasing the amount of heat removal from skin by optimization of CSC can potentially allow individuals with
darker skin to benefit from cutaneous laser therapy, and allow for higher incident fluences than those currently used
in patients with lighter skin types.
The first goal of this study was to determine how cryogen droplet velocity and diameter influenced heat removal.
Motivation to alter the droplet velocity and diameter was
as follows. Because of the low thermal diffusivity of skin
and the skin phantom used in this study, a cryogen film
build-up on the surface occurs during spraying.1,28,35 We
speculate that due to the cold droplets (55 C) impinging the residing cryogen film on the surface and the deeper
portions of the film warmed by the substrate, a thermal gradient exists within the cryogen film. This gradient may also
be augmented by the low thermal diffusivity of the cryogen
film.18 Therefore, the cryogen film itself may act as barrier to heat transfer between the impinging droplets and the
substrate.

AbstractCryogen spray cooling (CSC) is an effective method

to reduce or eliminate epidermal damage during laser treatment
of various dermatoses. This study sought to determine the effects
of specific cryogen properties on heat removal. Heat removal was
quantified using an algorithm that solved an inverse heat conduction problem from internal temperature measurements made
within a skin phantom. A nondimensional parameter, the Weber
number, characterized the combined effects of droplet velocity,
diameter, and surface tension. CSC experiments with laser irradiation were conducted on ex vivo human skin samples to assess
the effect of Weber number on epidermal protection. An empirical
relationship between heat removal and the difference in droplet
temperature and the substrate, droplet velocity, and diameter was
obtained. Histological sections of irradiated ex vivo human skin
demonstrated that sprays with higher Weber numbers increased
epidermal protection. Results indicate that the cryogen film acts
as an impediment to heat transfer between the impinging droplets
and the substrate. This study offers the importance of Weber number in heat removal and epidermal protection.

KeywordsDermatology, Laser therapy, Skin, Surface tension,

Surfactants, Weber number.

INTRODUCTION
Laser radiation is currently utilized to remove unwanted cutaneous structures such as hypervascular lesions,
rhytides, and hair.13,20,30,39 The unwanted structures are
heated by prescribing a particular wavelength which is
preferentially absorbed by the targeted chromophore (e.g.,
hemoglobin, water, or melanin), and specifying an appropriately short laser pulse to limit thermal diffusion from
the chromophore. These targeted chromophores undergo
photothermolysis, in which the energy deposited to the desired chromophore results in its thermal destruction.3 Despite the spatial confinement of heat within the targeted
chromophore, light absorption by the overlying epidermal
melanin which takes place over a broad spectral range26 can
Address correspondence to Bahman Anvari, Department of Bioengineering, MS 142, Rice University, P.O. Box 1892, Houston, TX 77251.
Electronic mail: anvari@rice.edu

1131
C 2004 Biomedical Engineering Society
0090-6964/04/0800-1131/1 

P1: KEE
Annals of Biomedical Engineering [AMBE]

pp1282-ambe-490790

July 28, 2004

1132

17:49

Style file version 14 Oct, 2003

PIKKULA et al.

Verkruysse et al.40 postulated that droplets possessing

greater momentum penetrate deeper into the cryogen film,
and come to closer proximity or even direct contact with the
substrate. Since droplets have lower temperatures relative
to the film, the reduction in separation distance between the
droplet and the substrate will increase the thermal conductance, and result in enhanced heat removal.
Increasing the droplet penetration into the film (or creating a thinner film) would allow the cold droplets to come in
closer proximity or even in direct contact with the substrate,
thus enhancing the rate of heat removal by reducing the distance the cold wave must propagate through a low thermally
diffusive cryogen film. The cryogen film composed of the
residing and impinging droplets is hypothesized to impede
heat removal; therefore, it was our aim to modify the droplet
properties, namely, increase droplet velocity or diameter to
augment the droplet penetration depth into the film.
Our second objective was to investigate the effect of the
cryogen film height reduction on heat removal by reducing
the cryogen surface tension. We speculated that creating
a thinner film would allow the cryogen droplets to come
in closer proximity to the substrate surface, resulting in
an increased rate of heat removal by reducing the distance
of heat diffusion between the impinging cold droplets and
substrate.
Lastly, we sought to determine if increased heat removal
would result in greater protection of the epidermis in response to laser irradiation of ex vivo human skin samples.
Results of this study can be used to optimize the appropriate range of cryogen droplet velocity and diameter for
use in conjunction with laser treatment of individuals with
various skin types. In summary, the objective of this work
was to investigate how droplet velocity and diameter, surface tension, and cryogen film height influence heat removal
and affect thermal damage to human skin epidermis during
laser irradiation.
THERMOPHYSICAL PARAMETERS AND HEAT
REMOVAL ESTIMATION
Weber Number
To account for the changes in both the altered droplet
velocity and diameter, we used the Weber number. Studies
on the general subject of spray cooling have shown that the
dimensionless Weber number (We) influences the instantaneous heat flux (q We0.6 when using Freon-113 sprayed
on heated metal substrates).15 We is proportional to the ratio
of the kinetic energy to the surface energy of the droplets:
 
 
dimpact 3
1
4
2
3
vimpact
2
2
 
We =
,
(1)
2 
d
4 impact
2
3

where is the liquid density (kg m ), dimpact the droplet

diameter (m), vimpact the droplet velocity (m s1 ), and is

the liquid surface tension (N m1 ). Eliminating constants

and reducing gives
We =

2
dimpact vimpact

(2)

Heat Removal Estimation

Inasmuch as thermal boundary conditions such as a timevarying heat flux [q(t)] are difficult to measure directly, indirect techniques using internal temperature measurements
are often used.32 With these techniques, an inverse heat conduction problem (IHCP) is solved to estimate the boundary
condition from internal temperature measurements.
We used the sequential function specification (SFS)
method6 to solve the IHCP from internal temperature measurements at a known depth within the phantom substrate
(described in the next section) at discrete times. This method
estimates the surface heat flux as a piecewise function of
time, sequentially solving for q(t) at each time point. In
doing so, the method uses future temperature data to estimate the surface heat flux at the current time point,31 and
minimizes the following least squares expression over future time steps. The sensitivity and accuracy of the IHCP is
described fully in a previous article on this technique by our
group.38 The total heat removal per unit area, Q (J m2 ),
the time-integral of q(t) at the substrate surface (z = 0),
was determined over the 200-ms cryogen spurt from the
solution of the IHCP for various spraying conditions.
Thermal Conductance
Thermal conductance, (t) (W m2 K1 ), between the
impinging droplet and substrate was calculated as
(t) =

q(t)
,
Ts (t) Tdroplet

(3)

where Ts is the temperature of the substrate at z = 0 (computed by solving the forward heat transfer problem once q(t)
was estimated) and Tdroplet is the cryogen droplet temperature just prior to impacting the substrate, measured by inserting a 60-m bead diameter thermocouple (CHAL-001,
OMEGA Engineering, Inc., Stamford, CT) into the
cryogen spray. Since calculation of utilizes measured
and estimated temperatures [i.e., Tdroplet and T (z = 0, t),
respectively], all thermal resistances between the droplet
and substrate (i.e., at cryogen-substrate interface and within
the cryogen film on the surface) are taken into account.
The dynamic thermal conductance was time averaged over
the duration of the cryogen spurt, resulting in an average
value, .
Surface Tension
The single capillary rise method is a classical method
of quantification of surface tension.22 To simplify the

P1: KEE
Annals of Biomedical Engineering [AMBE]

pp1282-ambe-490790

July 28, 2004

17:49

Style file version 14 Oct, 2003

Effects of Specific Cryogenic Properties on Heat Removal

measurement technique,16 the differential capillary rise

method (DCRM) was used in this study, in which two capillaries, arranged vertically and with different inner diameters, were lowered into a pool of the liquid being tested.
With dissimilar diameters, the column of liquid rises higher
in the capillary with the smaller inner diameter and the
height difference in the two menisci are measured.
The downward and upward forces, respectively, acting
on a column of fluid in a single capillary are
force = [hr 2 ( )]g
force = [2r ] cos ,

(4)

where is the surface tension of the liquid Nm1 ; g the

gravitational constant (9.8 ms2 ); and are the liquid
and vapor densities of the cryogen (1206 and 5.3 kgm3 ,
respectively); h 1 h 2 is the difference in height of the two
menisci (m); r1 and r2 (m) are the inner radii of each of the
capillaries; and is the angle the liquid contacts the inner
wall of the capillary. After equating the force balance and
substituting the differences in the heights and differences
in the area and circumferences of the two capillaries, the
DCRM yields the surface tension using simplified method
versus that of the single capillary method.29
=

g( )(h 1 h 2 )
2(1/r1 1/r2 ) cos

(5)

Images of the capillaries were captured with a CCD camera and an attached a zoom-macro lens, described in the
next section, using ambient lighting, and NIH Scion Image (www.scioncorp.com) was utilized to measure h 1 h 2
and .
EXPERIMENTAL METHODS
Cryogen Type and Delivery Systems
The cryogen used in this study was refrigerant R134a, 1,1,1,2 tetrafluoroethane (National Refrigerants, Inc.,
Rosenhayn, NJ), which is a nontoxic, environmentally compatible, Freon substitute2,23 (boiling point 26 C at 1
atm) used clinically for epidermal protection and pain reduction during cutaneous laser therapy. Spurt duration was
set to 200 ms, and controlled by a programmable digital
delay generator (DG535, Stanford Research Systems, Sunnyvale, CA).
Droplet velocity and diameter were altered by three
methods. First, immersing the cryogen canister into a warm
water bath led to increased pressure within the canister,
causing a more forceful release of the cryogen at a higher
ejection velocity. The second method was to change the
spraying distance between the substrate and cryogen delivery device. Since droplets evaporate and cool in flight,
adjusting the spraying distance also alters droplet impact
temperature. However, droplet velocity and diameter could
be independently varied from the droplet temperature when

1133

changing the pressure of the cryogen canister. Third, different cryogen delivery devices were used to produce droplet
diameter and velocity dependent on the design of the device.
Three types of cryogen delivery devices were used. The
first device (Device 1) was a standard automobile fuel injector with a 1-mm diameter orifice and an attachment nozzle
with a length of 2 mm and 1.5-mm orifice diameter that
produced a uniform spray cone with the following spray
characteristics (standard deviation) at ambient conditions:
average droplet diameter of 26 2.1 m and Sauter mean
diameter of 66 17 m at a spraying distance of 100 mm.27
The Sauter mean diameter is commonly used to quantify
droplet size during spraying processes and is calculated as
the diameter of a single droplet whose volume to surface
area ratio is equal to the sum of the volume over the sum of
the surface area of all droplets in the droplet set:
N
SMD = i=1
N
i=1

Di3 n i
Di2 n i

(6)

where D is the median droplet diameter of the given bin,

n the number of droplets in that bin, i the bin number, and
N is the total number of bins. The second device (Device
2) consisted of a fuel injector with a 1.3-mm orifice diameter (without an attachment nozzle) producing an average
droplet diameter of 34.0 3.7 m and Sauter mean diameter of 116.1 16 m at a spraying distance of 90 mm.27
The third spraying system (Device 3) consisted of the second device above with a 40-mm long attachment nozzle and
a 1.1-mm diameter orifice. This device produced an average
droplet diameter of 32 4 m and a Sauter mean diameter
of 101 22 m at a spraying distance of 90 mm.
Droplet Impact Velocity and Diameter
Droplet impact velocity ( impact ) measurements were
conducted by imaging the droplets (at spraying distances
of 40130 mm or 90 mm for the case when the cryogen canister temperature was altered) using a CCD camera, a 2 focal length extender (EX2C, Computar CBC
AMERICA Corp., Commack, NY), and a macrozoom lens
(Zoom 7000, Navitar, Japan) (Fig. 1). Resolution of the
imaging system (interpixel distance) was 3.7 m/pixel, determined by dividing the image height (1.9 mm) by the
number of pixel rows (515). Illumination sources were two
10 ns flashlamps (TWINLITE, High-Speed Photo-Systeme,
Wedel, Germany), which were coupled with a beam splitter
to project along the optical axis of the CCD camera. The
cryogen was sprayed perpendicular to the optical axis between the flashlamp and CCD camera. Triggering the two
flashlamps to spark within predetermined time intervals (1
4 s with respect to each other) exposed the droplets twice
on the same image. Velocity was subsequently calculated as
the ratio of the droplet travel distance to the delay between
the two flashes. The droplet size was determined from the

P1: KEE
Annals of Biomedical Engineering [AMBE]

pp1282-ambe-490790

1134

July 28, 2004

17:49

Style file version 14 Oct, 2003

PIKKULA et al.

FIGURE 2. Schematic of experimental setup used to measure

internal temperatures within the phantom substrate.

FIGURE 1. Schematic of the setup used for droplet velocity

measurements.

same images by quantifying the diameters in pixels and

converting the values into microns.
Phantom Substrate
A phantom substrate composed of epoxy resin with
imbedded microthermocouples was used to record internal temperature profiles in response to CSC (Fig. 2). The
thermal diffusivity () of the epoxy substrate, 0.843
107 m2 s1 (EP30, Master Bond Inc., Hackensack NJ),
was within the range of human skin, 6.9 108 to 1.1
107 m2 s1 .14 This phantom type has been used extensively in previous studies.27,33,34,40 Type K thermocouples
(ChromegaAlomega, Omega Engineering, Inc., Stamford, CT) with 30-m bead diameters, positioned at 20, 90,
200, and 400 m (all 5 m) below the phantom surface
were used to measure internal temperatures in response to
CSC. The measured temperature profiles were subsequently
used as input data for the inverse model, described in the
previous section, to estimate the heat flux at the substratecryogen interface. The entire substrate was warmed to 33 C
prior to each experiment by applying heated air to the
phantom.

removal. Three surfactants (Table 1), Brij 30, Tween 20, and
polyethylene glycol (PEG) 300, were chosen for this study
based on their solubility in R-134a, current applications in
drug delivery, and low toxicity.7 These three surfactants are
among the highest for solubility in R-134a.7
Several concentrations (w/w) were used for each surfactant (Table 1). The upper limit of each surfactant concentration in R-134a was determined by the maximum reported
solubility.7 If excessive foaming was present after spurt termination, a decreased concentration was used.
Surfactants were added to the pressurized canisters by
cold filling which consisted of attaching a regulator to
the canister, and placing it in a cooler filled with dry
ice to reduce the vapor pressure of the cryogen to atmospheric levels. After several minutes, temperature of the
fluid within the canister was well below the boiling point of
R-134a allowing the removal of the regulator and exposing
the contents to atmospheric pressure without releasing its
contents. Surfactants were subsequently added via a calibrated pipette to create a surfactant/R-134a mixture in a w/w
proportion.
In some experiments, surfactants were spread on the substrate surface to reduce the cryogen surface tension. In these
cases, two drops (5060 l) of surfactant were placed onto
a Kimwipe (Kimberly-Clark, Roswell, GA), and spread
manually on the substrate to provide an even distribution of
surfactant on the surface (area 16 cm2 ). When the surfactant was spread on the surface, the resulting surface tension
was not measurable. Whether the surfactant was added to
the cryogen or spread on the substrate surface, a Kimwipe
was used to remove any residue from the substrate from the
prior spurt.
Normalized Cryogen Film Height

Reduction of Cryogen Surface Tension

Surfactants reduce the intermolecular forces within a liquid, and when in contact with a surface, cause the spreading
of the liquid under gravity or other external forces. Surfactants were added to the cryogen liquid to reduce the cryogen film height and increase the droplet Weber number, and
subsequently, investigate the effects on the amount of heat

Although the absolute cryogen film height was not measured, we normalized the film height for a certain spraying
condition compared to those under different spraying conditions. The apparent cryogen film height (H ) on the substrate
was estimated as
m
H=
,
(7)
A

P1: KEE
Annals of Biomedical Engineering [AMBE]

pp1282-ambe-490790

July 28, 2004

17:49

Style file version 14 Oct, 2003

Effects of Specific Cryogenic Properties on Heat Removal

1135

TABLE 1. Surfactants used in the study.

Common name

Chemical name

Chemical
composition

Maximum reported7 solubility in

R-134a (% w/w proportion)

% w/w Proportions
used

Brij 30

Polyoxyethylene(4)
lauryl ether

C12 H25 (OCH2 CH2 )4 OH

1.8%

0.3%, 1.8%

Tween 20

Polyoxyethylene(20)
sorbitan monolaurate

C58 H114 O26

0.1%

0.02%, 0.1%

PEG 300

Poly(oxy-1,2-ethanediyl),
alpha-hydro-omega-hydroxy

H(CH2 CH2 O)6 OH

where m (kg) is the cryogen mass output from the orifice

of the cryogen delivery device, A (m2 ) is the area covered
by the cryogen film on the substrate, and (kg m3 ) is the
cryogen density.
The value of apparent height was normalized to that of
Device 2, which produced the lowest Weber number at room
temperature to obtain the normalized cryogen film height,
. Cryogen mass output from the orifice for each of the
three delivery devices under different canister temperatures
was measured for 200 ms spurts. The total mass loss of
the cryogen from the canister was divided by the number
of 200 ms spurts actuated (approximately 10), resulting in
mass output per spurt. The sprayed area on the surface of the
substrate immediately after spurt termination was measured
by a CCD image under ambient lighting conditions. This
methodology was based on the assumption that regardless
of the mass output for each of the cryogen delivery devices,
an equal fraction of cryogen was evaporated in flight.

4%

0.5%, 1%

spurts immediately preceding the laser pulse. Each cooling and irradiation combination utilized either of the two
cryogen delivery devices which produced droplets with the
greatest (Device 1) and least (Device 2) Weber numbers
with the cryogen canister at room temperature. Following
laser irradiation, 6-mm punch biopsies were excised and
fixed in 10% buffered formalin and hematoxylin & eosin
stained for histological analysis of thermal damage. Thermally mediated morphological damage was characterized
by a score of 0 (no observable damage), 1 (nuclei shrinkage), 2 (<10% dermal-epidermal separation), 3 (10% to
50% dermal-epidermal separation), and 4 (>50% dermalepidermal separation). The percentage of separation was
quantified by the linear distance ratio of the separated to intact tissue at the dermalepidermal junction for the portion
of the histological section exposed to laser irritation.

RESULTS
Laser Irradiation of ex Vivo Human Skin
in Conjunction with CSC
We sought to determine if altering the cryogen droplet
and velocity (hence, Weber number) could change the
amount of epidermal protection ex vivo in response to laser
irradiation. Skin samples were obtained from consenting patients undergoing autograft breast reconstruction using abdominal skin. CSC and laser irradiation were subsequently
performed on the samples. The study was carried out under
a protocol approved by the institutional review boards of
Rice University and The University of Texas M. D. Anderson Cancer Center.
One of each Fitzpatrick skin types V and VI (medium
brown and dark brown, respectively) at room temperature
were irradiated using the Candela (Wayland, MA) Vbeam
pulsed dye laser (595 nm wavelength). The temporal pulse
profile in this laser consists of a macropulse in which a
train of four 100-s micropulses are placed in equal time
intervals within the macropulse. In our experiments, the
macropulse was set to 6 ms; hence, delivering four 100-s
micropulses. Depending on the size of the excised skin,
one or two sites were cooled and irradiated using a combination of 10 or 15 J cm2 , and 100- or 200-ms cryogen

Cryogen Surface Tension

Published surface tension values for R-134a from
McLinden et al.21 and Chae et al.8 were extrapolated to
26 C which resulted in values of 15.5 and 15.4 mNm1 ,
respectively. Published values concur with the surface
tension measurements of R-134a in this study, 15.5
0.3 mNm1 (Table 2), verifying our method and results. For
the cases where the surfactant was spread on the substrate
surface, cryogen surface tension was not measurable. The
maximum reduction in surface tension, when the surfactant
was added to the cryogen canister was approximately 16%.
We nevertheless found no correlation between heat removal
and cryogen surface tension (data not shown). Additionally,
no correlation was found with and cryogen surface tension
(data not shown).
Droplet impact velocity ( impact ) data for pure R-134a
and with a 1% w/w PEG 300/cryogen mixture were 33.9
3.5 ms1 and 35.9 3.7 ms,1 respectively, at a spraying
distance of 90 mm. A standard t-test ( p < 0.05) showed no
statistical difference between the resulting velocities. However, the standard deviation of 10% provides a relatively
large variation in the spurt to spurt velocities. When analyzing the Weber number, discussed in the next section, a 10%

P1: KEE
Annals of Biomedical Engineering [AMBE]

pp1282-ambe-490790

July 28, 2004

1136

17:49

Style file version 14 Oct, 2003

PIKKULA et al.

TABLE 2. Surface tension of R-134a after addition of

surfactants.
Concentration of surfactant
in R-134a (% w/w)

Surface tension
SD (mNm1 )

0 (pure R-134a)
0.3 (Brij 30)
1.8 (Brij 30)

15.5 0.3
13.1 0.9
Not measurable
due to foaming
13.7 0.9
15.2 0.4
15.6 0.8
13 0.6

0.02 (Tween 20)

0.1 (Tween 20)
0.5 (PEG 300)
1 (PEG 300)

standard deviation in velocity is sufficient to overshadow

the effects of the reduction in cryogen surface tension.
Weber Number
As the cryogen canister temperature was increased, the
cryogen was expelled with a greater velocity, but produced
finer droplets. As the spraying distance was increased, both
droplet velocity and diameter decreased due to air-induced
drag force and droplet evaporation in flight. By only changing the temperature of the cryogen canister, altering the
spraying distance, or using one of the three different cryogen delivery devices, we were unable to independently vary
vimpact and droplet diameter just prior to impact (dimpact );
however, using a combination of the three methods we were
able to independently vary these spray parameters. Ranges
of vimpact and dimpact were 1272 m s1 and 2352 m, respectively, for cryogen canister temperatures of 1550 C
and spraying distances of 40130 mm while using the three
cryogen delivery devices.
The influence of Weber number on heat removal is shown
in Fig. 3. In general, heat removal increased with higher

FIGURE 4. Loglog plot of time-averaged thermal conductance, , versus Weber number for a 200-ms spurt.

Weber number. A similar trend was present between the

Weber number and the time averaged thermal conductance
(Fig. 4), indicating that there was reduced thermal resistance
between the impinging droplets and substrate when We was
increased. The time-averaged thermal conductance for the
duration of the cryogen spurt resulted in a coefficient of
variation of less than 0.08 for the cryogen delivery devices.
Normalized Cryogen Film Height
The scatter plot in Fig. 5 illustrates that increased Weber
number reduced on the surface of the substrate. Similarly,
as was decreased, heat removal and increased (Fig. 6).
Results demonstrated no correlation between actual cryogen mass output and heat removal (data not shown) which
is due to the substrate being in a flooded state.
Predicting Heat Removal
Altering the spraying distance, cryogen canister temperature, and using three different cryogen delivery devices
provided 50 different sets of experimental parameters (i.e.,
Q for a given combination of T , dimpact , and impact )
where T (in the range of 7592 C) is the temperature
difference between the impinging droplet just prior to impact and the initial substrate temperature (33 C). A nonlinear least-squares data fitting algorithm, using the Gauss
Newton method in MatLab (Natick, MA), produced the
proportionality
0.1261 0.1601
Q 0.739T dimpact
vimpact ,

FIGURE 3. Loglog plot of heat removal, Q, versus Weber number for a 200-ms cryogen spurt.

(8)

where dimpact is the droplet diameter (m), and vimpact is

the cryogen droplet velocity just prior to impact (m s1 ).
Analysis of dimpact and vimpact indicated that they were independent values (R 2 = 0.16). Equation (8) is dependent
on the droplet parameters prior to impact regardless of how
the droplets were generated. The sensitivity of Q on dimpact
and impact is less than 13% for an exponent increase of 10%.

P1: KEE
Annals of Biomedical Engineering [AMBE]

pp1282-ambe-490790

July 28, 2004

17:49

Style file version 14 Oct, 2003

Effects of Specific Cryogenic Properties on Heat Removal

FIGURE 5. The normalized cryogen film height, , versus the

Weber number of the cryogen spray.

We emphasize that Eq. (8) is valid for the epoxy substrate

used in this study with = 0.843 107 m2 s1 .
Ex Vivo Epidermal Protection
Protection of epidermis from laser-induced thermal injury in Fitzpatrick type V skin (Fig. 7) was increased
when using Device 1, which produced the spray with a
greater Weber number when using a 100-ms spurt and a
fluence of 10 Jcm2 . Note the complete dermalepidermal
separation [Fig. 7(A)] and areas of collagen coagulation
in the upper papillary dermis due to heat diffusing from
the melanosomes in combination with suboptimal cooling.
Most of the epidermis was spared when using a device (Device 1) producing a greater Weber number (Fig. 7B). Using
equal spurt durations and laser fluences on each treatment
pair, histological damage scores were compared between
the two cryogen delivery devices which produced the greatest and least Weber number droplets in this study. Results
demonstrated an average increase in epidermal protection

FIGURE 6. Heat removal (black) and average thermal conductance, (grey), for a 200-ms spurt versus the normalized cryogen film height, .

1137

(standard deviation) of protect = 1.93 1.77 when using the highest Weber number device (Device 1) as compared to the lowest Weber number device (Device 2). For
instance, a decrease in the thermal damage score of approximately 2 would reduce epidermal damage from between
10 and 50% dermalepidermal separation to that of just nuclei shrinkage. Comparing protect to a null hypothesis (no
increase in epidermal protection when using a spray with
higher Weber number droplets), protect was statistically
larger ( p < 0.05). The relatively large standard deviation
from the mean value of protect is likely due to biological
variation and a digression of up to 3% from nominal laser
fluences. In all cases, the device that produced the highest
Weber number provided equal or greater epidermal protection when compared to the device that produced the lowest
Weber number droplets.
DISCUSSION
The Weber number is sensitive to the variations in droplet
velocity (especially at higher velocities), and slight changes
in velocity could completely mask the effects of reducing
the cryogen surface tension. On the basis of the results of
this study, increased heat removal was not observed with
decreased cryogen surface tension due to the effects of the
variations in velocity of the cryogen droplets. These variations appear to obscure the expected increase in heat removal by the altered cryogen surface tension; nevertheless,
we have observed that a thinner film generated by a high Weber number spray allows for increased heat removal (Fig. 6).
Additionally, at high Weber numbers (i.e., We > 10) such
as the values described in this text (We > 800), the changes
in kinetic energy dominate the changes in surface energy of
the droplet. When We 1, we would expect an equal effect
of the two energies, and if We < 1, the surface tension will
dominate.
The spray impacts the film and substrate, and its energy
is transferred in the radial direction creating a larger sprayed
area; therefore, film diameters increase with greater Weber
numbers, resulting in thinner films.11 Our results agree with
Ref. 11 and indicate that a thinner film is beneficial in increasing heat removal (Fig. 6), supporting the theory that the
cryogen film acts as an impediment to heat transfer (Fig. 4),
where thermal conductance increases with increased Weber
number (i.e., thermal resistance decreases with higher Weber number values). Possible methods to enhance heat removal by CSC would be to decrease the cryogen film height,
or equivalently, induce deeper droplet penetration into the
film, both of which may be occurring using droplets with
greater Weber numbers.
From Eq. (8), the relative contribution of the parameters
influencing heat removal can be determined. Considering
the range of dimpact and impact and their respective exponents, the influence of impact on heat removal is approximately six times that of dimpact , which was ascertained by

P1: KEE
Annals of Biomedical Engineering [AMBE]

1138

pp1282-ambe-490790

July 28, 2004

17:49

Style file version 14 Oct, 2003

PIKKULA et al.

FIGURE 7. H&E histological sections of ex vivo Fitzpatrick type V skin cooled with a 100-ms cryogen spurt immediately prior to
pulsed laser irradiation with a fluence of 10 J cm2 for Weber numbers of (A) 1,100 (Q = 26.9 kJ/m2 ), and (B) 5,100 (Q = 33.2 kJ/m2 ).
Arrows indicate areas of slight nuclei shrinkage and dermal-epidermal separation.

the ratio of the mean values of dimpact to impact raised to their

respective powers in Eq. (8). The parameter T was by far
the most influential factor in determining heat removal.
Our experiments have shown that sprays with a higher
Weber number increase the heat removal from a skin phantom. Results using ex vivo human skin samples suggest
sprays with a higher Weber number does augment heat removal from skin, resulting in a statistically significant increased epidermal protection in response to laser irradiation.
These results are encouraging in that the ability to increase
heat removal may allow individuals with darker skin types
to benefit from cutaneous laser therapy, and lighter skinned

patients will be able to tolerate greater light dosages, which

in turn, may reduce the number of treatment sessions. Additionally, transient adverse effects such as hyperpigmentation (lasting up to 1 year) in darker skin tones10 by laser
irradiation are expected to be minimized or eliminated with
increased heat removal due to a further reduction in temperature at the basal layer.
CONCLUSION
Results of this study indicate the cryogen film acts as an
impediment to heat transfer between the impinging droplets

P1: KEE
Annals of Biomedical Engineering [AMBE]

pp1282-ambe-490790

July 28, 2004

17:49

Style file version 14 Oct, 2003

Effects of Specific Cryogenic Properties on Heat Removal

and the substrate. The cryogen droplet Weber number is

a vital spray parameter that influences heat removal. Ex
vivo epidermal protection by CSC is augmented by utilizing
droplets with higher Weber numbers.

ACKNOWLEDGMENTS
This study was supported in part by the Institute
of Arthritis, Musculoskeletal, and Skin Disease (1R01AR47996) at NIH; Texas Higher Education Coordinating
Board, Candela Corporation to B.A.; and a Student Research Grant from the American Society for Lasers in
Medicine and Surgery to B.M.P.

REFERENCES
1

Aguilar, G., B. Majaron, K. Pope, L. O. Svaasand, E. J. Lavernia,

and J. S. Nelson. Influence of nozzle-to-skin distance in cryogen
spray cooling for dermatologic laser surgery. Lasers Surg. Med.
28:113120, 2001.
2
Alexander, D. J., and S. E. Libretto. An overview of the toxicology of HFA-134a (1,1,1,2-tetrafluoroethane). Hum. Exp. Toxicol. 14:715720, 1995.
3
Anderson, R. R., and J. A. Parrish. Selective photothermolysis:
Precise microsurgery by selective absorption of pulsed radiation.
Science 220:524527, 1983.
4
Anvari, B., T. E. Milner, B. S. Tanenbaum, S. Kimel, L. O.
Svaasand, and J. S. Nelson. Selective cooling of biological
tissues: Application for thermally mediated therapeutic procedures. Phys. Med. Biol. 40:541252, 1995.
5
Ashinoff, R., and R. G. Geronemus. Treatment of a port-wine
stain in a black patient with the pulsed dye laser. J. Dermatol.
Surg. Oncol. 18:147148, 1992.
6
Beck, J. V., B. Blackwell, and C. R. St. Clair Jr. Inverse Heat
Conduction: Ill-Posed Problems.New York: Wiley, 1985.
7
Blondino, F. E., and P. R. Byron. Surfactant dissolution and
water solubilization in chlorine-free liquified gas propellants.
Drug Dev. Ind. Pharm. 24:935945, 1998.
8
Chae, H. B., J. W. Schmidt, and M. R. Moldover. Surface tension
of refrigerants R123 and R134a. J. Chem. Eng. Data 35:68,
1990.
9
Chang, C. J., B. Anvari, and J. S. Nelson. Cryogen spray cooling
for spatially selective photocoagulation of hemangiomas: A new
methodology with preliminary clinical reports. Plast. Reconstr.
Surg. 102:459463, 1998.
10
Chang, C. J., K. M. Kelly, M. J. C. van Gemert, and J. S. Nelson.
Comparing the effectiveness of 585-nm vs. 595-nm wavelength
pulsed dye laser treatment of port wine stains in conjunction with
cryogen spray cooling. Lasers Surg. Med. 31:352358, 2002.
11
Chaves, H., A. M. Kubitzek, and F. Obermeier. Dynamic process
occurring during the spreading of thin liquid films produced by
drop impact on hot walls. Int. J. Heat Fluid Flow 20:470476,
1999.
12
Chung, J. H., W. S. Koh, and J. I. Youn. Histological responses
of port wine stains in brown skin after 578 nm copper vapor laser
treatment. Lasers Surg. Med. 18:358366, 1996.
13
Dover, J. S., and K. A. Arndt. New approaches to the treatment
of vascular lesions. Lasers Surg. Med. 26:158163, 2000.
14
Duck, F. A. Physical Properties of Tissue. A Comprehensive
Reference Book. London: Academic Press, 1990.

15

1139

Ghodbane, M., and J. P. Holman. Experimental study of spray

cooling with Freon-113. Int. J. Heat Mass Transfer 34:1163
1174, 1991.
16
Higashi, Y., and T. Shibata. Surface tension for 1,1,1trifluoroethane (R-143a), 1,1,1,2-tetraflluoroethane (R134a),
1,1-dichloro-2,2,3,3,3-pentafluoropropane (R-225ca), and 1,3dichloro-1,2,2,3,3-pentafluoropropane (R-225cb). J. Chem.
Eng. Data 42:438440, 1997.
17
Ho, W. S., H. H. Chan, S. Y. Ying, and P. C. Chan. Laser treatment
of congenital facial port-wine stains: Long-term efficacy and
complication in Chinese patients. Lasers Surg. Med. 30:4447,
2002.
18
Ibreighith, M., M. Fiebig, L. Leipertz, and G. Wu. Thermal diffusivity of the alternative refrigerants R123, R134a, R142b, and
R152a in the liquid phase. Fluid Phase Equlib. 80:323332,
1992.
19
Kelly, K. M., J. S. Nelson, G. P. Lask, R. G. Geronemus, and L. J.
Bernstein. Cryogen spray cooling in combination with nonablative laser treatment of facial rhytides. Arch. Dermatol. 135:691
694, 1999.
20
Lask, G., M. Elman, M. Slatkine, A. Waldman, and Z. Rozenberg. Laser-assisted hair removal by selective photothermolysis.
Dermatol. Surg. 23:737739, 1997.
21
McLinden, M. O., J. S. Gallagher, L. A. Weber, G. Morrison,
D. Ward, A. R. H. Goodwin, M. R. Moldover, J. W. Schmidt,
H. B. Chae, T. J. Bruno, J. F. Ely, and M. L. Huber. Measurement and formulation of the thermodynamic properties of refrigerants 134a (1,1,1,2-tetrafluoroethane) and 123 (1,1-dichloro2,2,2-trifluoroethane). ASHRAE Trans. 2:263283, 1989.
22
Miller, C. A., and P. Neogi. Interfacial Phenomena. New York.
Marcel Dekker, 1985.
23
Nelson, J. S., and S. Kimel. Safety of cryogen spray cooling during pulsed laser treatment of selected dermatoses. Lasers Surg.
Med. 26:23, 2000.
24
Nelson, J. S., T. E. Milner, B. Anvari, B. S. Tanenbaum, S.
Kimel, L. O. Svaasand, and S. L. Jacques. Dynamic epidermal cooling during pulsed laser treatment of port-wine stain.
A new methodology with preliminary clinical evaluation. Arch.
Dermatol. 131:695700, 1995.
25
Nelson, J. S., T. E. Milner, B. Anvari, B. S. Tanenbaum, L. O.
Svaasand, and S. Kimel. Dynamic epidermal cooling in conjunction with laser-induced photothermolysis of port wine stain
blood vessels. Lasers Surg. Med. 19:224229, 1996.
26
Niemz, M. H. Laser-Tissue Interactions. Berlin: SpringerVerlag, 1996.
27
Pikkula, B. M., J. H. Torres, J. W. Tunnell, and B. Anvari. Cryogen spray cooling: Effects of droplet size and spray density on
heat removal. Lasers Surg. Med. 28:103112, 2001.
28
Pikkula, B. M., J. W. Tunnell, and B. Anvari. Methodology for
characterizing heat removal mechanism in human skin during
cryogen spray cooling. Ann. Biomed. Eng. 31:493504, 2003.
29
Rayleigh, O. M. On the theory of the capillary tube. Proc. R.
Soc. Lond. Ser. A 92:184195, 1915.
30
Ross, E. V., F. P. Sajben, J. Hsia, D. Barnette, C. H. Hiller, and
J. R. McKinlay. Nonablative skin remodeling: Selective dermal
heating with a mid-infrared laser and contact cooling combination. Lasers Surg. Med. 26:186195, 2000.
31
Stolz, G., Jr. Numerical solutions to an inverse problem of heat
conduction for simple shapes. J. Heat Transfer 82:2026, 1960.
32
Taler, J. Theory of transient experimental techniques for surface
heat transfer. Int. J. Heat Mass Transfer 39:37333748, 1996.
33
Torres, J. H., B. Anvari, S. B. Tanenbaum, T. E. Milner, J. C.
Yu, and J. S. Nelson. Internal temperature measurements in response to cryogen spray cooling of skin. SPIE Proc. 3590:1119,
1999.

P1: KEE
Annals of Biomedical Engineering [AMBE]

1140
34

pp1282-ambe-490790

July 28, 2004

17:49

Style file version 14 Oct, 2003

PIKKULA et al.

Torres, J. H., J. S. Nelson, B. S. Tanenbaum, T. E. Milner, D. M.

Goodman, and B. Anvari. Estimation of internal skin temperatures in response to cryogen spray cooling: Implications for laser
therapy of port wine stains. IEEE J. Special Topics Quant. Elect.
5:10581066, 1999.
35
Torres, J. H., J. W. Tunnell, B. M. Pikkula, and B. Anvari. An
analysis of heat removal during cryogen spray cooling and effects of simultaneous airflow application. Lasers Surg. Med.
28:477486, 2001.
36
Tunnell, J. W., D. Chang, C. Johnston, J. H. Torres, C. W. Patrick,
M. J. Miller, S. L. Thomsen, and B. Anvari. In-vivo study of
epidermal protection by cryogen spray cooling during pulsed
laser irradiation at high radiant exposures. SPIE Proc. 4609:67
74, 2002.
37
Tunnell, J. W., J. S. Nelson, J. H. Torres, and B. Anvari. Epidermal protection with cryogen spray cooling during high fluence

pulsed dye laser irradiation: An ex vivo study. Lasers Surg. Med.

27:373383, 2000.
38
Tunnell, J. W., J. H. Torres, and B. Anvari. Methodology for
estimation of time-dependent surface heat flux due to cryogen
spray cooling. Ann. Biomed. Eng. 30:1933, 2002.
39
Valeria, B. C., C. C. Dierickx, W. A. Farinelli, D. L. Tai-Yuan,
M. Woraphong, and R. R. Anderson. Ruby laser hair removal:
Evaluation of long-term efficacy and side effects. Lasers Surg.
Med. 26:177185, 2000.
40
Verkruysse, W., B. Majaron, G. Aguilar, L. O. Svaasand, and J. S.
Nelson. Dynamics of cryogen deposition relative to heat extraction rate during cryogen spray cooling. SPIE Proc. 3907:3748,
2000.
41
Zenzie, H. H., G. B. Altshuler, M. Z. Smirnov, and R. R.
Anderson. Evaluation of cooling methods for laser dermatology.
Lasers Surg. Med. 26:130144, 2000.