Gas-phase molecular recognition on functional monolayers immobilized on a highly sensitive quartz-crystal microbalance

Supramolecular Science 3 (1996) 16s169 0 1996 Elsevier Science Limited Printed in Great Britain. All rights reserved 0968-5677/96/$15.00

SO968-5677(96)00030-2

ELSEVIER

Gas-phase molecular recognition on functional monolayers immobilized on a highly sensitive quartz-crystal microbalance Yoshio Okahata, Kazunori Matsuura and Yasuhitto Ebara Department of Biomolecular Engineering, Midori-ku, Yokohama 226, Japan (Received 2 1 May 1996)

Tokyo Institute of Technology,

Nagatsuda,

Self-assembled monolayers of alkanethiols having functional groups (HS-(CH2)l-X, X =-H, --COOH, -CONHz, -NHz, thymine and adenine bases) were immobilized on the Au electrode of a quartz-crystal mucrobalance (QCM), and binding kinetics of guest molecules from the gas phase were studied from time courses of frequency decreases (mass increases) of the QCM. A highly sensitive, 63MHz overtone frequency of a conventional ~-MHZ at AT-cut QCM was developed to detect monolayer adsorption of small molecules. When acetic acid was employed as guest molecules, it adsorbed onto the --CONHz membrane as a Langmuir-type monolayer and adsorbed as multilayers onto the -COOH and -NH2 membranes, but hardly adsorbed onto the simple alkane membrane (-H membrane). When self-assembled monolayers bearing a thymine or adenine base as a terminal group were employed, selective binding, processes of complementary guest molecules were observed: 2-aminopyridine (an adenine model) and y-butyrolactam (a thymine model) selectively bound to the thyamine and adenine monolayers, respectively. 0 1996 Elsevier Science Limited (Keywords:quart~crystal microbalance;gas-phase molecularrecognition;acetic acid; nucleobase;self-assembledmonolayer)

INTRODUCTION Molecular recognition is an essential phenomenon in living systems, and many model systems have been proposed in host-guest chemistry and supramolecular chemistry. Numerous studies examined the premise that selective hydrogen bonds could be formed in hostguest combinations mainly in hydrophobic organic solvents’. Recently, several examples of molecular recognition based on hydrogen bonds at the air-water interface have been reported by using surface pressurearea (n-A) isotherms, FTIR and XPS techniques2. Molecular recognition on the membrane surface is interesting not only as a model of the biological membrane but also as a sensor device. Similar intermolecular recognition based on hydrogen bonds are also reported in adsorption onto a monolayer from the gas phase by using FTIR spectra3 and thermal desorption spectra4. Molecular recognition in the gas phase is the most simple and fundamental system devoid of solvent effects5Z6. A quartz-crystal microbalance (QCM) is one of several useful techniques to detect molecular adsorption onto a substrate by measuring in situ mass changes, both in gas4, ‘** and solution systems2(‘), ‘-’ I. In this paper, we report that gas-phase molecular recognition can be easily detected by using a highly sensitive QCM and a self-assembled monolayer. As a

simple example, a self-assembled monolayer of alkanethiolsi2 having a functional group at the oposition (HS-(CH2)lo-X, X =-H, -COOH, --CONHz, --NH3 is immobilized on a gold plate of a QCM and selective adsorption of acetic acid molecules is studied (see Figure I). As an exclusive molecular recognition in the gas phase, we prepared a selfassembled monolayer having thymine or adenine molecules at the terminal end, and complementary bindings of an adenine model (2-aminopyridine) or a thymine model (y-butyrolactam) are studied. It is interesting that complementary base-pair formation based on hydrogen bonding can be observed even in the gas phase. QCMs are sensitive mass-measuring devices because their resonance frequency decreases with increase of the mass on the electrode13. We have used the commercially available AT-cut, ~-MHZ QCM, in which the frequency decreases linearly by 1 Hz with an increase of ca. 1 ng of mass when the QCM is vibrated at its fundamental frequency’. ’ ’ . We have recently introduced a highly sensitive QCM, in which the ~-MHZ QCM is vibrated at seven-times overtone (63-MHz) of the fundamental frequency, in order to detect the adsorption of small molecules of acetic acid as a monolayer. The 63-MHz overtone QCM can detect ca. 0.1 ng of mass change through 1 Hz of frequency decrease.

SUPRAMOLECULAR

SCIENCE Volume 3 Numbers

l-3

1996

165

Molecular

recognition

on monolayers:

Y. Okahata et al.

rinsing with ethanol and Milli-Q water, the QCMs were dried and kept under a N2 atmosphere. The frequency was decreased by 35@-440Hz (Am = 4& 50ng) by immobilizing these monolayers. The mass changes shown in the parentheses were calculated from equation (2). The theoretical mass of the alkanethiol monolayers on the gold electrode was calculated to be 55 f 5 ng. These values indicate that the Au electrode was covered in 7&80% as a monolayer of alkanethiols. The monolayer-immobilized QCM was set in a flow cell (70cm3), in which a mixture of saturated vapour of guest molecules and N2 gas flowed at a rate of 2 1min-‘. The concentration of guest molecules in the flow cell was altered in the range of (l-70) x 10-j M by changing the mixture ratio of saturated vapour and dry N2 gas.

Gas Phase

X: -H -NH 2 -CCQH -CQNH 2

Gold Electrode RESULTS AND DISCUSSION -Quartz

Crystal

Figure 1 Schematic illustration of the adsorption behaviour of acetic acid molecules onto a self-assembled monolayer immobilized on a 63MHz overtone quartz-crystal microbalance (QCM) in the gas phase

EXPERIMENTAL QCM measurements The QCM employed is a commercially available 9MHz, AT-cut quartz (diameter 9mm), purchased from Sougo Pharmaceutical (Tokyo). The following equation has been established for AT-cut shear mode QCM13: AF=p

-2NF;

Am

(1)

A@&

where AF is the measured frequency shift (Hz), F0 the parent frequency of the QCM (9 x lo6 Hz), Am the mass change (g), A the electrode area (0.16cm2), N the overtone number (seventh), p4 the density of quartz (2.65 gcmM3) and mq the shear modulus of quartz (2.95 x 10” dyne cmp2). Calibration of the QCM used in our experiments by a polymer-coating method and an LB film transfer method gave the following equation7, 11*l4 Am = -(8.75 f 0.01) x 10m9AF

Selective adsorption of acetic acid onto the -CONHz membrane Figure 2 shows typical time series of the frequency changes of the 63-MHz QCM covered with the alkanethiol monolayer, responding to the exposure to acetic acid vapor (4.5 x 10m4M, P/P,,, = 0.54). When acetic acid is adsorbed as a Langmuir-type monolayer, the adsorption amount is calculated to be ca. 15 ng by assuming the molecular area of acetic acid is ca. 0.2nm2. Acetic acid was hardly adsorbed onto the simple alkane membrane (X = -H). In the case of the -CONH2 membrane, the adsorbed amount is saturated within 10s and reached a constant value (13 ng) that corresponded to 87% of the mass of the complete monolayer adsorption. When the -COOH and -NH2 membranes were employed, large adsorbed

-

0

-

5

- 10

tq-100 x

B - 15 ’ 5 - 20

2q -150

(2)

It is close to the theoretical equation calculated from equation (1) (Am = -7.35 x 10m9AF). Thus, when the ~-MHZ QCM was vibrated in the seventh overtone, the sensitivity for the mass change was increased about eight times, in comparison with the standard ~-MHZ vibration mode of the Q-4 [Am = -(1.05 f 0.01) x 10e9 AF]. The stability of the seventh-overtone frequency of the QCM was also examined: the standard deviation of the frequencies in air was 1.OHz for 1 h at 25°C. The QCM was immersed into 5mM ethanol solution of the alkanethiols (HS-(CH2)10-X) for 12 h. After

Figure 2 Typical time courses of frequency changes of the selfassembled monolayer-immobilized 63-MHz QCM responding to the exposure of acetic acid in the gas phase (4.5 x 1O-4M, P/Ps,, = 0.54) at 25°C

166

l-3 1996

SUPRAMOLECULAR

SCIENCE Volume 3 Number

- 25 _ 30

-350

0

500

1000

1500

Time /set

Molecular

[AcOH] 0

2

6

6

60

40

gl 30 \ 5

Y. Okahata et al.

on monolayers:

4.6, 5.6 and 1.9 for the -CONHz, --COOH, -NH2 and -H membranes, respectively. This indicates that acetic acid molecules tend to be adsorbed as a monolayer but not as multilayers onto the -CONH* membrane; on the contrary, they are easily adsorbed as multilayers onto the -COOH, -NH2 and the hydrophobic -H membranes. In summary, acetic acid molecules hardly bind to the hydrophobic surface of the -H membrane, and bind to the -NH2 membrane as multilayers with acid-base interactions. They show the strong monolayer adsorption onto the -CONH2 membrane using hydrogen bonding and tend to bind weakly to the -COOH membrane due to the formation of intra-membranous hydrogen bonds in the monolayer.

/ 1 O-’ M 4

recognition

20

10

Complementary binding between nucleic acid bases

0 0

0.2

0.4

0.6

0.8

1

P,/ Psat Adsorption isotherm:; of acetic acid molecules onto the functional monolayers immobilized on a 63-MHz QCM at 25°C

Figure 3

amounts (Am = 30-50 ng) were observed, indicating multilayer adsorption. Figure 3 shows the adsorption isotherms of acetic acid onto the various membranes on the QCM at 25°C. In the case of the -CONH2 membrane, the adsorbed amount (Am) reached a clonstant value of ca. long for concentrations below 4 x 10P4M. When the concentration increased above this point, the adsorption amount increased with increasing gas concentration. Thus, acetic acid can be adsorbed as a Langmuir-type monolayer (ca. 53% coverage) in the relatively low concentration range and then multilayer adsorption is observed in high concentrations for the -CONHz membrane. When the -COOH membrane was employed, the concentration range of the monolayer adsorption became very narrow and the multilayer adsorption was easily observed, above 2 x 1O-4M. In the case of the -NH:! membrane, the adsorbed amount increased linearly, which indicates acetic acid onto the -NH2 membrane as was adsorbed multilayers even at the low concentration. The adsorbed amount of acetic acid onto the hydrophobic -H membrane was small. Adsorption isotherms in Figure 3 were solved according to the BET equation’?

The QCMs immobilized with a self-assembled monolayer bearing a thymine 1 or adenine 2 at the terminal position was prepared, as were other membranes. As schematically shown in Figure 4, the binding behaviour of 2-aminopyridine (an adenine model), y-butyrolactam (a thymine model) and aniline having only one binding site were carried out in the gas phase. We used 2-aminopyridine and y-butyrolactam as models of adenine and thymine molecules, respectively, since nucleic acid bases are not volatile in air at normal temperature and pressure. Figure 5 shows typical time courses of frequency changes of a 63-MHz QCM immobilized with the thymine monolayer 1 responding to exposure of the same concentration (1.4 x 10e6 M) of 2-aminopyridine, y-butyrolactam and aniline in gas phase at 25°C. An adenine model of 2-aminopyridine bound

AQCM

Host Monolavers

Gas Phase

Guest Molecule

J-

o

Thymine

b/

(an Adenine Model) 0

(3) H-N

amount as a where Ammono is the adsorption monolayer, P and Psat are vapour pressure and the saturated vapour pressure of acetic acid, and C is the ratio of binding constants of the monolayer and multilayer adsorption. The ratio of the monolayer and the multilayer adsorption (C) was calculated to be 62,

Adenine

5

(a Thymine model)

HS-(CH2)&H3 3

0

-

NY \ /

Figure 4 Schematic illustration of binding of nucleo base models from the gas phase onto nucleo base monolayers immobilized on a highly sensitive 63-MHz QCM

SUPRAMOLECULAR

SCIENCE Volume 3 Numbers

l-3

1996

167

Molecular

recognition

on monolayers:

Y. Okahata et al. Table 1 Kinetic parameters for binding of guest molecules from gas phase onto the host monolayers at 25” C

J’ C)

0 - .. .....+.,

Host membranes

-0

‘;‘^““,...........,. ..........~....,.. *.........a..* a.... ..,,..... .-.-+-......_..._ +%...*“...+. ......“‘...._.....C ....IT 1

Guests

.a.

-lO-

kl I lC+ M-Is-1

k-1 I lo-2 s-1

190

026 .

Kl /l@ M-1

‘:

8

-20 -

s

-30 -

\(b)

is? -52. C**Z..*..,_ .-. **%_., ..__J

-40 1

_2

I

0

I

I

20

40

-3

‘-ii-&

(Adenine model)

: +ja&

(a) -4 “-w .. .... _.. -5 I 1 60 80

21 0

2.7

7.8

5.4

3.3

(Thymine model)

w

Time I set

---------__------------------

Figure 5

Time courses of frequency changes of the QCM immobilized with the thymine monolayer 1. (a) 2-Aminopyridine, (b) ybutyrolactam and (c) aniline. Gas-phase concentration of guest was 1.4 x 10m6M at 25°C

well onto the thymine monolayer 1 probably due to complementary two-point hydrogen bonding ability with the host membrane. Despite having two-point hydrogen bonding ability like thymine, y-butyrolactam (a thymine model) hardly adsorbed onto the thymine monolayer 1 as easily as aniline, which binds by one-point hydrogen bond with the thymine membrane. The time series of the binding behaviour of guest molecules onto a host membrane is expressed by the following equations: [host] + [guest] $

730

230

25

9.2

57

2.1 -_ -----

------_n--------------

D-CH3

7.9

3

00i

22 6.8

1.2

6.6

6.7 4.7

[host/guest]

Am, = [host/guest], = Am, { 1 - exp(--t/z)}

(5)

where z-’ = k,[guest] + k_l

(6)

The selectivity (about lo2 times) for K, values of the thymine monolayer 1 with 2-aminopyridine in the gas phase is comparable to the selectivity reported for association of 1-cyclohexyluracil (U) and g-ethyladenotine (A) in CHC13 obtained by IR spectroscopy and Kuu = 6.1 M-’ at 25”C)16. (&,, = 40M-’ However, binding constants obtained in the gas phase ( lo7 M-‘) are very large compared with those obtained in organic media ( lo2 M-l). This is probably due to the absence of solvent effects and large van der Waals forces in the gas phase.

The binding and dissociation rate constants (kl and were obtained from equation (6) at several different concentrations of guest molecules. Association constants (K,) were obtained from kl/k_l, and kinetic parameters are summarized in Table I. Complementary binding between the thymine monolayer 1 and 2-aminopyridine (an adenine model) showed about a 100 times larger K, value than other host-guest combinations. This large selectivity of the K, value is due to the larger binding rate constant (k,) and the smaller dissociation rate constant (k-,) than those of non-selective bindings. Similar selective binding kinetics were observed in the combination between the adenine monolayer 2 and a thymine model of y-butyrolactam, although the difference in K, value is not so large (5-20 times). When the self-assembled monolayer of 1-decanethiol 3 was used, all the guest molecules examined were hardly bound to this alkane membrane. These results clearly indicate that guest molecules selectively bind to the nucleic acid base membrane by using complementary hydrogen bonding, even in the gas phase.

We observed selecting binding behaviour and kinetics of small guest molecules onto the surface of functional self-assembled monolayers immobilized on a sensitive QCM plate in the gas phase. It is interesting that selective binding based on hydrogen bonds can be observed even in the gas phase. Molecular recognition in gas phase would be the most simple and fundamental system devoid of solvents effects. The highly sensitive QCM system is useful and will become a new tool for kinetic studies of molecular recognition on the membrane, especially for small guest molecules in the gas phase.

166

l-3

k-l, respectively)

SUPRAMOLECULAR

SCIENCE Volume 3 Number

CONCLUSION

1996

Molecular

recognition

on monolayers:

Y. Okahata et al.

Dey, M., Moritz, F., Grotemeyer, J. and Schlag, E.W. J. Am.

REFERENCES

Chem. Sot. 1994,116,9211

a) Rebek, Jr., J. Angew. Chem. Znt. Ed. Engl. 1990, 29, 245; b) Hamilton, A.D. J. Chem. Educ. 1990. 67. 821: c) Hamilton, A.‘D. Supramol. Chem. 1993, 1, 247, d) Huang: C.-Y., Cabell, L.A. and )4nslyn, E.V. J. Am. Chem. Sot. 1994, 116, 2778; e) Hayashi, T., Asai, T., Hokazono, H. and Ogoshi, H. J. Am. Chem. Sot. 1993, 115, 12210; f) Furuta, H., Magda, D. and Sessler, J.L. J. Am. Chem. Sot. 1991, 113, 978; g) Zimmerman, SC., Wu, W. and Zeng, Z. J. Am. Chem. Sot. 1991. 113. 196: h) Inouve. M.. Kim. K. and Kitano, T. J. Am. C&m. ioc.’ 1992,.114, 778; i) ‘Cochran, J.E., Parrott, T.J., Whithxk, B.J. and Whitlock, H.W. J. Am. Chem. Sot. 1992, 114, 2269 a) Rose, P.L., Harvey, N.G. and Amett, E.M. Adv. Phys. Org. Chem. 1993, 28, 45; b) Kurihara, K., Ohto, K., Tanaka, K., Aoyama, Y. and Kunitake, T. J. Am. Chem. Sot. 1991, 113, 444; c) Kurihara, K., Ohto, K., Honda, Y. and Kunitake, T. J. Am. Chem. Sot. 1991, 113, 5077; d) Ikeura, Y., Kurihara, K. and Kunitake, T. J. Am. Chem. Sot. 1991, 113, 7342; e) Sasaki, D.Y., Kurihara, K. and Kunitake, T. J. Am. Chem. Sot. 1991, 113, 9685; f) Sasaki, D.Y., Kurihara, K. and Kunitake, T. J. Am. Chem. Sot. 1992, 114, 10994; g) Ludwig, R., Ariga, K. and Shinkai, S. Chem. Lett. 1993, 1413; h) Kawabata, H. and Shinkai. S. Chem. Lett. 1994. 375: i) Ahuia. R.. Caruso. P.L.. Mobius; D., Paulus, W., Ringsdorf, H.- and’wildburg, G: Angew. Chem. Znt. Ed. Eng. 1992, 32, 1033; j) Weinbach, S.P., Jacquemain, D., Leveiller, F., Kjaer, K., Als-Nielsen, J. and Leiserowitz, L. J. Am. Chem. Sot. 1993, 115, 11110; k) Motesharei, K. and Myles, D.C. J. Am. Chem. Sot. 1994, 116, 7413; 1) Roush, J.A., Thacker, D.L. and Anderson, M.R. Langmuir 1994,10, 1642 a) Sun, L., Kepley, L.J. and Crooks, R.M. Langmuir 1992, 8, 2101; b) Xu, C., Sun, L., Kepley, L.J. Crooks, R.M. and Ricco, A.J. Analyt. Chem. 1993, 65, 2102; c) Sun, L., Crooks, R.M. and Ricco, A.J. Lungmuir 1993, 9, 1775 Schierbaum, K.D., Weiss, T., Thoden van Velzen, E.U., Engbersen, J.F.J., Reinhoudt, D.N. and GGpel, W. Science 1994,265, 1413

8 9

10 11

12

13 14

15 16

SUPRAMOLECULAR

Chu, I.-H., Dearden, D.V., Bradshaw, J.S., Huszthy, P. and Izatt, R.M. J. Am. Chem. Sot. 1993,115,4318 a) Okahata, Y. and Shimizu, 0. Langmuir 1987, 3, 1171; b) Okahata, Y., Shimizu, 0. and Ebato, H. Bull. Chem. Sot. Jpn 1990,63,3082 Guilbault, G.G. Analyt. Chem. 1983, 55, 1682 a) Ward, M.D. and Buttry, D.A. Science 1990, 249, 1000; b) Ebersole, R.C., Miller, J.A., Moran, J.R. and Ward, M.D. J. Am. Chem. Sot. 1990, 112, 3239 Redepenning, J., Schlesinger, T.K., Mechalke, E.J., Puleo, D.A. and Bizios, R. Analyt. Chem. 1993; 65,3378 a) Okahata, Y. and Ebato, H. J. Chem. Sot. Perkin Trans. 2 1991, 457; b) Okahata, Y., Matsunobu, Y., Ijiro, K., Mukae, M., Murakami, A. and Makino, K. J. Am. Chem. Sot. 1992, 114, 8299; c) Okahata, Y., Yasunaga, K. and Ogura, K. J. Chem. Sot. Chem. Commun. 1994, 469; d) Ebara, Y. and Okahata, Y. J. Am. Chem. Sot. 1994. 116., 11209 a) Folkers, J.P., Laibinis, P.E. and Whitesides, G.M. Lannmuir 1992. 8, 1330: b) Ulman. A.. Evans. SD.. Shindman, Y., Shartna, R:, Eilers, J.E.’ and’ Chang, ‘J.C. J: Am. Chem. Sot. 1991, 113, 1499; c) Nuzzo, R.G., Zegarski, B.R. and Dubois, L.H. J. Am. Chem. Sot. 1987, 109, 733; d) Zhang, L., Lu, T., Gokel, G.W. and Kaifer, A.E. Langmuir 1993, 9, 786 Sauerbrey, G. Z. Phys. 1959,155,206 a) Okahata, Y. and Ariga, K. J. Chem. Sot. Chem. Commun. 1987, 1535; b) Ariga, K. and Okahata, Y. Langmuir 1994, 10, 1535; c) Ariga, K. and Okahata, Y. J. Colloid Interface Sci. 1994, 167, 275; d) Ariga, K. and Okahata, Y. Langmuir 1994, 10, 2272: e) Okahata, Y.. Kimura. K. and Ariea. K. J. Am. &em. SW’ 1989; lli, 9190; f) dkahata, Y. aid Ebato, H. Analyt. Chem. 1989,61,2185 Brunauer, S., Emmett, P.H. and Teller, E. J. Am. Chem. Sot. 1938,60,309 Kyogoku, Y., Lord, R.C. and Rich, A. J. Am. Chem. Sot. 1967, 89,496

SCIENCE Volume 3 Numbers

l-3 1996

169