Molecular and Biomolecular Electronics - American Chemical Society

---

1

Introduction to Molecular and Biomolecular Electronics

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

Robert R. Birge W . M. Keck Center for Molecular Electronics and Department of Chemistry, Syracuse University, 111 College Place, Syracuse, NY 13244 This chapter provides a brief introduction to the field of molecular and biomolecular electronics. The advantages and disadvantages of implementing molecular electronics are discussed in terms of selected characteristics including size, speed, architecture, quantum statistics, nanoscale engineering, stability, nonlinear properties, and reliability.

MOLECULAR ELECTRONICS IS AN INTERDISCIPLINARY FIELD that lies at the interface of chemistry, electrical engineering, optical engineering, and solid-state science. There are two schools of thought regarding molecular electronics. One school emphasizes nanotechnology as the driving force and views molecular electronics as one avenue for achieving nanoscale devices (1). T h e second school emphasizes the properties of molecules and the unique architectures and capabilities that molecular systems provide as the principal motivation (2). T h e difference is one of emphasis rather than of direction, and most recent conferences (3) and books (including this volume) have sought an accommodation of both perspectives. F o r the purposes of this discussion, molecular electronics is defined as the encoding, manipulation, and retrieval of information at a molecular or macromolecular level. This approach contrasts w i t h current commercial techniques, fast approaching their practical limits, i n w h i c h these functions are accomplished v i a lithographic manipulation of bulk materials to generate integrated circuits. M o l e c u l a r electronics not only represents the final technological stage i n the miniaturization of c o m puter circuitry, but it also provides promising new methodologies for high-speed signal processing and communication, novel associative and neural architectures, as w e l l as linear and nonlinear devices and m e m 0065-2393/94/0240-0001 $08.00/0 © 1994 American Chemical Society Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

2

MOLECULAR AND BIOMOLECULAR ELECTRONICS

ories. T h e ability to explore new architectures unique to molecularbased systems has a potential equal to that p r o v i d e d by molecular scale engineering and miniaturization. Biomolecular electronics is a subfield of molecular electronics that investigates the use of native as w e l l as modified biological molecules (chromophores, proteins, etc.) i n place of the organic molecules synthesized i n the laboratory. Because natural selection processes have often solved problems of a nature similar to those that must be solved i n harnessing organic compounds and because self-assembly and genetic engineering provide sophisticated control and manipulation of large m o l ecules, biomolecular electronics has shown considerable promise. There is no clear d i v i d i n g line separating molecular from biomolecular electronics. F o r example, i f a synthetic chemist incorporates a p o r p h y r i n molecule into a synthetic organic logic gate, is this an example of b i o molecular electronics? Such examples are commonplace, and for simplicity we w i l l typically use the adjective " m o l e c u l a r " to cover both synthetic and natural sources of organic compounds. T h e preceding discussion conveys some of the advantages but few of the problems inherent i n the implementation of molecular electronics. O n e of the best ways to introduce this field is to examine the potential advantages and disadvantages as outlined i n Table I. T h e list presented i n Table I is neither exhaustive nor orthogonal. First, many additional characteristics could have been included. Those listed i n Table I are selected to provide the broadest coverage w i t h a m i n i m u m number of categories. Second, the characteristics are i n some cases overlapping. F o r example, the reliability of a device is a function of the size and stability of the component molecules, the speed of the device and the quantum mechanical properties of the molecule or molecular ensemble. Nevertheless, the characteristics listed i n Table I serve to represent the principal issues that not only encourage but also challenge scientists seeking to implement molecular electronics. W e w i l l discuss each separately.

Size and Speed Molecules are synthesized from the "bottom u p " by carrying out additive synthesis starting w i t h readily available organic compounds. B u l k semiconductor devices are generated " f r o m the top d o w n " v i a lithographic manipulation of bulk materials. A synthetic chemist can selectively add an oxygen atom to a chromophore w i t h a precision that is far greater than a comparable oxidation step facilitated b y using electron-beam or X - r a y lithography. F i g u r e 1 shows molecular-based gates, w h i c h are typically 1000 times smaller than their semiconductor equivalents (3-7). A t the same time, such gates have yet to approach a comparable

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

Reliability

Nonlinear properties

Stability

Nanoscale engineering

Small size makes connection to control, input, and output circuitry difficult.

Small size of molecular scale offers high intrinsic speed. Picosecond switching rates are common. Neural, associative, and parallel architectures can be implemented directly. The quantum mechanical properties can be engineered with high precision. Synthetic organic chemistry, selfassembly, and genetic engineering provide nanometer resolution. Some molecules and proteins offer thermal and photochemical stabilities comparable to bulk semiconductors. Intrinsic second- and third-order properties of molecules can be synthetically optimized. Ensemble averaging via optical coupling or state assignment averaging provides high reliability.

Most molecules and proteins are photochemically or thermally labile precluding general application. Lifetimes and damage thresholds of molecularbased nonlinear optical devices are not yet competitve. Thermal or photochemical stress, impurity effects, and quantum statistics limit reliability of many systems.

Quantized behavior limits electron current densities and architectural flexibility. Nanolithography provides higher scale factors and flexibility than current molecular techniques.

Three terminal devices and standard logic designs are difficult to implement.

Current Disadvantages

Potential Advantages

Characteristics, Potential Advantages, and Current Disadvantages of Implementing Molecular Electronics

Quantized behavior

Architecture

Size-speed

Characteristic

Table I.

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

CA3

I

s*

4

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

MOLECULAR AND BIOMOLECULAR ELECTRONICS

0.001

donor-1

acceptor

micron

donor-2

b Figure 1. Examples of optically coupled charge-transfer molecular gates currently under investigation. The top gate (a) operates through light activation of an electron from the donor (s) to acceptor (s) and generates an electrostatic shift of the output chromophore absorption maximum from 520 to 590 nm if both inputs have been activated simultaneously (4). The sigma separators are used to increase the back-tunneling time to ~3 ps. The bottom gate (b) also operates via electron-transfer reactions and results in single or double reduction of the acceptor depending on the light intensity or wavelength (6). The size bars are approximate.

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

1.

5

BIRGE Introduction

level of reliability or to interconnect capability compared to their semi­ conductor counterparts. The signal propagation times of molecular gates are due i n large part to their small sizes. W h e t h e r the gate is designed to operate via electron transfer, electron tunneling, or conformational photochromism, a decrease i n size w i l l y i e l d a comparable increase i n speed. This effect is due to the fact that all gates i n use, under study, or envisioned are activated b y the shift i n the position of a charge carrier, and all charge carriers have mass. W h e t h e r the device is classical or relativistic, the mass of the carrier places a limit on h o w rapidly the conformational change can take place. Thus, size and speed are intimately related. O n e can criticize this view as arbitrarily restrictive i n that electrostatic changes can be generated via optical excitation, and the generation of an excited electronic state can occur within a large chromophore i n less than 1 fs (1 fs = 10~ s, the time it takes light to travel ~0.3 μπι). Nevertheless, the reaction of the system to the charge shift is still a sizedependent property, and the relationship between the total size of the device and the response time remains valid, A detailed discussion of electron motion and current densities i n molecules is provided b y Nafie (8). 15

T h e ultimate speed of a device is d e t e r m i n e d b y other factors as w e l l . A s noted b y L a w r e n c e a n d B i r g e (9), H e i s e n b e r g u n ­ certainty limits the m a x i m u m frequency o f operation, f , of a m o n oelectronic or monomolecular devices o n the basis o f the f o l l o w i n g relationship: m3LX

Jmax

0.00800801t; 7r s

=

Γ

hM 2, + 2 t a n " (-2) + In 1

/

r m a x

/ηττ \ 0.963t5 (GHz) «

2

/~2\

p|j

/ Κ Λ 2\ 1

- In

U

a

/

ν (lb)

s

/ Ί ΐ

N

where v is the energy separation of the two states of the device i n wavenumbers, and Ν is the number of state assignments that must be averaged to achieve reliable state assignment. This equation applies only to monoelectronic or monomolecular devices; Heisenberg's uncertainty principle permits higher frequencies for ensemble averaged devices. F o r example, i f a device requires 1000 state assignment averages to achieve reliability and v ^ 1000 c m , it w i l l have a maximum operating frequency of ~ 9 6 0 M H z . T h e concept of state assignment averaging is defined and quantitatively examined i n reference 9. Virtually all mono­ molecular or monoelectronic devices require Ν > 5 0 0 at ambient t e m ­ perature, but cryogenic devices operating at 1.2 Κ can approach Ν = 1. Thus, although molecular devices have an inherent advantage w i t h r e s

s

1

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

MOLECULAR AND BIOMOLECULAR ELECTRONICS

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

6

spect to speed, quantum mechanics places constraints on the maximum operating frequency that can have a significant impact at ambient temperatures. It is interesting to examine the trends i n bit size that have characterized the past few decades of memory development. T h e results are shown i n F i g u r e 2 and indicate that the area per bit has decreased l o g arithmically since the early 1970s (10). F o r comparison F i g u r e 2 also shows the cross-sectional area per bit calculated for the human brain (assuming one neuron is equivalent to one bit) and for proposed threedimensional memories and proposed molecular memories. A l t h o u g h current technology has surpassed the cross-sectional density of the h u man brain, the major advantage of the neural system of the brain is that information is stored i n three dimensions. A t present, the m i n d of a human being can store more " i n f o r m a t i o n " than the disk storage allocated to the largest supercomputers. O f course, the human brain is not digital, and such comparisons are tenuous. Nevertheless, the analogy underscores the fact that current memory technology is still anemic compared to the technology that is inherent i n the human brain. It also demonstrates the rationale for, and potential of, the development of three-dimensional memories ( J J , 12). W e can also conclude from an analysis of Figure 2 that the t r e n d i n memory densities w i l l soon force the bulk semiconductor industry to address some of the same issues that confront scientists who seek to implement molecular electronics. Figure 2. Analysis of the area in square micrometers required to store a single bit of information as a function of the evolution of computer technology in years. The data for magnetic disk, magnetic bubble, thin-film, and silicon dynamic random access (DRAM) memories are taken from Key es (10). These data are compared with the cross-sectional area per bit (neuron) for the human brain as well as anticipated areas and implementation times for optical three-dimensional memories (9,11,12) and molecular memories. The optical three-dimensional memory, the brain, and the molecular memories are threedimensional, and therefore the crosssectional area per bit is plotted for comparison. The area (A) is calculated in terms of the volume per bit, \/bit, by the formula A = (W) . m

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

1.

BIRGE Introduction

7

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

Architecture M o l e c u l a r electronics offers significant potential for exploring new architectures (2, 13). M a n y of the chapters i n this volume examine new architectures that are based on self-assembly (14-19). O t h e r chapters explore architectures that, although not unique to molecular systems, can be implemented w i t h greater flexibility or performance by using molecular materials (7, 11, 20-28). T h e opportunity to explore new architectures is one of the key aspects of molecular electronics that has prompted the enthusiasm of researchers. This enthusiasm is somewhat tempered, however, by the recognition that the three-terminal transistor that represents the fundamental b u i l d ing block of current computer gates and signal processing circuitry is difficult to implement using molecules. This p r o b l e m , w h i c h also applies to Josephson junction devices, has one of two potential consequences. It c o u l d limit the role that molecular electronics w i l l play i n enhancing current computer and signal processing systems. Alternatively, it c o u l d encourage the investigation and development of new designs based on neural, associative, or parallel architectures and lead to h y b r i d systems w i t h enhanced capabilities relative to current technology. T h e latter alternative is far more likely. F o r example, optical associative memories and three-dimensional memories can be implemented w i t h unique capabilities based on molecular electronics (9,11,12). Implementation of these memories within h y b r i d systems is anticipated to have near-term application. Furthermore, the human brain, a computer w i t h capabilities that far exceed the most advanced supercomputer, is a prime example of the potential of molecular electronics. T h e development of an artificial neural computer is beyond our current technology, but it w o u l d be illogical to assume that such an accomplishment is impossible. Thus, we should view molecular electronics as opening new architectural opportunities that w i l l lead to advances i n computer and signal processing systems.

Quantized Behavior As noted i n the chapter b y R e e d and Seabaugh (29), band-gap engineering and nanofabrication techniques have made possible a new class of quantum devices w i t h unique functionalities. Q u a n t u m devices have the potential for greatly reducing the complexity of circuits w h i l e s i multaneously increasing the maximum frequency of operation. T h e fact that scientists and engineers w o r k i n g on b u l k semiconductor gates have endorsed the potential of quantum devices is an indirect endorsement of molecular electronics. This position follows from a recognition that the quantum mechanical properties of molecules can be o p t i m i z e d for particular applications w i t h considerable precision and w i t h growing

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

8

MOLECULAR AND BIOMOLECULAR ELECTRONICS

sophistication. A majority of the candidate molecules and devices discussed i n this book function i n a quantized rather than i n a classical format (7-9, 11, 13-27, 30). Q u a n t i z e d behavior is not always advantageous, however. Molecules invariably respond to the addition or subtraction of an electron w i t h reorganization of the core electrons and the movement of the atoms i n response to bonding changes. This characteristic limits the electron current a molecule can carry and complicates the design of three-terminal devices that provide amplification. Thus, quantized behavior can limit architectural flexibility.

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

Nanoscale Engineering T h e feature size of high-speed semiconductor devices has decreased dramatically during the evolution of computer technology (see F i g u r e 2). D r i v e n by the demand for higher speeds and densities, micrometer and even submicrometer feature sizes are now commonplace. Ultraviolet lithography can provide modest improvement over current densities, but the evolution toward nanoscale feature sizes w i l l require electronbeam or X - r a y lithography. A l t h o u g h such lithography is w e l l understood, it is very expensive to implement. A s noted, organic synthesis provides a " b o t t o m u p " approach that offers a 100- to 1000-fold i m provement i n resolution relative to the best lithographic methods (6, 7, 20, 21). Organic synthesis has been developed to a high level of sophistication in large part as a result of the efforts of natural product synthetic chemists to recreate a priori the complex molecules that nature has developed v i a billions of years of natural selection. A sophisticated synthetic effort already exists w i t h i n the d r u g industry, and thus a commercially viable molecular electronic device could likely be generated in large quantities using present commercial facilities. T w o alternatives to organic synthesis have had a significant impact on current efforts i n molecular electronics: self-assembly and genetic engineering. T h e use of the L a n g m u i r - B l o d g e t t technique to prepare organized structures is the best k n o w n example of self-assembly (16,18, 19). H o w e v e r , self-assembly can also be used i n the generation of m e m brane-based devices (14,17), microtubule-based devices (15), and l i q u i d crystal holographic films (28). Genetic engineering offers a unique approach to the generation and manipulation of large biological molecules (12, 22, 23, 26, 30-33). T h e current methods and procedures of sitedirected mutagenesis are r e v i e w e d i n the chapter by Stayton et al. (26), and a number of chapters examine the application of this technique to bacteriorhodopsin-based devices (22, 23, 25, 30). Thus, molecular electronics provides at least three discrete methods of generating nanoscale devices: organic synthesis, self-assembly, and site-directed mutagenesis. T h e fact that the latter two methods currently

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

1.

9

BIRGE Introduction

offer access to larger and often more complicated structures has been responsible i n part for the early success of biomolecular electronics. A l l three techniques offer resolutions significantly better than those possible via b u l k lithography. H i g h resolution is not the only criterion i n examining the quality of nanoscale engineering. Lithography offers an advantage that none of the techniques available to molecular electronics can duplicate. L i t h ography can be used to construct very large scale integrated (VLSI) devices involving more than 100,000 discrete components w i t h complex interconnections. This ability can be quantitatively analyzed by defining the scale factor, a ratio defined as the overall area of the device d i v i d e d by the size of the discrete gates or transistors that make up the device. A typical V L S I circuit has a scale factor of approximately 1 0 . Despite the fact that organic synthesis offers convenient access to a threedimensional structure, the preparation of extremely large molecules is a significant challenge. A comparable scale factor for large organic m o l ecules is approximately 1 0 - 1 0 . Genetic engineering provides access to m u c h larger structures, and scale factors of 1 0 and even 1 0 are common. Nevertheless the use of amino acid b u i l d i n g blocks limits flexibility. Self-assembly expands the size still further, but at present, the scale factors are small due to the use of identical molecules. In conclusion, nanoscale semiconductor engineering still provides the best combination of scale factor and flexibility. 5

3

4

5

6

Stability One of the commonly stated advantages of bulk semiconductor materials relative to organic molecules is thermal stability. Silicon and gallium arsenide can operate at temperatures that exceed those w h i c h most m o l ecules can withstand for extended periods of time. H o w e v e r , many m o l ecules and proteins can operate at very h i g h temperatures, and some have thermal stabilities that exceed those evidenced by silicon and gall i u m arsenide. F u r t h e r m o r e , the use of ensemble averaging, i n w h i c h many molecules are used to simultaneously represent a single bit of information, enhances system stability by allowing some molecules to decompose without adversely affecting system reliability. Similar observations apply to photochemical stability, an issue relevant to optical computing and optical memories. F o r example, the protein bacteriorhodopsin, w h i c h is the light-transducing protein i n the salt marsh bacterium Halohacterium halobium, exhibits outstanding thermal and photochemical stability (5, 9, 12, 22, 23, 25, 30-33). This feature is due i n part to natural selection and the i n vivo requirement that this protein operate within a bacterium inhabiting a hot salt marsh under intense solar radiation. T h e chromophore phthalocyanine, w h i c h has many de-

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

10

MOLECULAR AND BIOMOLECULAR ELECTRONICS

vice applications, has even greater stability (27). In summary, thermal and photochemical stability is an important issue i n implementing molecular electronics, but organic and biological molecules can be designed w i t h stabilities more than adequate for device applications.

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

Nonlinear Properties Many optical and electronic devices make use of the nonlinear properties of the constituent materials (9, 11, 12, 21, 24, 32). A majority of the recent work i n this area has concentrated on nonlinear optical properties because of the importance of these properties to the design of optical communication systems, optical computing, and optical memories. F o u r chapters of this book are devoted to this important area (11, 20, 21, 24). O n e of the principal advantages of using organic molecules i n nonlinear optical applications is the ability to tailor the properties of the molecules to suit specific applications. Synthetic organic chemistry offers a level of flexibility i n optimizing the dipole moment, transition moments, electronic symmetry, and conjugation length of a candidate material that exceeds that inherent i n manipulation of bulk inorganic materials. T h e principal problems encountered w i t h present-day nonlinear optical molecular materials are associated w i t h transparency, damage threshold, and lifetime. Thus, although organic materials have been prepared w i t h second-order hyperpolarizabilities m u c h higher than lithium niobate, the latter inorganic material has found greater commercial application i n second-harmonic generation. Organic materials, however, are rapidly closing the gap (20, 21), and commercial viability is fast approaching reality.

Reliability T h e issue of the reliability of molecular electronic devices is discussed in detail i n a separate chapter i n this book (9), and therefore the present discussion w i l l be b r i e f and selective. T h e topic is extremely important, however, and has been used repeatedly by semiconductor scientists and engineers as a reason to view molecular electronics as impractical. I n deed, an article by Birge et al. (4) has been referenced by others to argue (incorrectly) that the need to use ensemble averaging i n optically coupled molecular gates and switches demonstrates the inherent u n reliability of molecular electronic devices. This point of view is comparable to suggesting that transistors are inherently unreliable because more than one charge carrier must be used to provide satisfactory performance. T h e majority of ambient-temperature molecular and bulk semiconductor devices use more than one molecule or charge carrier to represent a bit for two reasons: (1) ensemble averaging improves reliability, and (2) ensemble averaging permits higher speeds (34). T h e

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

1.

BmGE

11

Introduction

nominal use of ensemble averaging does not, however, rule out reliable monomolecular or monoelectronic devices. W e explore the issue of re­ liability i n ensemble averaged and monomolecular devices briefly. T h e probability of correctly assigning the state of a single molecule, P i , is never exactly unity. This less than perfect assignment capability is due to quantum effects as w e l l as inherent limitations i n the state assignment process. T h e probability of an error i n state assignment, Perror, is a function of pi and the number of molecules, n, w i t h i n the ensemble used to represent a single bit of information. P can be approximated by the following equation (9):

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

e r r o r

Perror(n, Pi) S " E r f

(2p! + l ) V n L4V2 (l P l

Pl

(2pi -

l)Vn

(2)

) ' 4 V 2 ( l - ρχ) Pl

where E r f [Z ; Z ] is the differential error function defined by 0

x

E r f [Z ; Z J = E r f [ Z J » E r f [Z ] 0

(3)

0

where E r f [ZJ = ^

JT E x p (-**)&

(4)

Equation 2 is approximate and neglects error associated w i t h the p r o b ­ ability that the number of molecules i n the correct conformation can stray from their expectation values on the basis of statistical consider­ ations. Nevertheless it is sufficient to demonstrate the issue of reliability and ensemble size. First, we define a logarithmic reliability parameter, ξ, w h i c h is related to the probability of error i n the measurement of the state of the ensemble (device) by the function, P rror = 10~*. A value of £ = 10 is considered a minimal requirement for reliability i n non-errorcorrecting digital architectures. If we assume that the state of a single molecule can be assigned correctly w i t h a probability of 9 0 % (p = 0.9), then equation 1 indicates that 95 molecules must collectively represent a single bit to y i e l d ξ > 10 [P rror(95, 0.9) s 8 Χ 1 0 " ] . W e must recognize that a value of p = 0.9 is larger than is normally observed, and some examples of reliability analyses for specific molecular-based devices is p r o v i d e d i n Chapter 6 (9). In general, ensembles larger than 1 0 are r e q u i r e d for reliability unless fault-tolerant or fault-correcting architectures can be i m p l e ­ mented. T h e question then arises whether or not we can design a reliable computer or memory that uses a single molecule to represent a bit of information. T h e answer is yes p r o v i d e d one of two conditions apply. T h e first condition is architectural. It is possible to design fault-tolerant e

x

e

11

x

3

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

12

MOLECULAR AND BIOMOLECULAR ELECTRONICS

architectures that either recover from digital errors or simply operate reliably w i t h occasional error due to analog or neural-like environments. A n example of digital error correction is the use of additional bits beyond the number r e q u i r e d to represent a number. This approach is common in semiconductor memories, and under most implementations these ad­ ditional bits provide for single-bit error correction and multiple-bit error detection. Such architectures lower the required value of £ to values less than 4. A n example of analog error tolerance is embodied i n many optical computer designs that use holographic or F o u r i e r architectures to carry out complex functions. T h e second condition is more subtle. It is possible to design molec­ ular architectures that can undergo a state reading process that does not disturb the state of the molecule. F o r example, an electrostatic switch could be designed that can be " r e a d " without changing the state of the switch. Alternatively, an optically coupled device can be read by using a wavelength that is absorbed or diffracted but that does not initiate state conversion. U n d e r these conditions, the variable η that appears i n equation 1 can be defined as the number of nondestructive read " o p ­ erations" rather than the ensemble size. Thus our previous example indicating that 95 molecules must be included i n the ensemble to achieve reliability can be restated as follows: A single molecule can be used provided we can carry out 95 nondestructive measurements to define the state. Multiple-state measurements are equivalent to integrated measurements and should not be interpreted as a start-read-stop cycle repeated η number of times. A continuous read w i t h digital or analog averaging can achieve the same level of reliability.

Conclusions This discussion has outlined the advantages and disadvantages of i m p l e ­ menting molecular electronics. W e have chosen to examine this issue w i t h reference to the selected characteristics of size, speed, architecture, quantum statistics, nanoscale engineering, stability, nonlinear properties, and reliability. T h e key conclusions are summarized i n Table I. If one ignores the issue of commercial viability, an enthusiastic endorsement of molecular electronics is justified on the basis of the exciting basic research that has characterized this field. Commercial viability, however, remains an open question i n many areas of proposed implementation. Nevertheless, molecular-based biosensors and chiral photonic materials have already achieved commercial status. Molecular-based holographic and nonlinear optical materials, microwave assemblies, and optical memories have also been demonstrated to have commercial viability. Liquid-crystal displays ( L C D s ) , one of the first commercially successful molecular electronic devices, are now common. W h e n first introduced,

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

1. BIRGE Introduction

13

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

LCDs were slow, thermally labile, inefficient and exhibited poor contrast and limited viewing angles. The dramatic improvement in characteristics and the resulting current commercial success of L C D s are due to an extensive research and development effort. We should not expect similar levels of success from the current molecular electronic device candidates without similar time and effort. Those companies willing to invest in molecular electronics research now may soon represent the dominant growth companies in biomedical, photonic, signal processing, commu­ nications, and computer technology. References 1. Nanotechnology: Research and Perspective; Crandall, B. C.; Lewis, J.; Eds.; Massachusetts Institute of Technology: Cambridge, MA, 1992. 2. Computer (Special issue on Molecular Computing; Conrad, M., Ed.) 1992, 25, 6-67. 3. Molecular Electronics—Science and Technology; Aviram,A.,Ed.;American Institute of Physics: New York, 1992; Vol. 262, p 334. 4. Birge, R. R.; Ware, B. R.; Dowben, P. Α.; Lawrence, A. F. In Molecular Electronics—Science and Technology; Aviram, Α., Ed.; Engineering Foun­ dation: New York, 1989; pp 275-284. 5. Birge, R. R. In Nanotechnology: Research and Perspective; Crandall, B. C.; Lewis, J., Eds.; Massachusetts Institute of Technology: Cambridge, MA, 1992; p 381. 6. O'Neal, M. P.; Niemczyk, M. P.; Svec, W. Α.; Gosztola, D.; Gaines, G. L.; Wasielewski, M. R. Science (Washington, D.C.) 1992, 257, 63-65. 7. Metzger, R. M. Chapter 5 in this volume. 8. Nafie, L. A. Chapter 4 in this volume. 9. Lawrence, A. F.; Birge, R. R. Chapter 6 in this volume. 10. Keyes, R. W. AIP Conf. Proc. 1992, 262, 285-297. 11. Dvornikov, A. S., Rentzepis, P. M. Chapter 7 in this volume. 12. Birge, R. R. Computer 1992, 25, 56-67. 13. Conrad, M. Chapter 3 in this volume. 14. Ottova-Leitmannova,Α.;Martynski, T.; Warkak, Α.; Ti Tien, H. Chapter 17 in this volume. 15. Shashidar, R.; Schnur, J. M. Chapter 18 in this volume. 16. Marx, Κ. Α.; Samuelson, L. Α.; Kamath,M.;Sengupta, S.; Kaplan, D.; Kumar, J.; Tripathy, S. Chapter 15 in this volume. 17. Fendler, J. H. Chapter 16 in this volume. 18. Albrecht, O.; Sakai, K.; Takomoto, K.; Matsuda, H.; Eguchi, K.; Nakagiri, T. Chapter 13 in this volume. 19. Fujihira, M. Chapter 14 in this volume. 20. Di Bella, S.; Fragala, I. L.; Ratner, Μ. Α.; Marks, T. J. Chapter 9 in this volume. 21. Gorman, C. B.; Van Doremaele, G. H. J.; Marder, S. R. Chapter 8 in this volume. 22. Hampp, N.; Thoma, R.; Zeisel, D.; Brüchle, C.; Österhelt, D. Chapter 21 in this volume. 23. Hong, F. T. Chapter 22 in this volume. 24. Pierce, Β. M. Chapter 10 in this volume. 25. Rayfield, G. Chapter 23 in this volume.

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.

14

MOLECULAR AND BIOMOLECULAR ELECTRONICS

Downloaded by 191.101.54.113 on October 8, 2016 | http://pubs.acs.org Publication Date: May 5, 1994 | doi: 10.1021/ba-1994-0240.ch001

26. Stayton, P. S.; Olinger, J. M.; Wollman, S. T.; Bohn, P. W.; Sligar, S. G. Chapter 19 in this volume. 27. Wada, T.; Hosoda, M.; Sasabe, H. Chapter 11 in this volume. 28. Zhang, J.; Sponsler, M. B. Chapter 12 in this volume. 29. Reed, M.; Seabaugh, A. C. Chapter 2 in this volume. 30. Lanyi, J. K. Chapter 20 in this volume. 31. Schick, G. Α.; Lawrence, A. F.; Birge, R. R. Trends Biotechnol. 1988, 6, 159-163. 32. Birge, R. R. Annu. Rev. Phys. Chem. 1990, 41, 683-733. 33. Oesterhelt, D.; Brauchle, C.; Hampp, N . Quart. Rev. Biophys. 1991, 24, 425-478. 34. Birge, R. R.; Lawrence, A. F.; Tallent, J. A. Nanotechnology 1991, 2, 73-87. RECEIVED for review March 25, 1993. ACCEPTED revised manuscript May 11, 1993.

Birge; Molecular and Biomolecular Electronics Advances in Chemistry; American Chemical Society: Washington, DC, 1994.