S. KANG, S. KANG: CHARACTERISTICS OF THE TRADITIONAL KOREAN LIME PLASTER AFTER ...
479–489
CHARACTERISTICS OF THE TRADITIONAL KOREAN LIME
PLASTER AFTER AN ADDITION OF PERILLA OIL
ZNA^ILNOSTI TRADICIONALNEGA KOREJSKEGA APNENEGA
OMETA Z DODATKOM PERILOVEGA OLJA
Sanha Kang, Soyeong Kang
*
Research Division of Restoration Technology, National Research Institute of Cultural Heritage, Daejeon, Republic of Korea
Prejem rokopisa – received: 2022-03-18; sprejem za objavo – accepted for publication: 2022-06-13
doi:10.17222/mit.2022.528
Ancient Korean records indicate that a variety of organic materials (straw, hanji, sticky rice, oil, etc.) used to be added to lime
for different purposes, including the preservation and repair of monuments. In this study, aerial lime plaster with perilla oil was
produced. Before the sample preparation, a flow test was performed: lime was kneaded at various ratios of mixing water and
perilla oil to better understand the physical transformation of the material. After perilla oil was added, the fluidity of the lime
putty was decreased, allowing adjustments in the workability. Using the Uigwe as the reference, lime samples with perilla oil
were prepared. As even a small amount of perilla oil can cause color changes in the lime material, it is important to control the
amount of perilla oil added to the lime putty. Furthermore, the material became resistant to moisture. As freezing and thawing
due to the moisture content are the main causes of damage to lime materials, it was believed that the addition of perilla oil can
improve their performance. Neutralization tests, μ-CT imaging and an XRD analysis were performed to investigate the carbon-
ation of the lime materials. The test results revealed that when perilla oil was added to the material, the carbonation was de-
layed; furthermore, the strength of the material decreased initially but increased gradually as the curing progressed. This study
suggests that perilla oil can be added to lime to improve its freeze-thaw resistance, which will be helpful in the maintenance of
monuments.
Keywords: traditional Korean lime plaster, aerial lime, perilla oil
Starodavni korejski zapisi ka`ejo, da so se apnu dodajali razli~ni organski materiali (slama, hanji, lepljiv ri`, olje, itd.) z
razli~nimi nameni, ki vklju~ujejo tudi ohranjanje in sanacije spomenikov. V predstavljeni {tudiji so avtorji pripravili apneno
testo iz zra~nega apna z vsebnostjo perilovega olja. Pred pripravo vzorcev za preiskave, je bil opraviljen test razleza na stresalni
mizici, z razli~nimi dele`i vode in perila olja v apnenem testu. Test je pomagal razumeti fizikalno transformacijo materiala. Po
dodatku perilovega olja se je razlez apnene mase zmanj{al, kar omogo~a prilagajanje obdelovalnosti. Pri pripravi vzorcev
apnenega ometa z dodatkom perilovega olja je bila kot primer uporabiljena recepturo iz Uigwe. Vsebnost perilovega olja je bila
majhna, saj lahko dodatek tega olja hitro povzro~i spremembo barve apnenega testa. Dodatek olja je povzro~il odpornost
materiala proti vlagi. Ker so cikli zmrzovanja/tajanja pri pove~ani vlagi glavni vzrok po{kodb apnenih materialov, so avtorji
ocenili, da lahko dodatek perilovega olja izbolj{a zmrzlinsko odpornost materiala. Karbonatizacijo apnenih mas je bila preu~ena
z nevtralizacijskimi testi, μ-CT skeniranjem in XRD analizo. Rezultati so pokazali, da dodatek perilovega olja zakasni
karbonatizacijo, posledi~no pa je za~etna trdnost materiala manj{a. Trdnost materiala se nato pove~uje starostjo. [tudija je
pokazala, da se lahko zaradi izbolj{anja zmrzlinske odpornosti materialov z apnenim vezivom, me{anici doda perilovo olje, kar
je lahko bilo dobrodo{lo pri vzdr`evanju spomenikov.
Klju~ne besede: tradicionalni korejski apneni omet, zra~no apno, perilovo olje
1 INTRODUCTION
Lime has been used in various structures in Korea, in-
cluding buildings, fortresses, graves and mural painting.
During the Japanese colonial and modernization era,
however, the use of traditional lime in most of the manu-
facturing decreased and construction methods were dis-
continued. The use of lime also gradually decreased with
the introduction of modern materials such as Portland ce-
ment. A wide range of organic materials that used to be
added to both domestic and foreign limes to improve
their performance were identified. Typically, the research
on oil additives in western countries has mostly focused
on olive and linseed oils that are commonly used in
painting works.
1–5
The moisture absorption of lime
greatly decreases after an addition of oil, confirming that
oil has a waterproofing effect. However, it must be noted
that while the traditional Korean lime is aerial lime,
which reacts with atmospheric carbon dioxide and hard-
ens, most studies in western countries are focused on hy-
draulic lime. Interpreting the results of such a study
should therefore account for the differences in lime type.
The National Research Institute of Cultural Heritage
constructed the Namhansanseong Mock Yeojaing (Para-
pet) using aerial lime at the Namhansanseong Emer-
gency Palace in the mountains and monitored it for about
three years (Figure 1). It was found that the freeze-thaw
process of moisture repeatedly occurred during the win-
ter and played a significant role in the damage of the
constructed structure.
6
In such cases, cement (a modern
material) or waterproofing agents are added to increase
the freeze-thaw resistance of the materials. However,
these materials have not been evaluated for their stability,
and there is a possibility of secondary damage resulting
Materiali in tehnologije / Materials and technology 56 (2022) 5, 479–489 479
UDK 666.92:693.6:94(519) ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 56(5)479(2022)
*Corresponding author's e-mail:
soyeong.kang@korea.kr (Soyeong Kang)

from the incompatibility with the existing raw materials,
especially in the case of cement. As the need for environ-
mentally friendly and sustainable materials for monu-
ment restoration has increased, re-examining the useful-
ness of traditional materials has become necessary. In
this study, it was determined that oil additives, which are
natural products, can enhance the performance of lime,
so a literature review on the subject was conducted.
In the Uigwe records,
7
various oil additives were
identified, such as perilla oil, sesame oil, myoung oil and
tung oil. Perilla oil, which was mentioned the most, is of
particular interest. In particular, it is possible to infer in-
formation from the records of the 17
th
century,
8,9
such as
addition methods, number of applications and applica-
tion methods. In this study, perilla oil was added to aerial
lime and reproduced in the form of traditional Korean
lime plaster based on the contents of the records, and the
consequent changes in the properties of the lime materi-
als were investigated.
2 EXPERIMENTAL PART
2.1 Materials
The powder state of slaked lime (B company, Korea)
was used as the binder and an edible perilla oil (S com-
pany, Korea), commonly used for cooking in Korea, was
used as the additive. The perilla oil, used in this study,
was produced with the cold-pressed method.
2.2 Flow test
Prior to the creation of the samples, a flow test was
performed
10
and the results are in Table 1. Lime was
kneaded under various ratios of mixing water and perilla
oil to better understand the physical transformation of
the materials’ properties. The proportions of the mixing
water were (50, 60 and 70) w/% of the lime mass, while
those of perilla oil were (1, 3, 5 and 10) w/%. The mate-
rial could not be mixed with a 40 w/% proportion of mix-
ing water, not even with an addition of 10 w/% perilla
oil. The sample produced with 50 w/% of mixing water
had a very low water content and was destroyed during
the demolding process. Accordingly, it was determined
that the production of lime plaster samples would be fea-
sible if the proportion of mixing water was 60 w/% or
more. The flow test indicated that the lime putty that
contained 60 w/% or 70 w/% of mixing water exhibited
decreased fluidity as more perilla oil was added. In this
case, the larger the content of mixing water, the greater
was the decrease in fluidity. It is believed that perilla oil
has a greater effect on the lime putty for plastering than
on the lime for masonry, because the former needs a
higher proportion of mixing water. With the flow test, it
was confirmed that the workability of the lime putty
could be adjusted with the amount of perilla oil added,
and that perilla oil acted as a natural thickener.
Table 1: Flow test results for mixing water and perilla oil
Water
(w/%)
Perilla oil (w/%)
–1351 0
50 107.67 103.10 110.95 107.44 106.81
60 130.98 123.63 123.53 121.49 119.36
70 156.86 146.62 143.81 128.80 127.24
Average flow value (mm)
2.3 Sample preparation
The amount of mixing water was set based on the re-
sults of the flow test, and the proportions of water and
perilla oil were applied based on the lime weight. The
compositions of the samples used in the experiment and
applied curing conditions are shown in Table 2. The ex-
periment was designed to minimize the deviations
caused by the conditions other than the additive type,
such as lime type, slaking process and aggregate type.
Thus, a sample was prepared in the form of plaster with
no aggregates (such as sand). The details about the mix-
ing procedure were taken from the reference literature.
11
The lime putty was molded by filling it in a custom-
ized brass mold with dimensions of (20 × 20 × 20) mm
and (50 × 50 × 50) mm. In most cases, cube-shaped
specimens with a diameter of 20 mm were used, while
cube-shaped specimens with a diameter of 50 mm were
used for neutralization and freeze-thaw tests. Perilla oil
was added in two ways based on the descriptions from
S. KANG, S. KANG: CHARACTERISTICS OF THE TRADITIONAL KOREAN LIME PLASTER AFTER ...
480 Materiali in tehnologije / Materials and technology 56 (2022) 5, 479–489
Figure 1: Damage to Namhansanseong Mock Yeojang (Parapet) due to freeze-thaw: a) Namhansanseong Yeojang (Parapet) in winter (Gyeonggi
Cultural Foundation website), b) detailed photograph of the surface of Namhansanseong Mock Yeojaing (Parapet) damaged by freeze-thaw, c) ice
inside the surface crack of Namhansanseong Mock Yeojang (Parapet)

the records. The first method involves blending perilla
oil with lime and mixing water, and the second involves
applying perilla oil onto the surface of a sample. As rec-
ommended in the Uigwe, a hog bristle brush was used as
the application tool, and the application interval was 1 h.
The manufactured samples were dried at a constant
temperature and humidity (23 °C, 50±3% )a n dsepa-
rated from the mold on the fifth day after the manufac-
ture. On the seventh day, the samples were collected and
laid on an acrylic holder to be cured under two condi-
tions, curing and accelerated curing. Curing means that
the samples are cured with a 0.04 % concentration of
CO
2
. Accelerated curing (denoted by AC) means the
samples are cured with5%o fC O
2
in a chamber, to
check the tendency of long-term curing within a short
period of time as well as the curing conditions in an in-
door environment (Figure 2).
2.4 Analytical methods
To determine the state of the samples, the surfaces
were observed under 25× magnification using a portable
digital microscope, DG-3 (Scalar, Japan), and the CIE
value was obtained using a spectrophotometer
(CM-2600d, Konica Minolta, JPN). The standard light
source was set at D65, the viewing angle was 10° and the
analysis area was  16 mm. For measuring the chroma-
ticity, the average value of the measurements taken three
times at the same location was taken as the final value
for our data. The surface contact angle was measured to
confirm the moisture resistance. The sessile drop method
was applied using Phoenix 300 Touch (SEO), and the
contact angle was calculated using Tangent Line Method
2 in the Surfaceware software. In addition, a freeze-thaw
test was conducted. In this test, three samples, for which
curing was accelerated for 28 d, were used for each com-
position. The deposition freeze-thaw method was repeat-
edly applied as a cycle that included 18 h of immersion
under water,3hofcooling at –20 °C and3hofheating
at 50 °C. The state of the samples on each cycle was re-
corded and the ultrasound velocity was measured to
compare the changes over time.
A direct method was applied with a measurement dis-
tance of 50 mm (sample size) using a Pundit lab (Proceq,
Switzerland) device with a probe with a frequency of
54 kHz, voltage of 500 V and a sensitivity of 100×.
Three analysis methods were used to determine the de-
gree of carbonation. First, a neutralization test was per-
formed to confirm the discolored area by spraying
phenolphthalein on a sample cross-section. Additionally,
a μ-CT (XT H 225, Nikon, Japan) analysis was per-
formed to determine the carbonation area without de-
stroying the sample. The imaging conditions were
160 kV, 100 μA and 500 ms. As many as 3015 projection
images were collected. The information of the cross-sec-
tional image and internal structure of each axis was ob-
tained using the VG Studio software. The mineral com-
positions detected in each curing period were also
compared through an X-ray diffraction analysis. The
S. KANG, S. KANG: CHARACTERISTICS OF THE TRADITIONAL KOREAN LIME PLASTER AFTER ...
Materiali in tehnologije / Materials and technology 56 (2022) 5, 479–489 481
Figure 2: Sample preparation process
Table 2: Sample proportions and curing compositions
Sample Lime Water Perilla oil
L 1 0.6 -
Curing for
(7, 14, 28, 56 and 91) d
Accelerated curing (AC)
(7, 14, 28 and 56) d
M0.5 1 0.6 0.5 w/% mixed
M1 1 0.6 1 w/% mixed
M5 1 0.6 5 w/% mixed
B1 1 0.6 1 applied with a hog bristle brush
B5 1 0.6 5 applied with a hog bristle brush
B10 1 0.6 10 applied with a hog bristle brush
*Mixing proportions were based on w/%.

analysis was performed using Cu-K
 rays based on
X’Pert3 Powder (PANalytical, UK) with a 2 value of
33–65°, voltage of 40 kV and current of 30 mA.
To check the strength of the samples, both non-de-
structive and destructive methods were used. Regarding
the non-destructive test method, the Pundit lab device
was used in the same manner as for the previous sample
subjected to the freeze-thaw test. In this case, the ultra-
sound velocity measurement was directly performed at a
measurement distance of 20 mm (sample size), under a
voltage of 500 V, with a sensitivity of 100× using a
54 kHz probe. A UTM (AG-X plus, Shimadzu, Japan)
was used for the compression strength measurement with
a test rate of 0.5 mm/min and a load cell of 20 000 N.
3 RESULTS AND DISCUSSION
3.1 Physical properties
The longer the curing period, the lower was the
brightness of all samples (L*), regardless of the condi-
tions for adding perilla oil. In the case of L, there was a
difference of 1.18 between the brightness values of the
sample cured for 7 d and the sample that underwent ac-
celerated curing for 56 d without perilla oil, which
showed the smallest reduction. When perilla oil was
mixed in (M0.5, M1 and M5), differences of 1.80, 3.59,
and 3.09 were obtained, respectively. Conversely, when
perilla oil was applied (B1, B5 and B10), differences of
2.98, 4.29 and 6.68 were yielded, respectively. Thus, it
was confirmed that the range of brightness changes due
to prolonged curing was greater when perilla oil was ap-
plied. Additionally, samples with a relatively small
amount of perilla oil (M0.5, M1, B1 and B5) showed a
trend of decreasing redness and increasing yellowness.
In contrast, both redness and yellowness increased in the
samples containing a high amount of perilla oil (M5 and
B10).
Based on the color information for sample L (Ta-
ble 3), the color difference ( E) according to the curing
period was calculated for each sample. It was found that
the color difference increased with the amount of perilla
oil added and was greater when perilla oil was applied
with the brush compared to that when perilla oil was
mixed in. Particularly, sample B10 subjected to acceler-
ated curing for 56 d achieved a value of 21.51, the high-
est among all the samples.
Table 3: Color differences ( E)
Curing (d)
Accelerated curing (AC)
(d)
71 42 85 69 171 42 85 6
M0.5 1.41 1.20 1.38 1.52 1.19 3.94 3.83 1.43 2.00
M1 2.87 3.51 3.22 2.46 2.39 9.21 7.08 9.13 8.49
M5 12.48 14.18 14.35 13.99 10.39 14.72 17.68 17.12 17.09
B1 3.19 1.38 2.73 1.57 1.41 1.25 3.95 5.11 6.32
B5 2.47 2.50 5.01 5.37 4.07 8.01 9.51 10.48 7.84
B10 9.95 7.44 7.30 10.17 8.73 13.24 14.39 15.74 21.51
3.2 Water resistance
The moisture absorption properties of the samples
differed according to the conditions, under which perilla
oil was added (Table 4). Sample L was hydrophilic, that
is, water was absorbed immediately, before droplets were
formed on the surface. The wettability of the surface de-
creased for the sample that underwent AC for 28 d, and
the contact angle increased to 30.80°. For the sample that
underwent AC for 56 d, the contact angle was found to
be 40.33°. These changes are attributed to the increase in
the density. It is presumed that during carbonation the
matrix near the surface of a sample became quite dense.
Alternatively, the samples into which perilla oil was
mixed showed a contact angle of 90° or more in the early
stages of curing, and their surface showed hydro-
phobicity. It was confirmed that the samples to which the
oil was applied had lower surface wettability. In particu-
lar, samples B1 and B5 showed a high contact angle of
110° or more. The contact angle for B10, to which
perilla oil was applied for the maximum number of
times, was 89.22°, the lowest among all the samples with
added perilla oil.
As the curing time increased, a difference in the sur-
face contact angle was observed in the early stages of
curing. After curing for 91 d, the contact angles of M0.5,
M1 and M5, mixed with perilla oil, decreased to 90° or
less. In addition, the contact angle repeatedly increased
or decreased when the sample was exposed to acceler-
ated curing, but no specific pattern was observed. Mean-
while, B1, B5 and B10 with applied perilla oil showed a
slight decrease in contact angles, but their values re-
S. KANG, S. KANG: CHARACTERISTICS OF THE TRADITIONAL KOREAN LIME PLASTER AFTER ...
482 Materiali in tehnologije / Materials and technology 56 (2022) 5, 479–489
Table 4: Contact angle (°) measurement results by curing
Curing (d) Accelerated curing (AC) (d)
7 14 28 56 91 7 14 28 56
L––––––– 30.80 40.33
M0.5 104.43 95.32 102.81 88.45 84.26 83.63 110.39 101.44 85.96
M1 106.93 97.00 105.10 95.53 89.72 100.4 108.59 104.83 100.84
M5 105.32 98.14 90.17 95.51 84.71 106.79 106.56 104.57 99.33
B1 111.23 106.25 110.86 91.88 107.43 94.65 106.18 102.85 109.92
B5 115.62 100.79 95.82 95.80 97.37 109.33 105.82 105.19 105.43
B10 89.22 101.72 111.42 107.04 94.37 117.41 104.46 103.65 99.37

mained at 90° or more. Particularly, the contact angle of
B10 increased from 89.22° at the start of curing to
94.37°. The contact angle of the samples with perilla oil
applied continued to rise after the 7
th
day of AC. Among
the samples subjected to accelerated carbonation for
56 d, the final contact angle of the samples with perilla
oil (M0.5, M1, M5, B1, B5, B10) was higher than that of
sample L. Therefore, it can be determined that the sam-
ples to which perilla oil was applied have better surface
water resistance.
Each sample subjected to accelerated curing for 28 d
before the freeze-thaw test showed no identifiable physi-
cal damage. However, discoloration due to the addition
of perilla oil was noticed in B5, B10 and M5. Sample
M5 was especially dark in color and had a non-uniform
surface compared to the other samples. The state of the
sample was evaluated for the first time after the 3
rd
freeze-thaw cycle (Figure 3). When mixed with perilla
oil, samples (M0.5, M1, M5) were generally in good
condition. The samples under the condition in which
perilla oil was not added (L) or perilla oil was merely ap-
plied to the surface (B1, B5, B10) were physically dam-
aged. The samples in these cases were destroyed by
cracks, but detailed patterns of the cracks differed. Both
sample L and B1 had a large structural crack on the sur-
face. However, samples B5 and B10 had relatively small
and shallow cracks. Additionally, rather than having the
scaling peeled by frost, as was the case with samples L
and B1, the surfaces were separated into chunks. There-
fore, some freeze-thaw effects after the addition of
perilla oil were confirmed.
Damage was also observed on the surface of the sam-
ples with perilla oil mixed in after the 5
th
cycle of freez-
ing and thawing. Here, it was confirmed that sample M5,
which had the largest amount of perilla oil added,
showed more severe damage than M0.5 and M1. The re-
sults of the surface observation revealed that pulveriza-
tion resulted in cracks. At this point, similar to the sam-
ples to which perilla oil was applied, samples mixed with
perilla oil showed a damage pattern, indicating that the
surface was pulverized and peeled off, which differed
from the perilla oil-free sample L. The damage to all the
samples intensified during the 7
th
freeze-thaw cycle. As
the surfaces were powdered and destroyed, all the sam-
ples showed the same patterns, and their corners fell off.
Using a visual and portable microscope, the damage to
M5 was determined to be the most significant. In all the
samples, physical damage such as surface exfoliation,
powdering, falls and cracks increased after the 10
th
freeze-thaw cycle. In particular, sample M5 was on the
verge of destruction due to the cracks.
Table 5 illustrates the changes in the physical proper-
ties obtained with ultrasound velocity measurements for
different numbers of cycles. In the case of the samples
for which accelerated curing was performed for 28 d be-
fore the freeze-thaw test, the ultrasound velocity for B1
was 2088 m/s (the highest), while that for sample L was
2085 m/s, which was almost the same as for B1. In con-
trast, the samples with a relatively large amount of
perilla oil added showed an ultrasound velocity range of
1128–1567 m/s.
Table 5: Ultrasound velocity of the samples by cycle
Sample
Ultrasound velocity (m/s)
0 cycle 3
rd
cycle 5
th
cycle 7
th
cycle 10
th
cycle
L 2085 2728 – – –
M0.5 1514 1251 1313 1141 1073
M1 1567 1756 1725 1512 1070
M5 1488 1531 1535 1514 1360
B1 2088 1699 – – –
B5 1373 795 – – –
B10 1128 479 – – –
After the 3
rd
cycle of the freeze-thaw test, most of the
samples containing perilla oil had a reduced ultrasound
value, but M1 and M5 had slightly increased ultrasound
values. Here, sample L showed an increase of 643 m/s in
the ultrasound velocity despite substantial physical dam-
S. KANG, S. KANG: CHARACTERISTICS OF THE TRADITIONAL KOREAN LIME PLASTER AFTER ...
Materiali in tehnologije / Materials and technology 56 (2022) 5, 479–489 483
Figure 3: Freeze-thaw test results after 3
rd
cycle

age. It is difficult to interpret this increase as an increase
in the physical properties of the sample. The ultrasound
velocity of B1, B5 and B10 decreased by 389 m/s,
578 m/s and 649 m/s, respectively. As described above,
both the samples to which perilla oil was not added and
the samples to which it was scarcely applied were signif-
icantly damaged by the 3
rd
cycle. Consequently, it was
determined that further measurements of the ultrasound
velocity would be difficult.
The results of the ultrasound velocity measurements
of the samples mixed with perilla oil after the 5
th
cycles
of freeze-thaw showed that the average ultrasound veloc-
ity of M0.5 and M5 increased, whereas that of M1 de-
creased. In spite of the differences in the conditions,
there is very little change in the physical properties evi-
denced by the deviation in the ultrasound velocity. After
the 7
th
cycles of freeze-thaw, the ultrasound velocity
started to differ. First, the ultrasound velocity of M0.5
decreased by 173 m/s compared to the previous cycle,
while that of M1 and M5 decreased by 213 m/s and
21 m/s, respectively. Samples M0.5 and M1 exhibited
values considerably different from their previous cycle,
while M5 with the highest amount of mixed perilla oil
showed a smaller difference. All the samples exhibited a
decrease in the ultrasound velocity after the 10
th
freeze-thaw cycle. For M0.5, M1 and M5, the ultrasound
velocity decreased by 68 m/s, 442 m/s and 154 m/s, re-
spectively. Thus, we determined that M5 had the highest
ultrasound velocity. At this point, signs of physical dam-
age were discovered in the samples, so the freeze-thaw
experiment was ended.
3.3 Carbonation properties
The neutralization test is a traditional technique for
measuring the progress of carbonation in cement and
concrete. As lime also includes calcium oxide, a neutral-
ization test can be applied. Whena1%phenolphthalein
solution is sprayed on the cross-section after cutting the
center of the sample’s Z-axis, the color of the region con-
taining alkalinity changes depending on the pH of the
sample. Thus, the degree of the carbonation progress can
be determined by checking the discolored region (Fig-
ure 4).
Until the 56
th
day of indoor curing, no color change
appeared on the cross-section of L, but color differences
appeared outside the sample after the 91
st
day of indoor
curing. As the accelerated curing period increased, the
area with no color change increased, indicating that car-
bonation occurred. Unlike L, samples containing perilla
S. KANG, S. KANG: CHARACTERISTICS OF THE TRADITIONAL KOREAN LIME PLASTER AFTER ...
484 Materiali in tehnologije / Materials and technology 56 (2022) 5, 479–489
Figure 4: Neutralization test results (sample size of (50 × 50 × 50) mm)

oil failed to clearly show any discolored areas. De-
pending on the sample, there may be a difference in the
shade of discoloration, or a ring-shaped pattern may ap-
pear in a liesegang pattern.
12
As a result, the neutraliza-
tion test could not clearly identify the carbonated areas
of the lime plaster samples containing perilla oil.
Furthermore, it is possible to obtain a 3D voxel im-
age and a cross-sectional image without destroying the
samples. The 2D images from Figure 5 were acquired
with a μ-CT analysis where we observed the samples
along the Z-axis. These μ-CT images have different gray
values depending on the density of the materials. The ar-
eas with a low gray value (a relatively dark part) usually
have a lower density, which means that there are gaps in
the samples or lime materials that have not been carbon-
ated yet. However, a region with a high gray value (a rel-
atively bright part) denotes a dense part, such as a lime
material that has undergone carbonation. In this way, the
carbonation region can be estimated by measuring the
shade difference.
From the 28
th
day of the indoor curing onwards, the
carbonation of sample L can be seen progressing around
the outside shading. As the carbonation region gradually
expands throughout the curing period, no shade differ-
ence is detected on the cross-sections of the samples on
the 14
th
day of accelerated carbonation, and carbonation
might therefore be considered complete. The M0.5 cured
for 91 d has a slight shading difference at the center, but
the samples mixed with perilla oil have areas of ambigu-
ous shading. When perilla oil is applied, samples show a
more obvious shade difference than the samples with
mixed perilla oil. Nevertheless, the area where carbon-
ation has not yet been completed is wider than that of
sample L. The carbonation area of B5 is unclear on the
91
st
day of indoor curing, but it becomes distinct on the
7
th
day of accelerated curing, and there is no shade dif-
ference on the cross-section on the 14
th
day of acceler-
ated curing. A noticeable shade difference is also evident
on sample B10 from the 7
th
day of accelerated curing on-
wards. In contrast to the other conditions, the area where
carbonation has not progressed is wider. It is therefore
confirmed that the carbonation of B10 was occurring un-
til the 14
th
day of indoor curing and 28
th
day of acceler-
ated curing. According to the μ-CT analysis results, car-
bonation is delayed when increasing amounts of perilla
S. KANG, S. KANG: CHARACTERISTICS OF THE TRADITIONAL KOREAN LIME PLASTER AFTER ...
Materiali in tehnologije / Materials and technology 56 (2022) 5, 479–489 485
Figure 5: μ-CT analysis images (sample size of (20 × 20 × 20) mm)

oil are applied. In the case of mixed perilla oil, it is diffi-
cult to locate carbonation areas on the CT images.
The (20 × 20 × 20) mm samples used for the μ-CT
analysis were pulverized, and an X-ray diffraction analy-
sis was performed on them. The main minerals detected
in all the samples include portlandite, which is calcium
hydroxide (Ca(OH)
2
), and calcite, a calcium carbonate
(CaCO
3
). Calcium hydroxide is converted into calcium
carbonate in the process of carbonation by reacting with
carbon dioxide in the atmosphere. Therefore, it is possi-
ble to determine whether the material is carbonized
based on the changing diffraction pattern (Figure 6).
The portlandite peak was mostly observed for sample
Lonthe7
th
day of indoor curing. From the 14
th
day of
indoor curing onwards, the frequency and intensity of the
calcite peak began to increase. From the 91
st
day of in-
door curing onward, its diffraction peak became more
dominant than the portlandite peak. The longer the cur-
ing period, the sharper the peak of the calcite becomes,
while the portlandite peak tends to gradually flatten, in-
dicating that carbonation is in progress.
Sample M0.5 shows an increasing pattern of the cal-
cite peak after the 14
th
day of curing, and its diffraction
pattern based on the curing period is almost identical to
that of sample L. On the 7
th
day of accelerated carbon-
ation, the portlandite peaks became flatter, while the cal-
cite peaks became sharper. The portlandite peaks of M1
were mostly detected in the early stages of curing. The
frequency of the calcite peaks began to increase on the
14
th
day of indoor curing, but they were weaker than
those of M0.5, which was cured during the same time
period. As the curing period was prolonged, the portlan-
dite peak gradually became planar, and the diffraction
pattern with sharp calcite peaks is similar to that ob-
tained under the previous condition. In the early stages
of curing, M5 also exhibited the same diffraction pattern.
In this case, however, the diffraction peaks of calcite
were difficult to detect. That is, calcite peaks appeared
weakly on the 28
th
day of curing, but the portlandite
peaks were dominant until the 91
st
day of indoor curing.
A calcite peak was clearly visible on the 7
th
day of accel-
erated curing along with a decrease in the portlandite
peak. It can be assumed from this diffraction pattern that
carbonation reactions are delayed as the amount of
perilla oil added increases.
The diffraction pattern of sample B1, which does not
contain much perilla oil, is almost identical to that of
sample L. From the 14
th
day of indoor curing, the calcite
peak amplitude began to increase, and its diffraction
peak became more dominant than the portlandite peak by
the 91
st
day of indoor curing, similar to oil-free condi-
tions. The diffraction patterns of B5 were similar to
those of L and B1 during the early stages of curing.
However, on the 91
st
day of indoor curing, the calcite dif-
fraction peak was weaker than those of L and B1 during
the same period, and portlandite still dominated. The
portlandite peak became flat, and the calcite peak be-
came sharp only after accelerated curing was performed
for 7 days. From the 7
th
to the 14
th
day of indoor curing,
B10 exhibited the same diffraction pattern, with
S. KANG, S. KANG: CHARACTERISTICS OF THE TRADITIONAL KOREAN LIME PLASTER AFTER ...
486 Materiali in tehnologije / Materials and technology 56 (2022) 5, 479–489
Figure 6: XRD analysis results (P – portlandite, Ca – calcite)

portlandite being dominant. Under the conditions in
which a small amount of perilla oil was added (L, M0.5,
M1, B1), it was more difficult to confirm the calcite dif-
fraction peak. Calcite was only slightly detectable after
the 28
th
day of indoor curing. As late as the 91
st
day of
indoor curing, the portlandite peak was still dominant.
The calcite peak was dominant only from the 7
th
day of
accelerated curing onwards. In this way, the carbonation
delay due to perilla oil was also observed in the diffrac-
tion pattern of B10 to which oil was applied most fre-
quently.
3.4 Strength properties
The strength of the lime-plaster samples with respect
to the method of adding perilla oil and the curing period
was compared with the ultrasound velocity measure-
ments (non-destructive analysis) and compressive
strength measurements (destructive analysis). The ultra-
sound velocity measurement results revealed that the
strength of the samples to which perilla oil was not
added, or only a small amount was added, was relatively
high (Figure 7). The increase in the strength was quickly
detected for the samples into which the oil was mixed,
while the increase in the strength for the samples to
which perilla oil was only applied onto the surface was
delayed as the number of times the perilla oil was ap-
plied increased. Furthermore, the strength of the sample
without perilla oil was higher for about three months af-
ter the sample preparation. However, the longer the
curing period, the higher was the strength in the case of
included perilla oil. Due to the delay in the increase in
the strength, B10 required a long curing period before it
reached a physical strength comparable to that of the
samples under other conditions. The compressive
strength of the samples was determined through a de-
structive analysis of the samples, for which the ultra-
sound velocity was measured (Figure 8). When the com-
pressive strength was measured, the increase in the
strength of the samples without perilla oil, or with small
amounts of perilla oil, was high. The increase in the
strength was low when perilla oil was applied compared
to when a small amount, or no amount, of perilla oil was
added. Accordingly, it can be determined that the ultra-
sound velocity and compressive strength measurements
show similar trends.
The linear model confirmed the correlation between
the non-destructive analysis (ultrasound velocity mea-
surement) and destructive analysis (compressive strength
measurement), like in Figure 9. The coefficient of deter-
mination (R
2
) of the condition in which perilla oil was
not added was 0.78, representing the lowest value among
all the conditions. The conditions in which perilla oil
was mixed into the samples showed a correlation of 0.96
or more, and the conditions in which perilla oil was just
applied to the surface also showed a coefficient of deter-
S. KANG, S. KANG: CHARACTERISTICS OF THE TRADITIONAL KOREAN LIME PLASTER AFTER ...
Materiali in tehnologije / Materials and technology 56 (2022) 5, 479–489 487
Figure 7: Ultrasound velocity (m/s)
Figure 8: Compressive strength (N/mm
2
)

mination of 0.82–0.98. Thus, it was confirmed that the
two measurements used for the strength analysis of the
lime plaster samples containing perilla oil exhibited high
reproducibility.
4 CONCLUSIONS
In this study, perilla oil was added to aerial lime
based on the records of the Korean Uigwe, and the re-
sulting characteristics were verified. In the flow test, the
fluidity of the material (lime putty) decreased as the
amount of perilla oil added increased. Subsequently, the
fluidity decreased significantly as the moisture content in
the lime putty increased. The fluidity may have de-
creased because the absorption ratio of lime decreased
and perilla oil acted as a natural thickener. The thickener
suppresses material separation and can be used to adjust
the flow of a material with only a small amount added.
Therefore, an addition of perilla oil will most likely in-
crease workability. The brightness of the samples pro-
duced by adding perilla oil decreased as the curing pe-
riod increased; and the tendency of the redness or
yellowness to increase or decrease appeared to differ, de-
pending on the method of adding perilla oil to the sam-
ples (mixed vs. applied). The samples with added perilla
oil showed more color difference compared to those
without perilla oil. Hence, if perilla oil is added to lime
in a construction field, care should be taken to prevent
excessive discoloration. Above all, as previous studies
indicated, the moisture absorption properties of the aerial
lime plaster samples decreased with an addition of
perilla oil.
In addition, water resistance was confirmed via con-
tact-angle measurements. Water resistance was better
when perilla oil was applied than when it was mixed.
However, the freeze-thaw test revealed a difference in the
destruction pattern between the samples containing
perilla oil and those without it, whereas the samples
mixed with perilla oil were more durable. This is be-
cause the samples to which perilla oil was applied were
waterproof only on the surface where the oil was ap-
plied.
13,14
In addition, the carbonation of lime was de-
layed when perilla oil was added. Consequently, in the
early stages of curing, the samples without perilla oil
showed the greatest strength. As curing progressed, the
strength increased significantly. Here, the strength of the
samples was estimated by measuring ultrasound velocity
and compressive strength, and a high correlation between
those two measurement methods was confirmed for par-
ticular samples.
Overall, this study demonstrated that it is possible to
enhance the performance of traditional lime by adding
perilla oil to it, which leads to improved workability and
water resistance. The lime used for monuments can be
severely damaged during winter or in mountainous areas.
In this regard, lime with water resistance and freeze-thaw
resistance can be advantageous for the maintenance pur-
poses. Hence, by adding a small amount of perilla oil
during a construction using lime mortar, the damage
caused by freezing and thawing in winter can be mini-
mized. However, the study has some limitations. It is still
necessary to examine the perilla oil applying method,
which made the material highly waterproof only when
the construction surface was flat, such as wall plaster.
Further, perilla oil applied in the field should be studied
with regard to mortar samples mixed with aggregates,
such as sand, via additional experiments. It is also impor-
tant to examine the performance of lime when various
organic additives identified in the Uigwe are added in
combination with perilla oil.
Acknowledgment
This research is supported by the projects National
Research Institute of Cultural Heritage, Republic of Ko-
rea NRICH-2105-A38F-1 and NRICH-2205-A62F-1.
5 REFERENCES
1
C. Nunes, Z. Slí`ková, D. Køivánková, Lime-based mortars with lin-
seed oil: sodium chloride resistance assessment and characterization
of the degraded material, Periodico di Mineralogia, 82 (2013)3 ,
411–427, doi:10.2451/2013PM0024
2
C. Nunes, Z. Slí`ková, Hydrophobic lime based mortars with linseed
oil: Characterization and durability assessment, Cement and Con-
crete Research, 61–62 (2014), 28–39, doi:10.1016/j.cemconres.
2014.03.011
3
C. Nunes, P. Mácová, D. Frankeová, R. [ev~ík, A. Viani, Influence
of linseed oil on the microstructure and composition of lime and
lime-metakaolin pastes after a long curing time, Construction and
S. KANG, S. KANG: CHARACTERISTICS OF THE TRADITIONAL KOREAN LIME PLASTER AFTER ...
488 Materiali in tehnologije / Materials and technology 56 (2022) 5, 479–489
Figure 9: Strength correlation of lime plasters based on ultrasound velocity and compressive strength measurement: a) without perilla oil, b) with
perilla oil mixed, c) with perilla oil applied

Building Materials, 189 (2018) 20, 787–796, doi:10.1016/
j.conbuildmat.2018.09.054
4
L. Ventolà, M. Vendrell, P. Giraldez, L. Merino, Traditional organic
additives improve lime mortars: New old materials for restoration
and building natural stone fabrics, Construction and Building Mate-
rials, 25 (2011) 8, 3313–3318, doi:10.1016/j.conbuildmat.2011.
03.020
5
P. Zhao, M. D. Jackson, Y. Zhang, G. Li, P. J. M. Monteiro, L. Yang,
Material characteristics of ancient Chinese lime binder and experi-
mental reproductions with organic admixtures, Construction and
Building Materials, 84 (2015) 1, 477–488, doi:10.1016/
j.conbuildmat.2015.03.065
6
National Research Institute of Cultural Heritage, Trial applications of
traditional lime mortar, 2020, 6–327
7
Royal Construction Projects (Repairs) of Oegyujanggak Uigwe: The
Royal Protocols of the Joseon Dynasty
8
Uigwe for Renovating the Changdeokgung Palace and
Changgyeonggung Palace, 1652, 177, 223
9
Uigwe for Installing the Royal Coffin Hall and Spirit Hall of King
Hyojong, 1659, 1091
10
KS L 5111: Flow table for use in tests of hydraulic cement, 2017
11
KS L 5109: Practice for mechanical mixing of hydraulic cement
pastes and mortars of plastic consistency, 2017
12
C. Rodriguez-Navarro, O. Cazalla, E. Sebastian, Liesegang pattern
development in carbonating traditional lime mortars, Materials Sci-
ence, 458 (2002), 2261–2273, doi:10.1098/rspa.2002.0975
13
S. H. Kang, S. Y. Kang, A preliminary study on the effect of perilla
oil additive to traditional lime based plaster, Proceedings of the an-
nual conference of the Architectural Institute of Korea, 2020, 538
14
National Research Institute of Cultural Heritage, Value and Future of
Traditional Materials, 2020, 42–48
S. KANG, S. KANG: CHARACTERISTICS OF THE TRADITIONAL KOREAN LIME PLASTER AFTER ...
Materiali in tehnologije / Materials and technology 56 (2022) 5, 479–489 489