Journal of Petroleum Science and Engineering 142 (2016) 13–20

Contents lists available at ScienceDirect

Journal of Petroleum Science and Engineering

journal homepage: www.elsevier.com/locate/petrol

Water adsorption on kaolinite and illite after polyamine adsorption

Shifeng Zhang a, James J. Sheng b, Zhengsong Qiu a,n
a
School of Petroleum Engineering, China University of Petroleum (Huadong), Qingdao, 266580, China
b
School of Petroleum Engineering, Texas Tech University, TX, United States

art ic l e i nf o a b s t r a c t

Article history: Polyamine (PA) is a novel inhibitor applied to water-based drilling fluid in order to reduce clay hydration.
Received 31 October 2015 In this paper, water adsorption behavior on kaolinite and illite after PA adsorption (PA-kaolinite, PA-illite)
Received in revised form was characterized by X-ray diffraction (XRD), diffuse reflectance infrared Fourier transform spectroscopy
16 January 2016
(DRIFT), and adsorption isotherms tests using water, benzene, and heptane. After PA adsorption, water
Accepted 29 January 2016
adsorption on kaolinite was enhanced while water adsorption on illite was reduced. Water vapor ad-
Available online 30 January 2016
sorption affinity on kaolinite in an entire range was significantly increased due to H-bonds between the
Keywords: ammonium groups of PA and water molecules. As a result, kaolinite surface tension was increased. Water
Kaolinite and illite hydration adsorption affinity on illite in the low relative humidity (RH) regimes was decreased after PA adsorption
Surface tension analysis
and illite surface tension decreased because fewer water molecules could adsorb on tetrahedral (SiO4) of
Water adsorption
illite. Therefore, PA can be used as inhibitor while drilling shale formations mainly composed of illite, not
Polyamine adsorption
kaolinite.
& 2016 Elsevier B.V. All rights reserved.

1. Introduction containing low molar masses of polyether amine compounds, as

proposed by Klein et al. (2005), Patel et al. (2001), and Qu et al.
During drilling, shale accounts for approximately 75% of the (2009). Extensive research (Zhong et al., 2011; Patel et al., 2007;
drilled footage and contributes to approximately 90% of the well- Wang et al., 2011; Cui and van Jeroen, 2010) has been conducted
bore stability problems encountered, such as hole collapses, tight concerning water adsorption on Montmonrillonite (MMT) after PA
holes, stuck pipes, poor hole cleaning, hole enlargement, plastic intercalation. However, water adsorption behavior of kaolinite and
flow, fracturing, and loss of well control, as reported by Steiger and illite after PA adsorption has not been discussed yet and whether
Leung (1992) and Zhong et al. (2011). As reported by Lal (1999), or not PA can dehydrate kaolinite and illite remains unclear. As
the amount and type of clay minerals in shale decide the affinity of reported by Šolc et al. (2011), kaolinite is a dioctahedral layered
shale for water and are directly related to wellbore stability. A aluminosilicate of 1:1 type, therefore, each layer has two different
critical factor in designing drilling fluid is the ability to dehydrate surfaces parallel to the (001) plane. First one, tetrahedral, is
clay minerals to prevent wellbore stability problems while drilling formed from basal oxygen atoms of the tetrahedral (SiO4) sheet,
in shale formations (Rojas et al., 2006; Morton et al., 2005). In while the second one, octahedral, is formed from surface hydroxyl
recent decades, various drilling fluid additives have been pro- groups of the octahedral (AlO6) sheet. As reported by Benco et al.
posed, including relatively high concentrations of inorganic salts, (2001), these hydroxyl groups of octahedral surface are very
organic salts, and many types of polymeric additives, as reported flexible and allow effective formation of hydrogen bonds with the
by Bruton and Mclaurine (1993), Rosa et al. (2005), Caenn and polar molecules. They have amphoteric character and are able to
Chillingar (1996), and Souza et al. (2010). It has been reported that act as proton donors and/or acceptors, respectively (Tunega et al.,
poly (ethyleneoxide) (PEO) polymers, PEO-PPO-PEO copolymers 2004). For kaolinite octahedral surface, water molecule can only
can adsorb strongly to clay surfaces by hydrogen bonding with form weak hydrogen bonds with the basal oxygen atoms sur-
silanol groups on the clay surface (Burchill et al., 1983; Nelson and rounding the ditrigonal hole. As reported by Cheng and Sun
(2012), illite sheets consist of an octahedral aluminum layer
Cosgrove, 2005) to decrease hydration of minerals. Nevertheless,
sandwiched between two silica layers. In each unit cell, the K þ ion
as Young and Iskander (2006) concluded, these approaches have
at the surface balances the substitution of Si with Al and the po-
exhibited various limitations and remained somewhat un-
sition of the K þ ion is independent of the specific Si atom being
successful with regard to the dehydration of clay minerals. A re-
substituted. For water molecules on illite surface at low RH, two
cent advance was the introduction of water-based drilling fluids
configurations exist (Malani and Ayappa, 2009). In one config-
uration, two hydrogen atoms interact with the surface oxygen to
n
Corresponding author. form two hydrogen bonds, whereas in the other configuration,

http://dx.doi.org/10.1016/j.petrol.2016.01.040
0920-4105/& 2016 Elsevier B.V. All rights reserved.
14 S. Zhang et al. / Journal of Petroleum Science and Engineering 142 (2016) 13–20

only one hydrogen atom of water interacts with the surface oxy- potential of the redispersed clay mineral paste was measured with
gen, the other forms hydrogen bonds with other water molecules a Zetasizer 3000 zeta potential analyzer (Malvern Instruments
laterally. This configuration also facilitates the hydration of K þ . Ltd.; UK). After that, the clay mineral paste was dried, ground and
With RH increased, multilayers of water molecules adsorbed on sieved to generate a powder sample. Before each test, the desig-
illite. In this study, XRD analyses were used to study PA adsorption nated sample was kept at 100 °C for 5 h until the weight remained
or water adsorption in the interlayers of kaolinite, and illite. Water, constant. The carbon content in the dried clay-polyamine com-
benzene, and heptane vapor adsorption on clay minerals was also posite was analyzed using a Euro EA3000 elemental analyzer
tested, and the associated surface tension was analyzed using the (Euro Vector S.P.A.; Italy). The adsorbed quantity of PA or PEG400
VCG theory developed by Van Oss et al. (1988) to study water-clay was calculated based on the carbon content in the clay-polymer
minerals thermodynamical interaction. The Langmuir, Freundlich, composite versus the pure polymer. The data were corrected using
and BET adsorption models were used to fit the experimental data the carbon content obtained from the untreated clay sample.
in order to characterize water adsorption on clay minerals at dif-
ferent RH, while DFTIR spectroscopy was used to gather bonding 2.4. XRD measurement
information regarding the water molecules adsorbed on clay mi-
nerals. For comparison, water adsorption behaviors of kaolinite The clay and clay-polymer composite solutions were washed
and illite after polyethylene glycol (PEG400) adsorption were also twice and centrifuged with 5000g centrifugal force for 30 min.
discussed. Following this, the pastes produced were applied liberally and
evenly to glass slides that were subsequently placed in desiccators
with a saturated K2SO4 solution at  98% relative humidity (RH)
2. Experimental and silica gel at  0 RH for one month. XRD analysis was used to
measure the d-spacings of clay composites at room temperature
2.1. Clay and clay suspensions with an X’Pert-Pro MPD of PANalytical B.V. using a source of Cu-Kα
radiation equipped with solid-state detector and operating at
Well-crystallized kaolinite was obtained from clay repository of 40 KV and 40 mA.
Maoming Kaolin Science and Technology CO. Ltd in China. The
o2 μm fraction was isolated from the bulk material by sedi- 2.5. DRIFT analysis
mentation for 8 h under ambient temperature. The kaolinite paste
was then washed with deionized water 3 times until the con- The DRIFT spectroscopy of the treated samples recorded be-
ductivity of the supernatant equaled that of deionized water. After tween 1000 and 2000 cm  1 and summations over 30 scans was
washing, the kaolinite was dispersed and stored as a stock solu- obtained using a diffuse reflectance accessory ‘Collector’ from
tion. The solid content was measured after drying at 100 °C for 5 h Spectra-Tech with a NEXUS FT-IR Spectrometer (Thermo Nicolet
until the weight remained constant. Instrument Corporation; USA). The samples stored in the  98% RH
Well-crystallized illite was obtained from clay repository of and  0% RH desiccators were analyzed, respectively, at room
Xintai Nonmetal Material Co. Ltd. in China. The o2 μm fraction temperature without any KBr dilution. The materials of interest
was isolated from the bulk material by sedimentation for 8 h un- were poured loosely into a sample cup approximately 1 mm deep
der ambient temperature. The illite paste was then washed with and 3 mm in diameter and randomly oriented.
deionized water 3 times until the conductivity of the supernatant
equaled that of deionized water. After washing, the illite was 2.6. Water, benzene, and heptane vapor adsorption isotherm
dispersed and stored as a stock solution. The solid content was measurements
measured after drying at 100 °C for 5 h until the weight remained
constant. The clay composites were dried at 100 °C for 5 h before being
ground and sieved to generate the powder samples used for the
2.2. Chemicals analytical tests. The water, benzene, and heptane vapor adsorption
isotherms in the relative pressure range of 0–1 were measured
Polyamines (PA) (developed by the drilling fluid research team using an IGA-100B intelligent gravimetric analyzer (Hiden Iso-
in China University of Petroleum) contains primary amino groups chema Instrument Corporation; Warrington, UK). The results re-
attached to the ends of a polyether backbone. The block copolymer present the average of three adsorption experiments.
backbone is usually composed of propyleneoxide, ethyleneoxide,
or a propylene oxide/ethylene oxide mixture. The PA molarmass 2.7. Bulk specific surface area (SSA) by N2 gas adsorption
was approximately 400 g/mol and had a typical chemical formula
of CH3–[OCH2CH2]x–[OCH2 (CH3) CH]y–NH2. PEG400 was pur- N2 gas adsorption was performed at 77 K in duplicate using an
chased from Sinapharm Chemical Reagent Co. Ltd. ASAP 2010 device (Micromeritics Instrument Corporation; USA).
The SSA was calculated according to the BET equation from seven
2.3. Adsorption isotherms of PA, PEG400 on kaolinite, and illite data points recorded in the relative pressure range of 0.05–0.2.
Prior to the measurements, the samples were outgassed for at
Aqueous PA and PEG400 stock solutions were prepared and least 24 h at 105 °C under vacuum to remove adsorbed water.
mixed with deionized water and the clay mineral suspension to
prepare 0.2 wt% clay mineral suspensions with various con-
centrations of PA or PEG400. The samples were sealed in glass 3. Results and discussion
vials and rotated end-over-end for 24 h to allow the adsorption
process to reach equilibrium. The samples were centrifuged at 3.1. Adsorption models and the analysis of PA, PEG400 adsorption on
5000g centrifugal force for 30 min. The supernatant was collected kaolinite, and illite
and pH of the supernatant was measured using a PHS-3C PH Meter
(Shanghai REX Instrument Factory; China) at room temperature. The Langmuir and Freundlich adsorption models were used, in
The clay mineral paste was collected and redispersed in deionized the same manner as described in the work of Dada et al. (2012), to
water. This washing procedure was repeated twice. Then, zeta fit the experimental data (Table 1). Langmuir adsorption model
 S. Zhang et al. / Journal of Petroleum Science and Engineering 142 (2016) 13–20 15

Table 1
Fitting results of the PA, PEG400 adsorption isotherms on kaolinite, and illite.

Adsorbent Clay mineral Langmuir model constants Frendlich model constants

Q0 (mmol/g) KL R2 ΔG (KJ/mol) Kf n R2

PA Kaolinite 0.00040 261.62 0.993  13.79 0.000031 4.39 0.867

Illite 0.00042 135.15 0.995  8.08 0.00077 4.13 0.898
PEG400 Kaolinite 0.00014 176.39 0.980  12.82 0.0015 4.02 0.716
Illite 0.00023 57.94 0.988  6.69 0.00059 2.41 0.912

was described by Dada et al. (2012) in its linear form in Eq. (1). Q e is the amount of adsorbate adsorbed on adsorbent at equili-
brium (mmol/g).
1 1 1
= + Langmuir and Frendlich model were used to fit data for PA and
qe Q0 Q 0KLCe (1) PEG400 adsorption on kaolinite or illite.
As is shown in Table 1, the adsorption isotherms of PA on
where Ce is the equilibrium concentration of adsorbate (mmol/L),
kaolinite and illite fit the Langmuir model better. ΔG for PA ad-
qe is the amount of adsorbate adsorbed on adsorbent at equili-
sorption on kaolinite was greater than that on illite, indicating
brium (mmol/g), KL is Langmuir isotherm constant, Q 0 is max-
more energetically favorable adsorption on kaolinite. According to
imum monolayer coverage capacity (mmol/g). The free energy
the assumptions of the Langmuir adsorption model (Dada et al.,
change of the adsorption reaction is given by Eq. (2) (Özer et al.,
2012), it can be known only a monolayer of PA adsorbed on kao-
2004).
linite or illite. As is shown in Fig. 1, with PA addition, pH of clay
ΔG = − RT ln KL (2) mineral solution increased a lot and the solution was highly basic.
With PA concentration increasing, more PA adsorbed on illite or
where ΔG is standard free energy change (KJ/mol), R is the universal kaolinite and pH of the solution exhibited no obvious variation,
gas constant (8.314 J mol  1K  1), and T the absolute temperature (K). suggesting no obvious release of OH  or H þ during adsorption
The free energy change indicates the degree of spontaneity of the process. With PEG400 concentration increasing, more PEG400
adsorption process and the higher negative value reflects a more adsorbed on illite or kaolinite and pH of the solution increased at
energetically favorable adsorption (Özer et al., 2004). low concentration, which may be caused by OH  release during
Freundlich adsorption model was described by Dada et al. the process of adsorption and with greater PEG400 concentration,
(2012) in its linear form in Eq. (3). pH of the solution became similar to that of clay mineral solution.
1 According to Fig. 1, no obvious zeta potential change of illite or
log Q e = log Kf + log Ce kaolinite was observed after PA or PEG400 adsorption, suggesting
n (3)
illite or kaolinite charge changed little and no electrostatic inter-
where Kf is Freundlich constant (mmol/g), n is adsorption in- action was present in the adsorption process. As reported by Xu
tensity, Ce is the equilibrium concentration of adsorbate (mmol/L), et al. (2013), PA NH2/NH3 þ head groups can form hydrogen bonds

Fig. 1. Adsorption isotherms of PA, PEG400 on kaolinite, and illite and pH, zeta potential variation during adsorption process.
16 S. Zhang et al. / Journal of Petroleum Science and Engineering 142 (2016) 13–20

to oxygen atoms of the tetrahedral (SiO4) sheet and PA PPO/PEO not intercalate into the interlayers of kaolinite or illite. As seen in
backbone can adsorb strongly to clay surfaces by hydrogen Fig. 2, when p/p0 increased from 0% to 98%, the interlayer
bonding with silanol groups on the clay surface. As can be inferred d-spacing for PA-kaolinite, PEG-kaolinite, PA-illite, or PEG-illite
that PA adsorbed on the tetrahedral (SiO4) sheet of illite or kaoli- exhibited no increase, suggesting little water adsorption in inter-
nite through hydrogen bonds between PA NH2/NH3 þ and tetra- layers of PA-kaolinite, PEG-kaolinite, PA-illite, PEG-illite and PA,
hedral sheet O atoms, or hydrogen bonds between PA O atoms and and PEG400 adsorption caused no change to the water adsorption
silanol groups. As reported by Greenwell et al. (2005), Na þ cations in interlayers of kaolinite or illite.
in MMT can be closely coordinated by the PA O, N atoms, water
molecules, and the clay sheet. Similar to Na þ , K þ in illite may also 3.3. Vapor adsorption models and fitting of water vapor adsorption
be coordinated in the same manner. Kaolinite octahedral surface isotherms of clay composites for water adsorption mechanism ana-
hydroxyl groups can act as proton acceptor or proton donator lysis at different RH
(Tunega et al., 2004). As a result, PA can adsorb onto octahedral
surface of kaolinite by hydrogen bonds between PA NH2/NH3 þ and This paper reports on the application of three models to gen-
O(H) of kaolinite octahedral surface or between PA O atoms and erate isotherms for vapor adsorption on clay composites. The lin-
kaolinite surface hydroxyl group. ear form of the Langmuir (Shang et al., 1995) adsorption model is
The adsorption isotherms of PEG400 on kaolinite, and illite fit described by Eq. (4).
Langmuir model better, indicating only a monolayer of PEG400 p 1 1
adsorbed on kaolinite or illite. ΔG for PEG400 adsorption on = p+
V Vm VmK (4)
kaolinite was greater than that on illite, indicating more en-
ergetically favorable adsorption on kaolinite. where p is the water vapor pressure, V is the volume of water
Similar to interaction between ethanol (Cheng and Sun, 2012) adsorbed by the clay determined by multiplying the mass ratio of
or PEO polymers (Burchill et al., 1983) and clay minerals, H adsorbed water by the mass of the clay sample and converting the
(O) atom of PEG400 can form a hydrogen bond with the surface mass, Vm is the volume (cm3) of water corresponding to one
oxygen of tetrahedral (SiO4) sheet of illite or kaolinite and O complete monolayer on the surface of the clay, and K is the ad-
(H) atom can interact with K þ in illite. PEG400 can adsorb onto sorption equilibrium constant.
octahedral surface of kaolinite surface by hydrogen bonding be- The linear form of the BET model (Hatch et al., 2012) is de-
tween PEG400 (O)H and kaolinite O atoms or between PEG400 O scribed by Eq. (5).
(H) and kaolinite hydroxyl group. p
ΔG for PA adsorption on kaolinite and illite was greater than po 1 (c − 1) ⎛ p ⎞
= + ⎜⎜ ⎟⎟
that for PEG400, suggesting more energetically favorable adsorp-
tion for PA.
( 1 − )V
p
po
Vmc Vmc ⎝ po ⎠
(5)

where p/ po is the relative humidity of water vapor expressed as a

3.2. Clay and clay composites d-spacing variation for interlayer ad- ratio, V is the volume of water adsorbed by the clay determined by
sorption analysis multiplying the mass ratio of adsorbed water by the mass of the
clay sample and converting the mass and Vm is the volume (cm3) of
D-spacing of PA-kaolinite or PA-illite following PA adsorption at water corresponding to one complete monolayer on the surface of
0 RH% showed that PA adsorption can cause no change of the clay. In the BET model, c is related to the enthalpy of adsorp-
d-spacing of kaolinite or illite. The same results were also observed tion for the first molecular layer of water and the enthalpy of
for PEG-kaolinite and PEG-illite. This suggests PA and PEG400 can adsorption for subsequent layers.

Fig. 2. Relationship between clay mineral composites d-spacing at P/P0 of  0% RH or 98% RH and the PA or PEG 400 adsorption capacity.
 S. Zhang et al. / Journal of Petroleum Science and Engineering 142 (2016) 13–20 17

The Freundlich isotherm is described by Hatch et al. (2012) in the experimental data in the nonlinear high RH% region (Fig. 3).
its linear form in Eq. (6). The Freundlich adsorption isotherm model has been regarded
as an empirical law designed to represent the isothermal varia-
⎛ x⎞ 1 ⎛ p⎞ tions in gaseous adsorption on a solid adsorbent under pressure. In
ln⎜ ⎟ = ln(k ) + ln⎜⎜ ⎟⎟
⎝ m⎠ n ⎝ po ⎠ (6) the work of Hatch et al. (2012), this model was successfully used to
describe the adsorption of water onto non-swelling clay minerals.
where x/m is the concentration of adsorbed water at equilibrium, After plotting the experimental data in the linear format of the
in g/gclay, p/ po is the relative pressure of the water vapor (i.e., the Freundlich model, there were two distinct linear regions for kao-
RH), and both k and n are Freundlich constants representing the linite and PA-kaolinite, indicating two different adsorption re-
adsorption capacity and adsorption strength, respectively. gimes. The low RH regimes ranged from 0% to 40.0% for kaolinite
PA-kaolinite and PEG-kaolinite adsorbed more water than while for PA-kaolinite, it occurred at 0–6.4% RH. In the high RH
kaolinite, indicating PA and PEG400 adsorption can enhance the regime, the linear region for kaolinite was from 40.0% to 94.6% and
water adsorption. PA-kaolinite adsorbed more water than PEG- it was from 6.4–94.6% for PA-kaolinite. The Langmuir and Freun-
kaolinite, suggesting PA adsorption increase more water adsorp- dlich models were used to fit the PEG-kaolinite water adsorption
tion on kaolinite than PEG400 adsorption. isotherms; in the entire RH regime, the isotherms are linear for
PA-illite and PEG-illite adsorbed less water than illite, indicat- both models. Therefore, one water adsorption mechanism is pre-
ing PA and PEG400 adsorption can decrease the water adsorption. sent on the PEG-kaolinite surface.
PA-illite adsorbed similar quantity of water to PEG-kaolinite, After plotting the experimental data in the linear format of the
suggesting PA adsorption has similar effect as PEG400 adsorption Freundlich model, there were two distinct linear regions for illite
to decrease water adsorption on illite. and PA-illite. The low RH regimes for illite ranged from 0% to 6.4%,
According to Zhao (2005), water adsorption isotherm of PEG- but for PA-illite, the linear regions occurred at 0–12.6% RH. In the
kaolinite was similar to that of Langmuir model and Langmuir high RH regime, the linear region for illite was from 6.4% to 94.8%
model was used to fit the isotherm in addition to Freundlich and and it was 12.4–94.6% for PA-illite. For PEG-illite, in the entire RH
BET models. The linear Freundlich and BET models were used to fit regime, the isotherm was linear for Freundlich model. Therefore,
the rest of water adsorption isotherms. The results were presented only one water adsorption mechanism is present on the PEG-illite
in Table 2 and Fig. 3. surface.
The water content curves versus RH% were fit to the linear form In general, an n value of Freundlich model above one indicates
of the BET model. The linear regions of the BET fit of the experi- a strong adsorption. The n values for kaolinite, PA-kaolinite, and
mental data for kaolinite, PA-kaolinite, PEG-kaolinite were 0– PEG-kaolinite (4.37, 50.00, and 2.95, respectively) were all above
40.0%, 0–40.0% and 0–41.0%, respectively. The linear regions of the one in the low RH% range, indicating that the clays have a strong
BET fit of the experimental data for illite, PA-illite, PEG-illite were affinity. Among those of kaolinite, PA-kaolinite, and PEG-kaolinite,
0–25.0%, 0–12.6% and 0–50.5%, respectively. According to Bru- n value for PA-kaolinite was the largest, suggesting PA adsorption
nauer et al. (1938), the major assumptions corresponding to the can significantly increase the water vapor adsorption on kaolinite
use of the BET adsorption model are that the adsorbent surface is in low RH% range. This may result from H bonding between water
composed of uniform, localized sites and that the adsorption at and N, O, or H atoms of PA adsorbed on tetrahedral kaolinite
one site does not affect the adsorption at a neighboring site. Due to surface. In the high RH% regimes, the n value drops below one for
the assumptions of the BET model, nonideal, porous clay minerals kaolinite, indicating H bonding between water molecules became
with nonuniform adsorption sites make it difficult to reproduce dominant in all the water configurations on kaolinite. However, for

Table 2
Fitting parameters of Langmuir, BET, and Freundlich models applied to water adsorption isotherms.

Clay composites Surface area (m2/g) Water adsorption RH range (%) R2 Langmuir model BET model constants Freundlich model
models constants constants

Vm K Vm c k n

Kaolinite 26.49 (N2/BET) BET 0–40.0 0.999 0.0010 328.88

3.73(H2O/BET-low Freundlich 0–40.0 0.971 0.0020 4.37
RH)
40.0–94.6 0.982 0.0037 0.94
PA-kaolinite 29.40 (N2/BET) BET 0–40.0 0.995 0.006 277.5
21.69(H2O/BET-low Freundlich 0–6.4 0.968 0.0058 50.00
RH)
6.4–94.6 0.991 0.013 3.04
PEG-kaolinite 25.22 (N2/BET) Langmuir 0–91.3 0.999 0.0058 0.0022
9.65(H2O/BET-low BET 0–41.0 0.972 0.0027 72.06
RH)
Freundlich 0–91.3 0.923 0.0055 2.95
Illite 12.30 (N2/BET) BET 0–25.0 0.994 0.0020 110.52
7.30(H2O/BET-low Freundlich 0–6.4 0.995 0.0022 11.49
RH)
6.4–94.7 0.980 0.011 1.20
PA-illite 13.59 (N2/BET) BET 0–12.6 0.973 0.00002 32.78
2.33(H2O/BET-low Freundlich 0–12.6 0.976 0.0012 3.69
RH)
12.6–94.6 0.980 0.0050 1.31
PEG-illite 11.83 (N2/BET) BET 0–50.5 0.994 0.0019 13.90
5.25(H2O/BET-low Freundlich 0–94.7 0.982 0.0062 1.47
RH)
18 S. Zhang et al. / Journal of Petroleum Science and Engineering 142 (2016) 13–20

Fig. 3. Comparison of clay mineral composites experimental water adsorption data and fitting data using Langmuir, BET and Freundlich model.

Fig. 4. DRIFT spectroscopy of kaolinite and illite before and after PA or PEG400 adsorption at different RHs.
 S. Zhang et al. / Journal of Petroleum Science and Engineering 142 (2016) 13–20 19

PA-kaolinite or PEG-kaolinite, the n values remained above one pe

(3.04 and 2.95, respectively), indicating PA and PEG400 adsorption

Δads G = − RT ∫p = 0 Γ ln p = − πe (7)
can increase water adsorption affinity on kaolinite in the high RH%
1 1
regimes due to H bonding between water and N, O, or H atoms of where R is the universal gas constant (8.314 J mol K ), T is the
PA or PEG400. The n value for illite, PA-illite, and PEG-illite is temperature (K), Γ is the amount adsorbed (mmol/m2), and p is
above one in the entire RH% regimes. Compared to n value (11.49) the vapor pressure (Pa). At saturation, p tends toward a maximum
for illite in low RH% regimes, the n values for PA-illite and PEG-illte noted as pe . The surface area of the clay mineral composites can be
(3.69 and 1.47, respectively) were much lower, indicating PA and determined using N2/BET analysis. After integrating the vapor
PEG400 adsorption can decrease water adsorption affinity on illite adsorption isotherm, for high energy solids, ΔimmG was calculated
in the low RH% regimes. This may be due to interaction between using Eq. (8).
PA or PEG400 and illite surface oxygen or K þ to decrease water
adsorption sites. In the high RH% regimes, n values for PA-illite, ΔimmG = γSL − γSo = − πe − γlv (8)
and PEG-illite (1.31, and 1.47, respectively) were similar to that of
where γSo
is the surface tension of the solid in equilibrium with its
illite (1.20), indicating similar water adsorption affinity on illite in
own vapor. The VCG model developed by Van Oss et al. (1988) is
the high RH% regimes after PA and PEG400 adsorption.
applied to actual solids, as was described in Eqs. (9) and (10).
3.4. DRIFT spectroscopy for water molecules H-bonding analysis ΔimmG = γlv − 2[(γs LW ⋅γl LW )1/2 + (γs +⋅γl−)1/2 + (γs−⋅γl +)1/2] (9)

Although the characteristic FTIR primary amine N–H stretching

modes (above 3000 cm  1) were observed in the amines without
clay, this region was overlain with complex O–H stretching ab- γso = γs LW + 2 γs +⋅γs− (10)
sorptions from the clay hydroxyl groups and the interlayer water LW
where γs is the Liftshitz–Vander Waals contribution; specifically,
after intercalation, revealing no useful information. As reported by
Greenwell et al. (2005), the primary amine N–H bond-bending the dispersive forces that exist in all cases, γs+ includes the Lewis
region (1640–1560 cm  1) proved to be more useful during ana- acid contributions, and γs− includes the Lewis base contributions.
lysis. The FTIR spectroscopy of PA in the absence of clay shows NH2 Surface tension of clay mineral composites was calculated and
bending mode absorptions at 1571 cm  1 (Fig. 4). Compared to that shown in Table 3. Compared with kaolinite, the surface tension of
of illite, the HOH bending vibrations of PA-illite and PEG-illite PA-kaolinite and PEG-kaolinite was higher, indicating that water
exhibited no obvious shift. The HOH bending of kaolinite at 98%RH adsorption ability for kaolinite increased after PA or PEG400 ad-
was at ∼1641 cm  1 in the DRIFT spectroscopy. Compared to kao- sorption. Among the surface tension components of PA-kaolinite,
linite, a new band near the HOH bending (∼1640 cm  1) mode γS þ increased to be 167.52 mJ/m2 compared with kaolinite
absorptions existed at ∼1616 cm  1, which should be assigned to (1.22 mJ/m2) due to H bonding between water and N, O, or H
be H-bonds between the ammonium groups and the interlayer atoms of PA while γS  is reduced to 89.64 mJ/m2 compared with
water. These results suggest an increase in H-bonding interactions kaolinite (132.32 mJ/m2) because the H bonding between water
commensurate with the creation of interactions between the
and hydroxyl groups of octahedral surface may be reduced. Among
ammonium H atoms and the siloxane surface oxygen atoms of the
the surface tension components of PEG-kaolinite, γS  increased to
aluminosilicate clay layer with additional H-bonds likely to form
419.29 mJ/m2 compared with kaolinite because H bonding be-
between the ammonium group and the interlayer water (Hatch
et al., 2012; Blachier et al., 2009). The HOH bending of PEG-kao- tween water and O or H atoms of PEG400 existed while γS 
linite exhibited no obvious shift. (4.88 mJ/m2) was similar to that of kaolinite. As a result, surface
tension of kaolinite increased due to PA or PEG400 adsorption.
3.5. Determination of clay composites surface tension for water ad- Compared with illite, the surface tension of PA-illite and PEG-
sorption thermodynamical mechanism analysis illite was lower, indicating that water adsorption ability for illite
decreased after PA and PEG400 adsorption. Among the surface
The surface area of clay minerals was analyzed (Table 2) ac- tension components of PA-illite, γS þ increased to be 112.22 mJ/m2
cording to BET equations using N2 and water adsorption data compared with that of illite (82.62 mJ/m2) due to H bonding be-
(Médout-Marére et al., 1998a). In comparison to the results using tween water and N, O, and H atoms of PA while γS  is reduced to
N2 adsorption data, kaolinite, PA-kaolinite, PEG-kaolinite, illite, PA- 13.34 mJ/m2 compared with that of illite (181.14 mJ/m2) due to
illite, and PEG-illite surface areas calculated using water adsorp- decreased number of water adsorption on the basal oxygen atom
tion data were lower due to their comparatively low water ad- of tetrahedral (SiO4). Among the surface tension components of
sorption capacity. According to Médout-Marére et al. (1998a), the PEG-illite, γS  increased to 239.05 mJ/m2 compared with illite
difference between specific surface areas obtained using N2 and
while γS þ was reduced to be 4.02 mJ/m2. Therefore, surface ten-
water adsorption data confirms that water molecules are present
sion of illite decreased due to PA adsorption or PEG adsorption.
in the interlayer region. No obvious variation of d-spacing for
these clay mineral composites was observed as RH changed from
Table 3
0% to 98%, suggesting few water molecules was present in the The surface tension components of illite and kaolinite before and after PA or
interlayer region of those samples. Consequently, the N2 adsorp- PEG400 adsorption (SD error).
tion data was used to determine the surface areas of clay mineral
γS0 (mJ/m2) γSLW (mJ/m2) γS þ (mJ/m2) γS  (mJ/m2)
composites.
According to Médout-Marére et al. (1998a,, 1998b), Salles et al. Kaolinite 57.65 7 1.28 32.2470.27 1.22 7 0.09 132.32 7 1.04
(2007), and Kwok and Neumann (2000), vapor adsorption mea- PA-kaolinite 289.117 5.55 44.03 70.29 167.52 7 3.05 89.647 2.08
surements enable calculation of the adsorption free enthalpy. The PEG-kaolinite 115.43 7 1.75 24.9870.10 4.88 7 0.08 419.29 7 1.56
Illite 328.85 7 10.16 84.18 71.29 82.62 7 2.04 181.147 6.08
film pressure πe can be determined using Gibbs’ equation, and the
PA-illite 151.55 7 4.63 74.17 71.56 112.22 7 1.48 13.34 7 0.87
amount of water adsorbed at the saturated vapor pressure can be
PEG-illite 147.88 7 4.01 45.78 71.12 10.05 7 0.48 239.05 7 2.21
predicted using Eq. (7).
20 S. Zhang et al. / Journal of Petroleum Science and Engineering 142 (2016) 13–20

4. Conclusions Greenwell, H.C., Harvey, M.J., Boulet, P., Bowden, A.A., Coveney, P.V., Whiting, A.,
2005. Interlayer structure and bonding in nonswelling primary amine inter-
calated clays. Macromolecules 38, 6189–6200.
PA adsorption can significantly increase water adsorption on Hatch, C.D., Wiese, J.S., Crane, C.C., Harris, K.J., Kloss, H.G., Baltrusaitis, J., 2012.
kaolinite. After PA adsorption, the water vapor adsorption affinity Water adsorption on clay minerals as a function of relative humidity: appli-
on kaolinite in entire RH range increased and γS þ of kaolinite cation of BET and Freundlich adsorption models. Langmuir 28, 1790–1803.
Klein, A.L., Aldea, C., Bruton, J.R., Dobbs, W.R., 2005. Field verification: invert-mud
surface tension rose due to H-bonds between the ammonium
performance from water-based mud in Gulf of Mexico Shelf. SPE Paper 84314,
groups of PA and water molecules. PEG400 adsorption can also SPE Annual Technical Conference and Exhibition, October, Denver.
increase the water adsorption on kaolinite, with more water vapor Kwok, D.Y., Neumann, A.W., 2000. Contact angle measurements and interpretation:
adsorption affinity on kaolinite in high RH% range and higher γS  wetting behavior and solid surface tensions for poly(alkylmethacrylate)poly-
mers. J. Adhes. Sci. Technol. 14, 719–743.
of kaolinite surface tension. PA or PEG400 adsorption can decrease Lal, M., 1999. Shale stability: drilling fluid interaction and shale strength. SPE Paper
water adsorption on illite. After PA adsorption, water adsorption 54356, Latin American and Caribbean Petroleum Engineering Conference, April
affinity on illite in the low RH% regimes was reduced and γS  of Caracas, Venezuela.
Médout-Marére, V., Malandrini, H., Zoungrana, T., Douillard, J.M., Partyka, S., 1998b.
surface tension of illite was decreased, for fewer water molecules Thermodynamic investigation of surface of minerals. J. Pet. Sci. Eng. 20,
were bound to the basal oxygen atom of tetrahedral (SiO4). Surface 223–231.
tension of illite was reduced after PEG400 adsorption due to lower Médout-Marére, V., Belarbi, H., Thomas, P., Morato, F., Giuntini, J.C., Douillard, J.M.,
γS þ because K þ in illite was coordinated by the O atoms of 1998a. Thermodynamic analysis of the immersion of swelling clay. J. Colloid
Interface Sci. 202, 139–148.
PEG400 and less bonding between water molecules and K þ Malani, A., Ayappa, K.G., 2009. Adsorption isotherms of water on mica: redis-
occurred. tribution and film growth. J. Phys. Chem. B 113, 1058–1067.
PA was more effective in the dehydration of MMT (Zhong et al., Morton, E.K., Bomar, B.B., Schiller, M.W., Gallet, J.D., Azar, S.J., Otto, M.J., Hansen, N.
A., Dye, W.M., Shoults, L., Daugereau K., 2005. Selection and evaluation criteria
2011; Wang et al., 2011; Cui and van Jeroen, 2010) and illite and for high-performance drilling fluids. SPE Paper 96342, SPE Annual Technical
exhibited many promising results for use in drilling shale forma- Conference and Exhibition, October, Texas.
tion mainly composed of MMT or illite. However, for shale for- Nelson, A., Cosgrove, T., 2005. Small-angle neutron scattering study of ad-
sorbedpluronictri-block copolymers on laponite. Langmuir 21, 9176–9182.
mation composed of large amount of kaolinite, PA may be not Patel, A., Stamatakis, S., Young, S., Friedheim, J., 2007. Advances in inhibitive water-
suitable to be used during drilling. based drilling fluids—can they replace oil-based muds. SPE Paper 106476, SPE
International Symposium on Oil field Chemistry, 28 February–2 March, Texas.
Patel, A.D., Stamatakis, E., Davis, E., 2001. Shale Hydration Inhibition Agent and
Method of Use. US 6247543.
Acknowledgments Qu, Y.Z., Lai, X.Q., Zou, L.F., Su, Y.N., 2009. Polyoxyalkyleneamine as shale inhibitor
in water-based drilling fluids. Appl. Clay Sci. 44, 265–268.
Rojas, J.C., Clark, D.E., Greene, B., Zhang, J., 2006. Optimized salinity delivers im-
We wish to thank the NSF of China (No. 41072094) for finan-
proved drilling performance. AADE 2006 Fluids Conference.
cially supporting this work. Rosa, R.S., Rosa, A., Farias, S.B., Garcia, M.H., Coelho, A.D.S., 2005. A new inhibitive
water-based fluid: a completely cationic system. SPE Paper 94523, SPE Latin
American and Caribbean Petroleum Engineering Conference, 20–23 June, Rio
de Janeiro, Brazil.
References Salles, F., Bildstein, O., Douillard, J.M., Jullien, M., Van Damme, H., 2007. Determi-
nation of the driving force for the hydration of the swelling clays from com-
Özer, A., Özer, D., Özer, A., 2004. The adsorption of copper (II) ions on to dehydrated putation of the hydration energy of the interlayer cations and the clay layer. J.
wheat bran (DWB): determination of the equilibrium and thermodynamic Phys. Chem. 111, 13170–13176.
parameters. Process Biochem. 39, 2183–2191. Shang, S.B., Horne, R.N., Ramey, H.J., 1995. Water vapor adsorption on geothermal
Šolc, R., Gerzabek, M.H., Lischka, H., Tunega, D., 2011. Wettability of kaolinite (001) reservoir rocks. Geothermics 24, 523–540.
surfaces-molecular dynamic study. Geoderma 169, 47–54. Souza, C.E.C.D., Lima, A.S., Nascimento, R.S.V., 2010. Hydrophobically modified poly
Benco, L., Tunega, D., Hafner, J., Lischka, H., 2001. Orientation of OH groups in (ethyleneglycol) as reactive clays inhibitor additive in water-based drilling,
kaolinite and dickite: Ab initio molecular dynamics study. J. Phys. Chem. B 105, fluids. J. Appl. Polym. Sci. 117, 857–864.
10812. Steiger, R.P., Leung, P.K., 1992. Quantitative determination of the mechanical
Blachier, C., Michot, L., Bihannic, I., Barrès, O., Jacquet, A., Mosquet, M., 2009. Ad- properties of shales. SPE Paper 18024, SPE Drilling Engineering, 7, 181–185.
sorption of polyamine on clay minerals. J. Colloid Interface Sci. 336, 599–606. Tunega, D., Gerzabek, M.H., Lischka, H., 2004. Ab initio molecular dynamics study of
Brunauer, S., Emmett, P.H., Teller, E., 1938. Adsorption of gases in multimolecular a monomolecular water layer on octahedral andtetrahedral kaolinite surfaces. J.
layers. J. Am. Chem. Soc. 60, 309–319. Phys. Chem. B 108, 5930–5936.
Bruton, J.R., Mclaurine, H.C., 1993. Modified poly-amino acid hydration suppressant Van Oss, C.J., Good, R.J., Chaudhury, M.K., 1988. Additive and nonadditive surface
proves successful in controlling reactive shales. SPE Paper 26327, Annual tension components and the interpretation of contact angles. Langmuir 4,
Technical Conference and Exhibition of the Society of Petroleum Engineers, 884–891.
Texas, October. Wang, L., Liu, S.Y., Wang, T., Sun, D.J., 2011. Effect of poly(oxypropylene) diamine
Burchill, S., Hall, P.L., Harrison, R., Hayes, M.H.B., Langford, J.I., Livingston, W.R., adsorption on hydration and dispersion of montmorillonite particles in aqu-
Smedley, R.J., Ross, D.K., Tuck, J.J., 1983. Smectite-polymer interactions in eous solution. Colloids Surf. A 381, 41–47.
aqueous systems. Clay Miner. 18, 373–397. Xu, Y., Liu, Y.L., He, D.D., Liu, G.S., 2013. Adsorption of cationic collectors and water
Caenn, R., Chillingar, G.V., 1996. Drilling fluids: state of the art. J. Pet. Sci. Eng. 14, on muscovite (001) surface: a molecular dynamics simulation study. Miner.
221–230. Eng. 53, 101–107.
Cheng, T., Sun, H., 2012. Adsorption of ethanol vapor on mica surface under dif- Young, S., Iskander, G.R., 2006. Drilling performance and environmental compliance
ferent relative humidities: a molecular simulation study. J. Phys. Chem. C 116, resolution of both with a unique water-based fluid. SPE/IADC Paper 103967,
16436–16446. SPE/IADC Indian Drilling Technology Conference and Exhibition, October,
Cui, Y.N., van Jeroen, S.D., 2010. Adsorption of polyetheramineson montmorillonite Mumbai.
at High pH. Langmuir 26, 17210–17217. Zhao, Z.G., 2005. Principles for Adsorption Application. Chemical Industry Press,
Dada, A.O., Olalekan, A.P., Olatunya, A.M., DADA, O., 2012. Langmuir, Freundlich, Beijing, China, pp. 184–194.
Temkin and Dubinin–Radushkevich isotherms studies of equilibrium sorption Zhong, H.Y., Qiu, Z.S., Huang, W.A., Cao, J., 2011. Shale inhibitive properties of
of Zn2 þ onto phosphoric acid modified rice husk. J. Appl. Chem. 3, 38–45. polyether diamine in water-based drilling fluid. J. Pet. Sci. Eng. 78, 510–515.