RESEARCH

Overview with selected publications.


1

MOLECULAR AND CONDENSED MATTER PHYSICS


1.1

Channeling of ions in solids (M.Sc. diploma work)

Channeling of ions in solids is a useful tool in condensed matter physics, e.g. to study lattice location of foreign atoms in solids, such as donor and acceptor atoms in semiconductors. In this work, the effects of nuclear and electronic scattering of channeling ions were studied theoretically in order to improve the applicability of the technique as an analytical tool. The work included experimental studies by means of an ion-accelerator at the Research Institute for Physics, Stockholm .

K. Björkqvist, B. Cartling and B. Domeij
Calculations on dechanneling of protons in Si and W
Rad. Effects 12, 267-276 (1972)

 

1.2

Localized description of the electronic structure of solids (Ph.D. thesis, part 1)

The possibility to control location of donor and acceptor atoms in semiconductors by means of the ion-implantation technique raised the need for new theoretical methods to determine the electronic properties of such systems. In this work, a localized description of the electronic structure of systems with broken symmetry was developed based on the definition of localized orbitals as eigenfunctions of the one-electron Hamiltonian for a finite cluster of atoms in a simulated crystalline environment. A local density approximation of exchange and correlation was employed, and one-electron orbitals were determined by a selfconsistent multiple-scattering theory. The method was applied to localized excitations in perfect solids, such as X-ray emissions, as well as to localized states in imperfect solids, such as those of donor and acceptor atoms in semiconductors. The main features of the density of states of the valence band, and also the energy gap between valence and conduction bands, of perfect covalent semiconductors were obtained in good agreement with observations. Apart from localized donor and acceptor states in the energy gap of doped semiconductors, new types of localized states below the valence band were predicted. This work was partly performed at the Center for Materials Science and Engineering, Massachusetts Institute of Technology, USA .

B. Cartling, B. Roos and U. Wahlgren
A model for self-consistent cluster calculations of the electronic structure of doped semiconductors by means of the SCF X a scattered wave method
Chem. Phys. Lett. 21, 380-384 (1973)

B. Cartling
Localized description of the electronic structure of covalent semiconductors:
I. Perfect crystals
J. Phys. C: Solid State Phys. 8, 3171-3182 (1975)

B. Cartling
Localized description of the electronic structure of covalent semiconductors:
II. Imperfect crystals
J. Phys. C: Solid State Phys. 8, 3183-3193 (1975)

B. Cartling
On the use of a localized description of the electronic structure of solids
Int. J. Quant. Chem. 9S, 117-121 (1975)

B. Cartling
Localized and relativistic descriptions of electronic structures
Ph.D. thesis, Department of Theoretical Physics, Royal Institute of Technology, Stockholm (1976)

 

1.3

Relativistic multiple-scattering theory of molecular electronic structures (Ph.D. thesis, part 2)

For molecular systems including heavy atoms, it is important to account for relativistic effects on the electronic structure. In this work, the molecular spinorbital concept was generalized to a molecular four-component spinor. The one-electron Dirac equations for the determination of optimal spinors were derived by a variational principle applied to a total energy functional. For the solution of the one-electron Dirac equations, the multiple-scattering formalism based on partial-wave expansions was relativistically generalized. The Green's function was evaluated in the representation of the free spherical solutions of the Dirac equation. The Green's function matrix elements describing propagation between scattering centers depend only on the geometric structure of the molecular system. The one-center scattering factors are given in terms of Dirac central field solutions. The ordinary, non-relativistic, multiple-scattering formalism is obtained as a special case in the limit of infinite velocity of light. This work was performed at the Quantum Theory Project, Department of Physics and Astronomy, University of Florida, Gainesville, USA .

B. Cartling and D. Whitmore
Multiple scattering formalism for a molecular Dirac equation
Chem. Phys. Lett. 35, 51-56 (1975)

B. Cartling and D. Whitmore
Relativistic molecular spinors by generalized multiple scattering theory
Int. J. Quant. Chem. 10, 393-412 (1976)

B. Cartling
Localized and relativistic descriptions of electronic structures
Ph.D. thesis, Department of   Theoretical Physics, Royal Institute of Technology, Stockholm (1976)

 

2

BIOLOGICAL PHYSICS


2.1

Mechanisms of biological energy conversions

Living systems are thermodynamically open systems with flows of energy and matter which underlie the emergence and maintenance of the highly complex structures characterizing these systems. The key physical questions regarding biological energy conversions concern inter- and intramolecular transfers of molecular energies without considerable dissipation. In both photosynthetic and oxidative energy conversions in plant and animal cells, respectively, chains of electron-transfer reactions are coupled to phosphorylation reactions that synthesize ATP-molecules for storage and subsequent utilization of energy. In this work, mechanisms based on electronic-conformational coupling in electron-transfer proteins were proposed and investigated. According to these mechanisms, the proteins involved operate as cyclic molecular machines. An important component of the proposed mechanisms is conformational gating of electron transfers to avoid dissipative processes. The prediction of intermediate redox states during the operating cycle of electron-transfer proteins prompted development of new types of experimental investigations, see Section 2.2. The description of conformational and electronic transitions required the development of new models of protein dynamics and electron-transfer reactions, see Section 2.3 and 2.4, respectively. The mechanisms studied are also of relevance to enzymatic reactions in general. Part of this work was performed at Department of Chemistry, Princeton University, USA and Division of Chemistry and Chemical Engineering, California Institute of Technology, USA .

B. Cartling and A. Ehrenberg
A molecular mechanism of the energetic coupling of a sequence of electron transfer reactions to endergonic reactions
Biophys. J. 23, 451-461 (1978)

B. Cartling and A. Ehrenberg
The energy transforming function associated with electron transfer reactions in biological systems
In Tunneling in Biological Systems, B. Chance, D. DeVault, H. Frauenfelder, R.A. Marcus, J.R. Schrieffer and N. Sutin, eds., Academic Press, New York, 575-580 (1979)

B. Cartling
A stochastic model of protein conformational dynamics and electronic-conformational coupling in biological energy transduction
J. Chem. Phys. 83, 5231-5241 (1985 )

 

2.2

Laser spectroscopy of protein dynamics and intermediates

To study the protein conformational transitions, as well as the intermediate protein redox states, that occur in the proposed mechanisms of biological energy conversions, see Section 2.1, various experimental techniques were developed in this work. The first investigation of the dynamics of a redox transition of an electron-transfer protein by time-resolved resonance Raman spectroscopy using a pulsed laser and vidicon-equipped spectrograph in combination with pulse-radiolytical reduction by means of a pulsed electron accelerator was performed on cytochrome c. A long-lived transient was observed upon reduction of the alkaline form of cytochrome c linked to replacement of the sixth ligand to the heme iron. This transient of cytochrome c was also studied by time-resolved resonance Raman spectrosopy in combination with continuous-flow chemical reduction. A technique to stabilize intermediate redox states of electron-transfer proteins at low temperature was also developed and permitted a detailed analysis of cytochrome c by resonance Raman spectroscopy. A new reaction sequence, involving generation of hydrated electrons as reductants by UV-laser pulse-photolysis, was proposed and allowed the first time-resolved studies of cytochrome c reduction in the ns time-regime. The results obtained by the different techniques are also included in an invited review chapter of the book "Biological Applications of Raman Spectroscopy". These works were performed at the Risö National Laboratory, Denmark; Department of Chemistry, University of York, England; Department of Chemistry, Princeton University, USA and the Regional Laser and Biotechnology Laboratories, University of Pennsylvania, Philadelphia, USA .

B. Cartling and R. Wilbrandt
Time-resolved resonance Raman spectroscopy of cytochrome c reduced by pulse radiolysis
Biochim. Biophys. Acta 637, 61-68 (1981)

M. Forster, R. Hester, B. Cartling and R. Wilbrandt
Continuous flow-resonance Raman spectroscopy of an intermediate redox state of cytochrome c
Biophys. J. 38, 111-116 (1982)

B. Cartling
Intermediate and stable redox states of cytochrome c studied by low temperature resonance Raman spectroscopy
Biophys. J. 43, 191-205 (1983)

B. Cartling, G. Holtom and T. Spiro
Photoelectron generation and transfer to cytochrome c studied by nanosecond transient absorption spectroscopy
J. Chem. Phys. 83, 3894-3905 (1985)

B. Cartling
Cytochrome c
In Biological Applications of Raman Spectroscopy, T.G. Spiro, ed., John Wiley, New York, vol. 3, 217-248 (1988)

 

2.3

Stochastic models of protein dynamics

The internal dynamics of proteins span a wide range of timescales, from fs to s. For the elucidation of the functions of proteins, it is important to be able to treat also the timescales beyond those accessible by direct molecular dynamics simulations. In this work, theories of stochastic processes were utilized to develop dynamics descriptions which cover all timescales. In the stochastic models, protein dynamics are decomposed into fluctuations in, and transitions between, conformational states. By a stochastic analysis of short-time fluctuations, the coupling strength between the reaction coordinate of a conformational transition and the bath that remaining degrees of freedom constitute is determined. The rate constant of a conformational transition is obtained from the nonstationary solutions of the Fokker-Planck equation, or Kramers´ equation, for Brownian motion of a particle in a potential and in contact with a heat bath. The methods developed, to obtain the probability density function and probability density current in the phase space for a large range of the coupling strength, are generally applicable to the kinetics of thermally activated processes. It was shown that protein dynamical variables referring to the normal coordinate system satisfy the stochastic equations of motion well, and that a reaction coordinate of a conformational transition is dominated by such a variable. For protein dynamics on still longer timescales, conformational relaxations are treated as diffusion processes described by the Smoluchowski equation.

B. Cartling
A stochastic model of protein conformational dynamics and electronic-conformational coupling in biological energy transduction
J. Chem. Phys. 83, 5231-5241 (1985)

B. Cartling
Kinetics of activated processes from nonstationary solutions of the Fokker-Planck equation for a bistable potential
J. Chem. Phys. 87, 2638-2648 (1987)

B. Cartling
Brownian motion model of activated transitions in a periodic potential
J. Chem. Phys. 90, 1819-1831 (1989)

B. Cartling
From short-time molecular dynamics to long-time stochastic dynamics of proteins
J. Chem. Phys. 91, 427-438 (1989)

B. Cartling
Stochastic model of intermode couplings in protein dynamics
J. Chem. Phys. 94, 6203-6210 (1991)

 

2.4

Electron transfer in biological systems

Electron-transfer reactions are fundamental in biological systems, e.g. in photosynthetic and oxidative energy-converting systems. A remarkable temperature dependence of an electron-transfer step in the reaction center of several photosynthetic bacteria has had a large impact on the development of electron-transfer theory. The electron transfer is nearly temperature-independent below a threshold and thermally activated above. In this work, a mechanism based on conformational control of electron transfer was proposed. Starting from the crystallographic structure of a photosynthetic reaction center, a detailed molecular mechanism was derived by studies of conformational states and electronic structure. The low temperature, unactivated, electron transfer is assigned to a direct, superexchange, mechanism involving intermediate orbitals on a bridging aromatic amino acid. In the high temperature, activated, region, hydrogen-bond formation and proton transfer between the bridging amino acid and another residue, in a high-energy conformational state of the reaction center, makes sequential electron transfer dominate the rate. This mechanism constitutes an electron-transfer switch since the electron-transfer rate differs by several orders of magnitude between the conformational states. The mechanism predicts the formation of a neutral radical of the bridging amino acid during electron transfer, which should be possible to detect by time-resolved spectroscopy. The influence of a bridging aromatic amino acid on long-range electron transfer is also of relevance to the design of artificial photosynthetic energy-converting systems.

B. Cartling
A mechanism of temperature dependent electron transfer reactions in biological systems
J. Chem. Phys. 95, 317-322 (1991)

B. Cartling
An electron transfer switch in photosynthetic reaction centra
Chem. Phys. Lett. 196, 128-132 (1992)

B. Cartling
A molecular mechanism of conformational gating of electron transfer in photosynthetic reaction centra
Biophys. Chem. 47, 123-138 (1993)

 

3

BIOLOGICAL NEURAL SYSTEMS


3.1

Biophysical models of neurons, synaptic interactions and neural microcircuits

Biophysical model neurons have been developed in terms of low numbers of dynamical variables. The model neurons incorporate neuronal adaptation, i.e. the coupling between neuronal activity and excitability that derives from the influence of intracellular calcium ions on the afterhyperpolarization phase of action potentials and thereby on the firing rate. It has been shown, that the model neurons closely reproduce the observed response characteristics of neocortical pyramidal cells and fast-spiking interneurons, two of the most frequent neuron types in the brain. By employing these models, particularly interesting results have been obtained regarding the role of neuronal adaptation in the generation and control of complex network dynamics, see Sections 3.2 and 3.3. Spiking versions of the low-dimensional adapting model neurons have also been formulated and shown to reproduce experimental observations. They allow exploration of the temporal aspects of neural coding, as discussed in Section 3.2.

Stochastic and reduced biophysical models of synaptic transmission have been developed and evaluated. In particular, facilitation of presynaptic release of neurotransmitter-containing vesicles due to intracellular calcium ions and depletion of readily releasable vesicles are incorporated. Both the stochastic and reduced biophysical models have been shown to display the principal dynamical characteristics of synaptic transmission experimentally observed. In particular, a stochastic model based on a master-equation formulation of synaptic transmission has permitted also the fluctuations of dynamical variables to be determined. The fluctuations are related to information-theoretic entropy, which has been utilized in an information-theoretic analysis of neural information transmission. The models of synaptic transmission have also been applied to studies of neural coding, see Secion 3.2.

Models of neocortical microcircuit systems have been formulated and employed in studies of neuromodulatory control of dynamics. The models are based on recent observations regarding reciprocal connections between pyramidal cells and inhibitory interneurons in cat and rat neocortices. A new type of activity-dependent short-term depression of synaptic couplings is also incorporated. It has recently been observed that back-propagating action potentials in interneuron dendrites in rat neocortex cause an activity-dependent release of GABA as a retrograde messenger depressing synaptic excitation of interneurons by pyramidal cells. Similarly, it has been observed that activity-dependent dendritic release of glutamate as a retrograde messenger from pyramidal cells suppresses inhibitory transmission between interneurons and pyramidal cells. The different dynamical modes of the neocortical microcircuit systems are linked to different functional modes, and the derived influence of neuromodulators on these are in accordance with many observations.

B. Cartling
A generalized neuronal activation function derived from ion-channel characteristics
Network 6, 389-401 (1995)

B. Cartling
Response characteristics of a low-dimensional model neuron
Neural Comput. 8, 1643-1652 (1996)

B. Cartling
A low-dimensional, time-resolved and adapting model neuron
Int. J. Neural Syst. 7, 237-246 (1996)

B. Cartling
Stochastic and reduced biophysical models of synaptic transmission
J. Biol. Phys. 26, 113-131 (2000)

B. Cartling
Control of neural information transmission by synaptic dynamics
J. Theor. Biol. 214, 275-292 (2002)

B. Cartling
Neuromodulatory control of neocortical microcircuits with activity-dependent short-term synaptic depression
J. Biol. Physics 30, 261-284 (2004)

 

3.2

Complex dynamics of neural systems. Neural coding and information processing

By means of the biophysical model neurons, it has been demonstrated that the complexity of neural network dynamics can be controlled by the adaptivity of excitatory neurons. At strong adaptivity, i.e. when the neuronal excitability is strongly influenced by the preceding activity, complex dynamics of either aperiodic, including chaotic, or limit-cycle character are generated. This regime corresponds to an exploratory mode of the system, in which the neural representation space can be searched. The frequency of limit-cycle dynamics is controlled by the adaptivity and is in the range of theta rhythms observed in the brain. At weak adaptivity, the dynamics are governed by fixed-point attractors in the neural representation space, and this corresponds to a mode for retrieval of particular patterns. The neuronal adaptivity in turn can be controlled by various neuromodulators in the brain. In particular, the model neural networks have been extended to also incorporate activity-dependent release of neuromodulators. It has been shown that an autonomously controlled sequence of bifurcations, from an initial exploratory to a final retrieval phase of an associative process, can result from such a mechanism.

By means of the biophysical models of synaptic transmission developed in this work (Section 3.1), the conditions for the two principal types of neural coding, i.e. rate and temporal coding, have been investigated. A rate coding based on the variation of the steady-state average level of postsynaptic membrane potential with presynaptic firing rate is possible at a low release probability, where the amplitude of the postsynaptic membrane potential response remains nearly constant for varying presynaptic firing rate. For higher release probabilities, the amplitude of the postsynaptic membrane potential is inversely proportional to the presynaptic firing rate above a limiting frequency. The resulting average level of the postsynaptic potential then remains nearly constant. In this limit, a temporal coding based on a phasic response of the postsynaptic neuron to a presynaptic action potential train can be realized. The synaptic information transmission has also been analyzed in terms of information-theoretic concepts. Recently, a new type of activity-dependent short-term depression of synaptic couplings in neocortical microcircuits has been incorporated.

B. Cartling
Control of the complexity of associative memory dynamics by neuronal adaptation
Int. J. Neural Syst. 4, 129-141 (1993)

B. Cartling
Generation of associative processes in a neural network with realistic features of architecture and units
Int. J. Neural Syst. 5, 181-194 (1994)

B. Cartling
Autonomous neuromodulatory control of associative processes
Network 6, 247-260 (1995)

B. Cartling
Dynamics control of semantic processes in a hierarchical associative memory
Biol. Cybern. 74, 63-71 (1996)

B. Cartling
Control of computational dynamics of coupled integrate-and-fire neurons
Biol. Cybern. 76, 383-395 (1997)

B. Cartling
A neural mechanism of the generation of meaning in cognitive processes
Behav. Brain Res. 87, 49-58 (1997)

B. Cartling
Control of resolution and perception in working memory
Behav. Brain Res. 100, 255-271 (1999)

B. Cartling
Stochastic and reduced biophysical models of synaptic transmission
J. Biol. Phys. 26, 113-131 (2000)

B. Cartling
Neuromodulatory control of interacting medial temporal lobe and neocortex in memory consolidation and working memory
Behav. Brain Res. 126, 65-80 (2001)

B. Cartling
Control of neural information transmission by synaptic dynamics
J. Theor. Biol. 214, 275-292 (2002)

B. Cartling
Neuromodulatory control of neocortical microcircuits with activity-dependent short-term synaptic depression
J. Biol. Physics 30, 261-284 (2004)

 

3.3

Cognitive processes: perception, learning, memory and language

In an investigation of working memory, retrieval from long-term memory to working memory has first been considered. Types and values of object features are represented by superassemblies and assemblies of neurons, respectively, possibly corresponding to cortical hypercolumns and minicolumns in the brain. Resolution is based on temporal segmentation of different items, and it has been shown that the observed capacity of working memory with respect to number of items resolved and also the observed temporal separation of different items can be accounted for by the model. Perception of an object is based on sufficiently intense coactivation of the neural representatives of the different features of an object. The temporal segmentation derives from the adaptation of excitatory neurons, and can therefore be controlled by neuromodulators in the brain regulating neuronal adaptivity. Neuromodulation of neural network dynamics is described in Section 3.2. A retrieval process thus can be controlled between exploratory and ultimate retrieval modes.

The above model of memory retrieval has been extended to also describe consolidation of long-term memories. A simplified model of the interacting medial temporal lobe and neocortical association areas has been formulated. It has been shown that memory consolidation by long-term potentiation of synaptic couplings, based on repeated activations of neocortical patterns, may be guided by neuromodulated dynamics of the medial temporal lobe via short-term couplings acting as pointers. Bifurcations, in this case with respect to developing synaptic couplings, have been shown to depend on the adaptivity of excitatory neurons in the medial temporal lobe and thus to be under neuromodulatory control. At weak adaptivity, after an initial temporal segmentation of several objects, accounting for the capacity of working memory to resolve several items, attention is selectively focused on a single object according to the model. At intermediate adaptivity, reactivations may persist and long-term synaptic couplings gradually develop. At strong adaptivity, the model predicts attention and memory consolidation to be subsequently terminated. The neuromodulatory control of the interacting medial temporal lobe and neocortical system via the adaptivity of excitatory neurons may account for several observations on the influence of neuromodulators on various cognitive processes and brain disorders.

Structured temporal sequences, in which the structure is derived from certain rules, occur in many cognitive contexts, such as human communication via language or music. For efficient cognition and interaction, it is important that individuals understand the structural rules. These are often acquired implicitly by the human brain from a stream of sequences. As an example, a child's learning of a language relies on an implicit extraction of the underlying grammar from hearing sentences. The specific objectives of this work are to devise and investigate neurobiologically plausible circuits for the representation and processing of rule-derived temporal sequences. The investigations primarily refer to human language acquisition, involving the neural inference and representation of grammar underlying language comprehension and production. Language processing also has a deeper significance in the sense that a mathematical problem can be formulated as the question whether a certain symbol string is accepted by a language or not. The neurobiological principles underlying the recognition or generation of a string, which conforms with the rules defining a language, thus may have a bearing on the general problem-solving (or computational or intelligent) capacity of the brain. Knowledge of the neurophysiological mechanisms of cognition may be helpful for improving the treatment of various brain disorders and for understanding the developmental situation. It may also contribute to the development of biologically inspired technical applications, such as brain-like information processing and decision-making.

B. Cartling
Control of the complexity of associative memory dynamics by neuronal adaptation
Int. J. Neural Syst. 4, 129-141 (1993)

B. Cartling
Generation of associative processes in a neural network with realistic features of architecture and units
Int. J. Neural Syst. 5, 181-194 (1994)

B. Cartling
Autonomous neuromodulatory control of associative processes
Network 6, 247-260 (1995)

B. Cartling
Dynamics control of semantic processes in a hierarchical associative memory
Biol. Cybern. 74, 63-71 (1996)

B. Cartling
Control of computational dynamics of coupled integrate-and-fire neurons
Biol. Cybern. 76, 383-395 (1997)

B. Cartling
A neural mechanism of the generation of meaning in cognitive processes
Behav. Brain Res. 87, 49-58 (1997)

B. Cartling
Control of resolution and perception in working memory
Behav. Brain Res. 100, 255-271 (1999)

B. Cartling
Neuromodulatory control of interacting medial temporal lobe and neocortex in memory consolidation and working memory
Behav. Brain Res. 126, 65-80 (2001)

B. Cartling
Neuromodulatory control of neocortical microcircuits with activity-dependent short-term synaptic depression
J. Biol. Physics 30, 261-284 (2004)

B. Cartling
On the implicit acquisition of a context-free grammar by a simple recurrent neural network
Neurocomputing 71, 1527-1537 (2008)

 

4

OTHER


4.1

Modulation of the piano tone

An analysis of the digitally recorded sound of piano tones, that permits a high temporal resolution of the frequency time-dependence, has revealed modulation of the piano tone. The modulation is of a new type displaying beating character of both frequency and amplitude modulation. The modulations have been shown to derive from weak coexcitation of damped adjacent tones due to coupling via the bridge. The frequency and amplitude modulation of the sound resulting from coexcitation of one strong and one or two weak tones has also been analyzed theoretically. One weak tone causes frequency and amplitude modulation of the sound, and two weak tones produce beating frequency and amplitude modulation, where the beatings of the two modulations are of opposite phase. The audibility of the observed frequency and amplitude modulation has been discussed in terms of previously determined detection thresholds. The beating character of both frequency and amplitude modulations, however, distinguishes the phenomena from those previously studied and prompts further psychoacoustic investigations. It has been shown that detuning of unison strings may significantly increase the frequency deviation of the frequency modulation in conjunction with affected amplitude modulation. The modulatory effects of coupling to adjacent tones therefore may possibly be utilized in the tuning process. A coupling of tones analogous to the situation in a piano may arise in other stringed musical instruments transferring string vibrations to a soundboard via a bridge. Prof. Anders Askenfelt, KTH is gratefully acknowledged for valuable and stimulating discussions.

B. Cartling
Beating frequency and amplitude modulation of the piano tone due to coupling of tones
J. Acoust. Soc. Am. 117, 2259-2267 (2005)

 

 

Links to descriptions of current research projects at KTH:

Low-dimensional biophysical models of neurons, synapses and neural microcircuits

Complex dynamics of neural systems. Neural coding and information processing

Cognitive processes, working memory, memory consolidation and language